Citation
Butterfly migration through the Florida Peninsula

Material Information

Title:
Butterfly migration through the Florida Peninsula
Creator:
Lenczewski, Barbara, 1952-
Publication Date:
Language:
English
Physical Description:
viii, 132 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Butterflies ( jstor )
Coasts ( jstor )
Gulfs ( jstor )
Insects ( jstor )
Latitude ( jstor )
Peninsulas ( jstor )
Species ( jstor )
Statistical bias ( jstor )
Statistical significance ( jstor )
Sulfur ( jstor )
Butterflies -- Migration -- Florida ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 124-131).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Barbara Lenczewski.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
29238700 ( OCLC )
ocm29238700

Downloads

This item has the following downloads:

EIBUVQCHO_WFD7AX.xml

AA00011452_00001.pdf

00006.txt

AA00011452_00001_0095.txt

AA00011452_00001_0091.txt

00026.txt

AA00011452_00001_0059.txt

00047.txt

00080.txt

AA00011452_00001_0057.txt

00058.txt

AA00011452_00001_0034.txt

00105.txt

00060.txt

00054.txt

00092.txt

AA00011452_00001_0013.txt

AA00011452_00001_0046.txt

00051.txt

AA00011452_00001_0119.txt

AA00011452_00001_0022.txt

00055.txt

00061.txt

AA00011452_00001_0136.txt

00137.txt

AA00011452_00001_0011.txt

00067.txt

00142.txt

00037.txt

AA00011452_00001_0125.txt

AA00011452_00001_0088.txt

AA00011452_00001_0061.txt

00033.txt

00100.txt

AA00011452_00001_0116.txt

00096.txt

AA00011452_00001_0133.txt

AA00011452_00001_0001.txt

AA00011452_00001_0130.txt

00108.txt

AA00011452_00001_0047.txt

AA00011452_00001_0089.txt

AA00011452_00001_0045.txt

AA00011452_00001_0106.txt

AA00011452_00001_0002.txt

AA00011452_00001_0142.txt

00062.txt

00002.txt

00112.txt

00076.txt

AA00011452_00001_0014.txt

00057.txt

AA00011452_00001_0018.txt

AA00011452_00001_0008.txt

AA00011452_00001_0055.txt

AA00011452_00001_0052.txt

00087.txt

00066.txt

AA00011452_00001_0072.txt

00073.txt

00075.txt

AA00011452_00001_0115.txt

AA00011452_00001_0085.txt

AA00011452_00001_0077.txt

AA00011452_00001_0124.txt

00007.txt

AA00011452_00001_0042.txt

AA00011452_00001_0108.txt

00127.txt

AA00011452_00001_0041.txt

AA00011452_00001_0037.txt

00027.txt

00063.txt

AA00011452_00001_0098.txt

AA00011452_00001_0075.txt

AA00011452_00001_0101.txt

00114.txt

00091.txt

00071.txt

00120.txt

AA00011452_00001_0122.txt

00059.txt

AA00011452_00001_0127.txt

00136.txt

AA00011452_00001_0017.txt

AA00011452_00001_0069.txt

AA00011452_00001_0033.txt

00042.txt

00012.txt

00125.txt

00023.txt

AA00011452_00001_0058.txt

AA00011452_00001_0111.txt

00039.txt

AA00011452_00001_0026.txt

00122.txt

AA00011452_00001_0103.txt

AA00011452_00001_0084.txt

00133.txt

00072.txt

00081.txt

00020.txt

AA00011452_00001_0097.txt

00038.txt

AA00011452_00001_0023.txt

AA00011452_00001_0028.txt

AA00011452_00001_0114.txt

AA00011452_00001_0141.txt

00101.txt

00011.txt

AA00011452_00001_0140.txt

AA00011452_00001_0074.txt

00034.txt

00010.txt

00083.txt

00143.txt

00024.txt

AA00011452_00001_0038.txt

00110.txt

00093.txt

AA00011452_00001_0093.txt

00117.txt

AA00011452_00001_0016.txt

EIBUVQCHO_WFD7AX_xml.txt

00022.txt

00119.txt

00111.txt

AA00011452_00001_0080.txt

00019.txt

00126.txt

00135.txt

AA00011452_00001_0078.txt

AA00011452_00001_0063.txt

AA00011452_00001_0100.txt

AA00011452_00001_0029.txt

00070.txt

00032.txt

AA00011452_00001_0030.txt

00138.txt

00068.txt

00107.txt

AA00011452_00001_0056.txt

00128.txt

00140.txt

AA00011452_00001_0005.txt

AA00011452_00001_0051.txt

00064.txt

00008.txt

AA00011452_00001_0110.txt

AA00011452_00001_0019.txt

00035.txt

AA00011452_00001_0139.txt

AA00011452_00001_0096.txt

00095.txt

AA00011452_00001_pdf.txt

AA00011452_00001_0120.txt

AA00011452_00001_0131.txt

AA00011452_00001_0143.txt

00090.txt

AA00011452_00001_0068.txt

00016.txt

AA00011452_00001_0039.txt

AA00011452_00001_0040.txt

00116.txt

00118.txt

00005.txt

AA00011452_00001_0135.txt

00103.txt

AA00011452_00001_0134.txt

00017.txt

00139.txt

AA00011452_00001_0035.txt

00097.txt

AA00011452_00001_0060.txt

00050.txt

00121.txt

00085.txt

00018.txt

AA00011452_00001_0086.txt

AA00011452_00001_0006.txt

00098.txt

AA00011452_00001_0132.txt

AA00011452_00001_0031.txt

00113.txt

00052.txt

AA00011452_00001_0099.txt

00144.txt

AA00011452_00001_0009.txt

00084.txt

AA00011452_00001_0062.txt

00069.txt

00134.txt

00004.txt

AA00011452_00001_0113.txt

AA00011452_00001_0102.txt

AA00011452_00001_0105.txt

00088.txt

AA00011452_00001_0126.txt

AA00011452_00001_0121.txt

AA00011452_00001_0007.txt

AA00011452_00001_0066.txt

AA00011452_00001_0081.txt

AA00011452_00001_0109.txt

AA00011452_00001_0104.txt

00029.txt

AA00011452_00001_0087.txt

AA00011452_00001_0070.txt

AA00011452_00001_0079.txt

00074.txt

00132.txt

AA00011452_00001_0044.txt

AA00011452_00001_0128.txt

AA00011452_00001_0036.txt

AA00011452_00001_0117.txt

00077.txt

AA00011452_00001_0083.txt

AA00011452_00001_0004.txt

00041.txt

AA00011452_00001_0094.txt

00053.txt

AA00011452_00001_0049.txt

AA00011452_00001_0092.txt

AA00011452_00001_0020.txt

00104.txt

00115.txt

00078.txt

00141.txt

AA00011452_00001_0123.txt

00131.txt

00021.txt

AA00011452_00001_0067.txt

00028.txt

AA00011452_00001_0054.txt

AA00011452_00001_0024.txt

00031.txt

00009.txt

00046.txt

AA00011452_00001_0043.txt

AA00011452_00001_0090.txt

AA00011452_00001_0015.txt

00044.txt

AA00011452_00001_0073.txt

00013.txt

AA00011452_00001_0012.txt

AA00011452_00001_0032.txt

AA00011452_00001_0027.txt

00109.txt

00099.txt

00102.txt

00040.txt

AA00011452_00001_0082.txt

00129.txt

00094.txt

00014.txt

00086.txt

00130.txt

00049.txt

00079.txt

00048.txt

AA00011452_00001_0003.txt

AA00011452_00001_0021.txt

00123.txt

AA00011452_00001_0112.txt

00065.txt

00106.txt

AA00011452_00001_0071.txt

00015.txt

AA00011452_00001_0053.txt

00056.txt

00045.txt

AA00011452_00001_0048.txt

AA00011452_00001_0107.txt

AA00011452_00001_0065.txt

AA00011452_00001_0118.txt

AA00011452_00001_0064.txt

AA00011452_00001_0138.txt

AA00011452_00001_0010.txt

00030.txt

AA00011452_00001_0129.txt

AA00011452_00001_0050.txt

00089.txt

00082.txt

AA00011452_00001_0137.txt

AA00011452_00001_0025.txt

00036.txt

00124.txt

00043.txt

00025.txt

AA00011452_00001_0076.txt

00003.txt


Full Text
/T/V
BUTTERFLY MIGRATION THROUGH THE FLORIDA PENINSULA
By
BARBARA LENCZEWSKI
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
1992


This work is lovingly dedicated to my parents, Lucjan and Stanislawa Lenczewski, who have waited half their lives to see it and to my husband, Joseph Jowers, for believing that he would.


ACKNOWLEDGEMENTS
A number of people were helpful during this research. Foremost, I would like to thank my major professor, Dr. Thomas J. Walker, for supplying the materials and workshop used in construction of the portable traps and for providing expertise and labor during their construction. The funds for travel and other expenses were arranged by Dr. Walker and he provided a great deal of enthusiasm and encouragement for this project throughout many years. I am also indebted to Dr. J. J. H. Frank, who, through the Mole Cricket Biocontrol Project, provided the opportunity for a graduate assistantship and facilitated the loan of travel vehicles. Dr. Lincoln P. Brower provided many opportunities for "butterfly" socializing, and the use of his property for mark-release-recapture studies, during which he kindly made many valuable observations. Dr. Brower and Dr. Frank Slansky generously allowed me the use of their laboratories and equipment. I also thank the other members of my committee, Drs. J. E. Lloyd and J. L. Nation, for their helpful suggestions and a critical review of the manuscript.
The following people made the trapping study feasible by providing locations for, and assistance with, flight traps. I am particularly grateful to Dr. J. Whitesell, Valdosta State College; Dr. G. W. Ellstrom and his assistant, Ms. Annette Chandler, JFAS/AREC Leesburg; Dr. Mohammed Abou-Setta and Dr. H. N. Nigg, IFAS/CREC; and Dr. Mark Deyrup, Archbold Biological Station. Special thanks go to Dr. Whitesell
iii


for field assistance with traps, Dr. Nigg for a tour of the Green Swamp and Dr. Deyrup for help during hurricanes and otherwise.
My first interest in butterflies and their migration through the Florida Peninsula was inspired by the late Dr. Dennis Leston, in whose memory this work is completed. His enthusiasm for nature and love for insects will remain with me always. I am greatly indebted to my parents, Lucjan and Stanislawa Lenczewski, who have encouraged my education. My sister, Eva Lenczewski, despite her dislike of insects, has sometimes even helped with this perversity of mine and she deserves recognition for her courage. Finally, and most importantly, I thank my husband, Joseph Jowers, for a great deal of assistance and companionship in the field. He and my son, Justin, held down the home front to really make my work possible.
iv


TABLE OF CONTENTS
ACKNOWLEDGEMENTS ......................................iii
ABSTRACT ...............................................vii
CHAPTER 1. INTRODUCTION.................................. 1
Literature Review ....................................... 1
General Migration................................... 1
Florida Butterfly Migration ............................. S
Research Plan.......................................... 10
Biology of the Butterflies................................... 11
Phoebis sennae..................................... 11
Agraulis vanillae.................................... 15
yjrJarms pipjejis.................................... 17
Precis penia...................................... 18
Other Migrants..................................... 20
CHAPTER 2. FLIGHT DIRECTION ...............................23
Introduction...........................................23
Materials and Methods ....................................25
Flight Azimuths....................................25
Mark-Release-Recapture...............................26
Results and Discussion ....................................26
Mark-Release-Recapture...............................37
CHAPTER 3. NUMBERS OF MIGRANTS ...........................38
Introduction...........................................38
Materials and Methods ....................................39
Latitudinal Pole Counts................................39
Migration Profiles...................................41
Results and Discussion....................................44
Latitudinal Pole Counts................................44
v


Cloudless sulphur............................... 44
Gulf fritillary ................................. 46
Long tailed skipper.............................. 46
Migration Profiles................................... 49
Cloudless sulphur............................... 49
Gulf fritillary ................................. 58
Numbers of Migrants................................. 66
CHAPTER 4. PHENOLOGY OF MOVEMENT ........................77
Introduction...........................................77
Materials and Methods ....................................80
Longitudinal Pole Counts ..............................80
Portable Flight Traps.................................81
Results and Discussion ....................................83
Cloudless Sulphur...................................83
Longitudinal pole counts ..........................83
Flight trap catches ..............................85
Gulf Fritillary .....................................93
Longtiudinal pole counts ..........................93
Flight trap catches ..............................96
Long Tailed Skipper ................................102
Longitudinal pole counts ......................... 102
Flight trap catches .............................105
Buckeye........................................112
Flight trap catches .............................112
CHAPTER 5. SUMMARY AND DISCUSSION........................ 118
REFERENCES ............................................124
BIOGRAPHICAL SKETCH.....................................132
vi


Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
BUTTERFLY MIGRATION THROUGH THE FLORIDA PENINSULA
By
Barbara Lenczewski December 1992
Chairman: Dr. Thomas J. Walker
Major Department: Entomology and Nematology
At least eight species of butterflies migrate southward during fall of each year
through the Florida peninsula. The four major migrants, representing four families of
Lepidoptera, are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae
(L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long tailed
skipper; and Precis coenia (Hubner) (Nymphalidae), the buckeye. Drive counts of the
net southward movement of the cloudless sulphur and gulf fritillary were made across
north central Florida at the latitude of Gainesville (29.65) during fall 1986-1988. This
migration profile, or cross section through the migratory stream, was used to estimate
the density of individuals moving south into central Florida annually. The highest
density of the cloudless sulphur was through the central parts of the state and that of the
gulf fritillary, along the western sections. Peak numbers were recorded just east and just
vii


west of Gainesville, respectively. Both species avoided the Atlantic coast. Using previous data from Gainesville flight trap catches in conjunction with density estimates from this study, the total number of cloudless sulphurs and gulf iritillaries moving south each fall into central Florida was estimated at 42 and 115 million individuals, respectively. When previous estimates for the long tailed skipper and buckeye were included, the total number of butterflies migrating south through the Florida peninsula was estimated to be 175 million individuals annually. Observations of flight azimuths for three species (excluding the buckeye) were made at various locations south of Gainesville during fall 1986. Most significant mean flight directions were consistent with the 141 previously described for Gainesville and the limit of significant southward movement for that year was observed just south of Lake Ockeechobee for all three species. The phenology of the migration was studied by trapping the four major migrant species using directional flight traps at five stations from south Georgia to south central Florida during fall 1987-1988. All significant net movement was southward, extending to the Lake Alfred site, about 192 km south of Gainesville. There was no evidence from trap catches or drive counts at the Lake Placid latitude (27.45) that large numbers of migrants move into extreme south Florida to overwinter.
viii


CHAPTER 1 INTRODUCTION
Literature Review
General Migration
Migration has been recognized in recent years as an important element in the dynamics of insect populations (Solbreck, 1985). Long distance movements of butterflies (Williams, 1930), locusts (Johnson, 1969), aphids (Kennedy, 1961), moths (Stinner si al, 1983), dragonflies (Dumont and Hinnekint, 1973) and insects of many other groups have been recorded by observers across practically every continent (Williams, 1958), at high altitudes (Rainey, 1951), and many miles out to sea (Baust el al, 1981). Although some insect flights seem to function as random dispersal by forces beyond the individual's control (Johnson, 1969), closer behavioral studies reveal that active initiation of flight occurs in many species and correlates with favorable wind conditions and appropriate seasons (Baker, 1978; Taylor and Reling, 1986). In other words, the migrants will "choose" the appropriate winds, during the appropriate times of year, by which they may be carried.
It was perhaps due to his interest in butterflies that Williams (1930) stressed the concept of individual control over flight direction in his early definition of migration. This individual control over flight direction is particularly evident in butterfly migrations.
1


2
These highly visible insects travel within the boundary layer, the layer of air near the
ground where wind velocity is less than the insect's air speed, and flight direction can
be controlled by the individual. The thickness of this boundary layer is variable,
determined by wind velocity and the air speed maintained by the individual (Taylor,
1958; Pedgley, 1982). This determined choice of flight track is perhaps easiest to
observe in butterflies and demonstrates that migration can be more under the individual
control of these, and perhaps other, insects than previously suspected.
There has been considerable controversy in theoretically extricating dispersion
from migration. Baker (1978) has suggested that there is a continuum in the expression
of the migratory habit, ranging from random movement to the evolution of highly
specific destinations with correspondingly adaptive physiological changes. In a recent
summary, Dan than arayana (1986, p. 1) claimed a consensus has emerged on the
terminology used to describe insect movements. His definition of migration will be
accepted for the purposes of this study:
Non-migratory movements involve travel within the habitat associated with such activities as feeding, mating and oviposition ... In contrast, migratory movements take insects beyond the habitat for the purposes of colonizing new habitats, re-colonizing old ones, aestivation or hibernation ....
The defining criterion is travel outside what is, or had previously served, as the
sustaining "habitat" to more favorable conditions. The behavior and physiology involved
in getting to these new habitats fall somewhere on a continuum in the evolution of the
migratory habit. Some insect migrants, such as the monarch butterfly, may be
reproductively inactive during these movements and maintain long distance flights on
stored body fats, but in others, mating and feeding occur in conjunction with the


3
migratory movement. Although not necessarily so, a cyclic, or seasonal pattern is often seen in these migratory movements. The same individual may return, at least part way, to re-colonize old habitats, or, the return may consist of a step-wise expansion of new generations into the former breeding range during favorable conditions (Walker, 1985a).
Baker (1969) has speculated on the evolution of migration and the maintenance of migratory behaviors in populations. The factors that initiate migration are varied and have been found to be finely attuned by genetic programming (Lamb and McKay, 1983) to changes in habitat quality over time and space (Southwood, 1962) and other environmental cues (Dingle, 1972). The influence of genetic factors on migratory tendency in members of a population has been investigated empirically by Dingle t al (1977) and Istock (1978), and theoretically by Roff (1975). It is known that individual dispersal tendencies can be highly variable within populations (Baker, 1978) and these characters have been influenced by experimental selection to some degree (Dingle, 1968; Rankin, 1978). There are a number of possible life history strategies available to an insect and how natural selection operates on these genetic factors in the evolution of the migratory habit is still unclear.
How migrants find their destinations, that is, their navigational techniques, (Schmidt-Koenig, 1975), also remain a mystery. Monarch butterflies are highly specific in their annual migration to overwintering sites in Mexico (Calvert and Brower, 1980). Walker (1985a) suggested that cloudless sulphurs from the southeast navigate toward the Florida peninsula.


4
Migrating butterflies can maintain their compass direction and therefore must have some orientation mechanism. It is possible that a time-compensated sun compass is used (Walker, 1985a; Baker, 1978), but although Baker (1968a) has shown possible azimuth orientation to the sun's position, by Pieris rapae. this remains to be proven (Able, 1980). That fall migrants can maintain their migratory direction on overcast days, although not many fly then, weakens, but does not eliminate, this theory (Verheijen, 1978). Another possibility is that they have a magnetic compass. Evidence of geomagnetic orientation has been found in the underwing moth, Noctua pronuba L. (Baker and Mather, 1982), and is well known in honey bees (Lindauer and Martin, 1972). Magnetic particles have been detected in the monarch butterfly, although their function has not been demonstrated (MacFadden and Jones, 1985). Jungreis (1987) tested two of the Florida migrant butterflies, Phoebis sennae (L.) and Agraulis vanillae (L.), and found no evidence of magnetic particles.
Migration requires a high degree of coordination among behavior, morphology and metabolism to function as an adaptive life history strategy. Migrant insects must undergo physiological changes that can sometimes be used to identify their migratory tendency. One metabolic preparation for sustained flight that has been studied is energy storage. A primary source of energy is fat, or lipids (Beenakkers, 1965), and increased levels have been found in migrating locusts and monarch butterflies (Johnson, 1974). It is known that monarchs build up these lipid stores prior to fall migration and deplete them during long flights to Mexico (Beall, 1948; Brower, 1977). Whether they replenish, or slow, the depletion of these stores by nectar feeding enroute has still not


5
been determined (Meier and Fivizzani, 1980). Beall (1948) reported that monarchs captured in Ontario, before flight, were heavier than those caught after flight in Louisiana. Fat stores are conserved carefully by thermoregulation through the overwintering period and must remain sufficient to at least fuel the return journey to new feeding or breeding grounds (Masters, 1988).
There has been very little experimental investigation of migration in the laboratory. Most studies in this area have been done with insects in which flight activity and/or orientation can be easily measured, and usually in tethered flight (Dingle, 1985). So far, researchers have been unable to measure a butterfly's level of flight activity and preferred direction in the laboratory. Flight traps (Walker, 1985b; Walker and Lenczewski, 1989), that exploit the behavioral characteristic of migrants to go up and over obstacles, rather than around, can segregate these individuals in the field, but the problems in the laboratory remain.
Florida Butterfly Migration
In a long term study using permanent flight traps in Gainesville, Walker (1991) has identified at least eight species of butterflies that migrate southward each fall through the Florida peninsula. The four principal migrants represent four families of Lepidoptera: Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long-tailed skipper; and Precis coenia (Hubner) (Nymphalidae), the buckeye. These species comprised 85% of Walker's (1991) total trap catch and they also exhibited bidirectional


6
seasonal movement i.e. north in the spring and south in the fall. The total size of the fall migration of these species has been estimated using preliminary data from this study by Walker (1991) at about 80 million individuals. This is comparable to the estimates for monarchs at Mexican overwintering sites (Calvert el si, 1979). The most significant and noticeable movements through the Florida peninsula are exhibited by the cloudless sulphur and the gulf fritillary, which have been estimated by Walker (1991) to comprise nearly 80% of the fall migration, or 64 million individuals annually.
A number of other species also show less directed or irregular migratory movements (Walker, 1991). These include four pierids, Pieris rapae (L.), the cabbage white; Eurema daira (Godart), the dainty sulphur; Eurema lisa (Boisduval & LeConte), the little sulphur; Eurema nicippe (Cramer), the sleepy orange, as well as a nymphalid, Vanessa virginiensis (Drury), the painted lady. Two hesperiid skippers are known to migrate southward in the fall through north Florida, although there is very little known of their movements since they are too small to be retained by Walker's (1991) traps. They are Lerema accius (Smith), the clouded skipper and Panoquina ocola (Edwards), the ocola skipper. Walker (1991) also reported some individuals of the monarch butterfly, Danaus plexippus (L.), as passing through Gainesville. The majority of the monarch migration, however, is southwest toward Texas and Mexico (Brower, 1977).
The recurring spectacle of huge butterfly migrations along the Gulf coast (Urquhart and Urquhart, 1976) and through north central Florida (Walker, 1978, 1980, 1985ab, 1991) is locally well known, but is inadequately represented in the literature. The Florida Gulf coast is a renowned stopover along the major North American


7
migratory route for birds (National Geographic, 1979). It seems to function this way, in some respects, for migrant butterflies also. Large numbers of monarchs, gulf fritillaries and long tailed skippers are often seen feeding at Baccharis halimifolia along the Gulf coast, particularly in the St. Marks or Appalachicola area (pers. obs.) during the fall. Cloudless sulphurs are also present, but not in such great numbers as the other species and there are some indications (Walker, pers. comm.) that they avoid coastlines. Despite the intriguing presence of these large aggregations of migrant butterflies which are obvious to the most casual observer, the subsequent movements of these individuals, except for the monarch, remain largely unknown. These species often exhibit erratic flight near the coastline, thus making it very difficult to find a preferred pattern. There are reports of movements west, following the coastline (Urquhart and Urquhart, 1976), but I have seen migration in both directions along the coast and also south across the water, as well as north back to land. The migrants seen massed along the Gulf are spectacular, and understandably, have attracted the most attention. Away from the Gulf coast, most north Florida residents realize that the numbers of these species increase in the fall, but are generally unaware of their persistent southward flights. Once the butterflies pass into the Florida peninsula, we know virtually nothing of the migration beyond Gainesville (Walker, 1991).
Most migrants pass through north peninsular Florida during September through November and it has been presumed they join resident populations in subtropical south Florida. Although the migratory flights are so conspicuous through north Florida during fall, this does not seem to be the case in the southern parts of the state. Although the


8
migrant species seem more common in south Florida during this time (Lenczewski, 1980), there has only been one report of Agraulis vanillae and Urbanus proteus migration in Melbourne (Williams, 1958). In addition, there are no data reporting any large aggregations of these species in south Florida, which is lepidopterologically very well explored, especially during the fall and winter months (Lenczewski, 1980). Perhaps, as suggested by Walker (1985a), the densities become insignificant when the migrants settle into the large areas that are available.
The only spring migrant trapped at Gainesville in large numbers has been Precis coenia. the buckeye. Although there must be some method of northern recolonization, there is only minimal direct evidence of spring return migrations by the long tailed skipper, Urbanus proteus. Spring migrations of Phoebis sennae and Agraulis vanillae are also small and little noticed except for significant flight trap catches (Walker, 1991). The total northern movement during spring of the four species has been estimated using preliminary data from this study as eight million individuals (Walker, 1991).
As stated previously, of the insect migrants, selection for precision in destination is perhaps most evident in the monarch butterfly (Brower, 1977), where large scale movements to several precisely located overwintering sites are coordinated for millions of individuals. The monarch completes the southward journey in reproductive diapause and the same individuals return, at least part of the way, during spring to repopulate the northern range (Brower, 1962). For the migrant butterflies observed moving through north Florida, it is clearly evident that the butterflies are mating, ovipositing and stopping to feed at nectar sources enroute. This reproductive activity must result in subsequent


s
generations that will also presumably join the migration south and perhaps colonize northern areas in which they have had no experience, the following spring. As Baker (1978) has suggested, the behavior and ecology of these species may demonstrate a sequence in evolutionary development along the spectrum of the migratory habit and can provide new insights into these behaviors.
Florida provides an ideal geographical location for the study of butterfly migration. The peculiar configuration of the state results in a number of migratory species from the southeastern United States being funneled through the peninsula presumably to the southern latitudes where they join resident populations and overwinter. Although there are reports of occasional butterflies some distance at sea (Baust gt al, 1981), the difficulty of long oceanic trips (Brower, 1962) most likely precludes the majority of butterflies from flying across the Gulf of Mexico, and certainly, the Atlantic. The peninsula of Florida, then, acts as a trap with a concentration of individuals contained by extensive water masses on three sides. The butterfly migration through Florida perhaps ranks among the world's phenomenal insect movements in that it involves a large number of unrelated species from a wide area within a short period of time. Larsen (1988ab) has reported a similar mixed butterfly migration during spring and fall in southern India. The movement involves more than twenty species, numbering approximately four million spring migrants and as many as 100 million fall migrants. The width of the migratory front reported by Larsen (1988ab) was, at most, only 12 km in the spring. It may also be true that the majority of the Florida spring return migration does not pass through Gainesville.


10
At the very least, understanding why, how and where these butterflies are going presents a fascinating puzzle. As stated previously, outside of Gainesville, Florida, virtually nothing is known of the phenology or extent of the migratory movements. The goal of this study was therefore to contribute to general knowledge of the migration, primarily of Phoebis sennae and Agraulis vanillae. elsewhere in peninsular Florida.
Research Plan
The investigations in this study focused on the migratory patterns in peninsular Florida of the two principal migrant butterflies, Agraulis vanillae and Phoebis sennae. Some attention was also given Urbanus proteus and Precis coenia. The research was designed as a four part study, carried out during the falls of 1986 through 1988 and addressed the following questions:
1. Flight direction Is the mean flight direction of migrants throughout Florida the same
as in Gainesville?
2. Numbers of migrants Is the migration density across Florida along the Gainesville
latitude constant? How many migrants pass into central Florida annually?
3. Phenology of movement Does the peak migration at southern sites occur at
approximately the same time as it does in Gainesville? Where does the migration stop?
Estimates of migration density for the cloudless sulphur and gulf fritillary were made along the Gainesville latitude (29.65). From these data, a "migration profile" or cross section, through the stream of migration for each species was constructed. This


11
information was used, together with Walker's (1991) data from Gainesville, to estimate the size of the annual fall migration of these two species moving through north central Florida. The phenology of the migration was studied by trapping the four principal migrant species using directional flight traps at five stations from south Georgia to south central Florida. Counts of north and south flying Phoebis sennae and Agraulis vanillae also were made on other west-east transects along the Florida peninsula.
The individual chapters have been written as drafts of separate publications. This has unavoidably resulted in some repetition. The numerical assignment of weeks for calendar dates during which sampling was done with all methods is given in Table 1-1. The location of all sites used for pole counts and flight trap stations are shown in Fig. 1-1.
Biology of the Butterflies
Phoebis sennae
Commonly known as the cloudless sulphur, this is a fast flying butterfly of temporary, disturbed habitats. It is found in the eastern and southern United States, rarely straying to Canada, through the West Indies and south to Argentina (Lenczewski, 1980). In both the temperate and tropical zones, the subspecies of sennae exhibit strong migratory habits (Howe, 1975) and C.B. Williams (1938) nicknamed this "The Travelling Butterfly." The subspecies eubule occurs commonly in the southeastern United States, straying as far north as Canada (Klots. 1951) during the summer and fall


Fig. 1. I-Location of pole count and flight trap sampling locations used during the fall of 1986, 1987 and 1988.


Table 1-1. Calendar dates of sampling and the corresponding assigned week number for all sampling methods used during 1986-1988.
1986 1987 1988
WEEKS
1 01-07 SEP AUG 30-05 SEP AUG 28-03 SEP
2 08-14 06-12 04-10
3 15-21 13-19 11-17
4 22-28 20-26 18-24
5 29-05 OCT 27-03 OCT 25-01 OCT
6 06-12 04-10 02-08
7 13-19 11-17 09-15
8 20-26 18-24 16-22
9 27-02 25-31 23-29
10 03-09 NOV 01-07 NOV 30-05 NOV
11 10-16 08-14 06-12
12 17-23 15-21 13-19
13 24-30 22-28 20-26
14 29-05 DEC 27-02 DEC
15 06-12 03-10
months. Generally, the species occurs in the Mississippi Valley to Central Illinois and along the Atlantic coastal plain to New Jersey (Calhoun et al, 1990). It is known to reproduce as far north as west central Illinois (Sedman and Hess, 1985) and Virginia (Clark and Clark, 1951). In Florida, this species has been recorded for every month of the year throughout the state, but individuals are most common during the late summer and fall in north central Florida (Lenczewski, 1980; D. Baggett, pers. comm.). This subspecies is also resident in the states adjacent to Mexico (Howe, 1975). During some years, individuals are frequent in eastern Kansas and Missouri where they may be seen flying in a southeast direction in late summer (Howe, 1975). Similar flights have been


14
recorded in Mississippi and Alabama (Mather and Mather, 1958; Lambremont, 1968), apparently heading southeast toward the Florida peninsula. Walker's (1985a) observation of systematic changes in P. sennae's flight direction at many locations in the southeast suggests that these migrants are navigating toward peninsular Florida. The cloudless sulphur does not exhibit the coastal build up of other species mentioned and, as Walker has suggested (pers. comm.), may detect the "proper" time to change its flight direction before hitting the Gulf or Atlantic coastlines from up to 60 miles away! Gaddy and Laurie (1983) note some consistent discrepancies in flight direction along the South Carolina coastline for Phoebis sennae. In August, September and early October of 1978-1980, they noted northeastern migrations along the immediate coast and, what seemed to be random movements, inland. As previously discussed, flights adjacent to coastlines can be erratic. Migrating butterflies often seem "confused" and fly in what appear to be inappropriate directions, more often than not following the coastline in either or both directions.
Adult males are yellow above and unmarked. Females are a deeper, more orange yellow, fringed with dark brown marginal spots. The adults are particularly attracted to red flowers, many of which flower in the fall and are an important source of nectar for migrating butterflies during that time (pers. obs.). Larvae are a pale yellowish green with a yellow lateral stripe along each side. They are most often found at the young shoots of their leguminous hostplants and sometimes web leaves together for shelter (Howe, 1975). The larvae of this butterfly feed on a wide range of leguminous hostplants. There are at least 50 species of hostplants recorded and adults


15
will opportunistically oviposit on plants according to availability. The preferred larval hostplants in Florida for this species include various species of Cassia, particularly obtusifolia (L.) (pers. obs.), and as Howe (1975) reported in Louisiana, the partridge pea, Chamaecrista cinerea (L.). Two generations reportedly occur in the northern part of the range and although Howe (1975) reports that breeding is continuous in the Gulf region, including Florida, there are no records of such activity during the winter months, at least in the latter area. During 1987, Calhoun ej al (1990) reported cloudless sulphur individuals sighted in Wisconsin and New York. Numbers of individuals seen that summer were unusually large and sightings extended the northern range significantly. The species was common in Mississippi, Illinois, Kentucky, Indiana, Ohio and West Virginia. He described temporary breeding colonies of Phoebis sennae in Ohio and West Virginia, where apparently, due to the unusually warm winter of 1986-1987, individuals overwintered and were available in early spring to utilize newly emerged hostplants. The larval hostplants of this species usually senesce in late fall and are absent through the winter months. Interestingly, these breeding populations in Ohio and West Virginia were located along river banks, which Calhoun (1990) considered to be primary corridors of dispersal.
Agraulis vanillae
This butterfly, also known as the gulf friullary, is commonly found in temporary, disturbed habitats such as roadsides and fields, usually near its larval foodplant, during the late summer and fall months in Florida. The species is widespread,


16
ranging throughout tropical America, from Argentina north to the southern United States (Howe, 1975) and throughout the West Indies (Riley, 1975). The eastern subspecies, nigrior. occurs from Florida, west to Louisiana, north to North Carolina and is also found on Bermuda. Although this species has been reported as far north as New York (Howe, 1975), it is not commonly found above southern Virginia (Clark and Clark, 1951) on the eastern coast. Unlike the cloudless sulphur, this species is not usually seen during winter months (December-February) in northern Florida. However, there was an exceptionally warm winter in Gainesville during 1986-87 and adults were seen flying in January (pers. obs.).
Like the other heliconiids, the larvae of Agraulis vanillae feed only on species of Passiflora. In Georgia and north central Florida, the caterpillars are commonly found on the passion flower, Passiflora incarnata L. (pers. obs.). The larva has a black body, with three pairs of dorso-lateral, orange-brown stripes. There are six rows of branching tubercles on the body and a pair of longer tubercles curving back on the head. The adult is distinctively colored with bright orange upper wings marked with black and the hindwing underside covered with metallic silver spots and they are particularly attracted to the flowers of Spanish needle, Bidens pilosa and lantana, Verbena spp. According to Klots (1951), there are three or more broods in the south. In Gainesville, Florida, the butterflies arrive and begin to oviposit as soon as plants are available in early spring, continuing to breed until the winter months, usually late November (pers. obs.).


17
Urbanus proteus
Urbanus proteus. the long tailed skipper, is a minor pest of cultivated beans, known as the bean leaf roller in agricultural literature. This species is distributed from Argentina through the United States and West Indies. The mainland populations comprise the nominate subspecies, with all of the insular populations falling into a second subspecies (Howe, 1975). In the United States, JJ. proteus proteus is found from Connecticut, south to Florida and west to Texas, Arkansas, Arizona and California. The larvae are green and have a dark mid-dorsal line with yellow lateral lines and green stripes below these. The area between the stripes is dotted with black and yellow spots. The head is large, reddish brown, with two yellow spots between the ocelli and the mouthparts. The caterpillars come out to feed from leaf shelters that they construct on their leguminous hostplants, moving to a larger one at each instar. The major foodplants include such legumes as Bauhinia. Clitoria mariana. Desmodium. Phaseolus. Soja. Vigna. Wisteria and Pueraria lobata (Lenczewski, 1980). In the adult, the thorax, wing bases and hind wings are a conspicuous metallic green. There are four separate square spots on the central band of the forewing and a larger square spot under the end of the cell. The male has a costal fold, the antennal club is yellowish brown below and the range of wingspread is 38-50 mm (Howe, 1975). Adults are found flying in open areas, particularly fields and along roadsides in disturbed sites. Three generations are reported in Florida but only representatives of the late-summer brood arrive in Virginia where this species is considered a casual visitor (Clark and Clark, 1951). Although Howe (1975) claims it occurs throughout the year, there are no records during the summer months of


18
June and July in south Florida (Dade or Monroe Cos.). Instead, the recently introduced, nonmigratory species, JJ. dorantes. with very similar habits is common during this period (Lenczewski, 1980).
Precis coenia
Commonly known as the buckeye, this butterfly is the only Florida migrant which demonstrates relatively large scale spring northward migration (Walker, 1991). The genus Precis is largely confined to the tropical regions of the world, only two species occur in North America. There has been considerable confusion concerning the systemaucs of this species, which can be quite variable in size and color. P. coenia is found from southern Canada, west to California, Arizona and south through tropical America and Cuba. Clark and Clark (1951) reported two seasonal forms in Virginia. The spring/autumn form had an average forewing length of 27 mm in females and 24 mm in males. The summer form averaged a slightly greater wing length and darker color, characterized by an irregular reddish band on the underside of the hind wings which the spring/autumn form lacks. Clark and Clark (1951) noted that the wings in nearly all of the spring/autumn specimens they collected in Virginia were damaged. These worn wings may indicate long distance travel and this corresponds well to what has been observed of migration periods for this species. Mather (1967) found light and dark forms to be correlated with seasonal changes. He found the major seasonal shift from dark to light occurs between March and April and from light to dark between August and September. The shifts were found to coincide with a change in the mean


19
temperature from below 16C to above 16C and from above 27C to below 27C and not with rainfall, as previously suspected. Scott (1975) investigated the issues of male territoriality and migratory tendency which seem to be at theoretical odds. To the casual observer, buckeye males seem highly territorial, dashing from their resting places, to chase intruders or meet females. Scott (1975) concluded that it is a "psuedo-territorial" tendency and is not maintained for any significant length of time at a particular site.
The adults are often seen in open country, especially sitting along dirt roads near disturbed, weedy fields. Three broods occur in the southeast (in Virginia from late May to late fall). Although the butterflies of the last brood may be seen on warm days through the first half of winter, they are not found in the spring in Virginia (Clark and Clark, 1951). This species is considered a summer breeding resident in Virginia and there are no records of overwintering individuals. However, Clark and Clark (1951) report having seen ragged, overwintering individuals in March, April and early May in Washington, D.C. Howe (1975) describes the adults as hibernating and sometimes migrating. In southern Florida at least, the subspecies, P. coenia coenia. is reported for every month of the year (Lenczewski, 1980).
The larva has a dark gray body, striped or spotted with orange-yellow. There are a number of short, branching spines on the body and one pair on top of the head. The larval foodplants are Ruellia (Acanthaceae); Plantago (Plantaginaceae); Antirrhinum. Buchnera. Gerardia harperi. Linaria. Mimulus. Scrophularia lanceolata and others (Scrophulariaceae); Ludwidgia (Onagraceae); Sedum (Crassulaceae); Verbena prostrata


20
and Acuba (Verbenaceae). The Scrophulariaceae appear to be the most important larval hostplants in Florida.
Other Migrants
Pieris rapae (L.). An introduction from England into Quebec in 1860, this well-known migrant has spread rapidly throughout most of North America. Also known as the cabbage white, it can be a serious agricultural pest on cruciferous crops. Although very common in agricultural fields in north central Florida, this species is now rather rare in south Florida (Lenczewski, 1980). Walker (1991) reported northward movement in the spring in Florida, but it seems this species is also capable of overwintering as far north as Canada (Scott, 1986).
Vanessa virginienis (Drury). Although the American painted lady is a common species in north Florida, this butterfly has been reported by Lenczewski (1980) as infrequent in south Florida (Dade and Broward Cos). The range of this species extends from coast to coast in the U.S., from southern Canada to Colombia, the Canary Islands and Hawaii. It is rare in the Antilles and an occasional vagrant in Europe. This species was not previously known to be migratory, at least to the extent of V. cardui. As found for P. rapae in Gainesville, Florida, V. virginiensis showed northward movement in the spring but individuals were seldom caught in the fall (Walker, 1991). Possibly their movements are not related to temperature since Scott (1986) reports this species as also overwintering in Canada.


21
Eurema lisa (Boisduval and Le Conte). Scott (1986) reports this species as overwintering in the southeastern U.S. and extending its range north to Canada in summer. At Gainesville, Walker (1991) trapped E. lisa only in the fall when it was moving south. This species is a well-known migrant and has been reported on occasions to fly over the Caribbean, the Atlantic and the Gulf of Mexico in huge swarms (Howe, 1975; Klots, 1951; Lenczewski, 1980).
Eurema daira (Godart). This species overwinters in the southeastern Coastal Plain and spreads north a few hundred kilometers in summer (Scott, 1986). Walker (1991) found no significant movement south in the fall through Gainesville except in the fall of 1985.
Eurema nicippe (Cramer). This species sometimes moves in a seemingly inappropriate direction, northward during fall flights (Walker, 1991) and reportedly overwinters throughout the southeastern U.S. (Howe, 1975). However, it has not been reported in December, January or March in south Florida Dade or Monroe Cos., although it is very common there in the fall (Lenczewski, 1980).
Danaus plexippus CL.). The monarch is undoubtedly the best studied migrant butterfly. Large numbers of monarchs aggregate along the Florida panhandle coast each fall. Most of this eastern population flies west along the Florida coast and down through Texas to overwintering sites in Mexico (Brower, 1977). The migrating monarchs generally fly at altitudes (Gibo, 1986) where flight traps are of little use, although Walker (1991) did trap 15 in ten years and reported a significant southward bias in the fall. Those individuals that enter Florida become physiologically "trapped" when warm


22
temperatures end their reproductive diapause, and subsequently, migratory tendency (Brower, 1961; T. Van Hook, pers. comm.). There are reports of several small overwintering colonies along the west coast of Florida (Brower, 1961; and pers. com.) but the breeding status of these is unknown. About 20% of female monarchs along the Florida Gulf coast during fall migrations are mated and will oviposit on available milkweeds in Gainesville also during that time (T. Van Hook, pers. comm.). This species is also known to breed in south Florida during the winter months, but the subsequent migratory tendencies of the offspring are not known (Lenczewski, 1980).
Panoquina ocola (Edwards). This small skipper is often found concentrated in large numbers along the Gulf coast with other migrants in the fall (pers. obs.) and has been reported by Walker (1978) as migrating through Gainesville. A mass movement of these skippers in Louisiana was described by Penn (1955). Found through most of the southeast, from Virginia to Florida, west to Texas, Kentucky and Arkansas, it has also been reported as far north as New Jersey (Howe, 1975). There are no records of this species in south Florida (Dade and Monroe Cos.) in January (Lenczewski, 1980).
Lerema accius (Smith). The range of this skipper is from New England to Florida, west to Illinois, Arkansas, Texas and south to northern South America. Howe (1975) reports it as scarce northward but common in the southern states, with records from February to November. Walker (1978) reported some fall southward movement for this species and it is present during every month of the year in south Florida (Lenczewski, 1980).


CHAPTER 2 FLIGHT DIRECTION
Introduction
Of all insect migratory movements, those of butterflies are the most evident and most accessible to study. Unlike their nocturnal counterparts, the moths, butterflies are highly visible, daytime fliers which move through the boundary layer (Williams, 1930; Baker, 1978). This is the layer of air near the ground where wind velocity is less than the insect's air speed. The thickness of this boundary layer is variable, determined by wind velocity as well as the air speed maintained by the individual (Taylor, 1958; Pedgley, 1982). Most other insect migrants, such as leafhoppers (Taylor and Reling, 1986), noctuid moths, locusts and aphids (Johnson, 1969), travel at much higher altitudes. Radar studies have shown that they are generally flying with the wind (Pedgley, 1982), although they may take off from the ground only when wind direction is favorable (Williams, 1958). Flight within a few meters of the ground, however, affords the individual more control over direction and also the opportunity to cease flying by clinging to vegetation should conditions become unfavorable. This type of flight allows an observer to easily identify species as well as to observe flight direction and behavior under variable environmental conditions.
23


24
There are at least eight species of migrating butterflies that pass throughnorth central Florida (Walker, 1978, 1980, 1985ab, 1991) each fall. The four most obvious and abundant species are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long-tailed skipper; and Precis coenia (Hubner) (Nymphalidae), the buckeye. Outside of Walker's studies (1978, 1980, 1985ab, 1991; Walker and Riordan, 1981), mostly based in Gainesville, Florida, there is very little known about the flight directions of these species. The conspicuously yellow cloudless sulphur, especially, has been reported travelling in southeasterly fall flights throughout the southeastern states (Lambremont, 1968; Shannon, 1916; Williams, 1930, 1958; Clark and Clark, 1951; Howe, 1975; Urquhart and Urquhart, 1976; Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981). Although there are reports of inappropriate northern flights (Gaddy and Laurie, 1983; Muller, 1977), or variable movements, especially along coastlines (Urquhart and Urquhart, 1976), the majority of individuals appear to be navigating toward the Florida peninsula (Walker, 1985a). Walker's trapping and observational data (1978, 1980, 1985ab, 1991) clearly demonstrate that the migrants pass through Gainesville, in north central Florida, each fall and maintain an average flight path approximating 141. However, for most of the rest of the state, flight paths are unknown. My study was undertaken to investigate previously unknown migratory flight directions of the four principal migrant species at various sites throughout the Florida peninsula.


25
Materials and Methods
Flight Azimuths
During the fall migratory season (September-November) of 1985-1988, flight azimuths for the four migrant species were recorded at a number of locations throughout the Florida peninsula. Sites selected for observations were large, open areas, such as high school football fields or large lawns. Hours for observation were peak flight periods established by Walker (1985a) from 0730 to 1430 HRS LMT. Observation periods were also limited to weather conditions that are favorable to migration and were terminated if more than 50% of the sky was obscured by clouds, temperatures dropped below 21 C or wind speed exceeded 3 m/sec. For each observation, these conditions were noted as described in detail by Walker (1985a): civil time, species, appearance of sun, percentage of blue sky visible, wind speed, wind direction and air temperature. The butterfly passing closest to the observer was selected and a sighting of its flight path was then made from the point where it had passed, to where it disappeared on the horizon, noting the magnetic bearing with a Suunto KB 14 compass. Such observations continued with the next individual spotted and so forth, also as described by Walker (1985a). Magnetic bearings were adjusted for magnetic declination, resulting in true bearings and civil time was converted to local mean time. Walker (1985a) found that mean migratory directions at Gainesville remained approximately 1410 regardless of species, season, time of day or wind. For the purposes of this study, migration data for all dates at a locality were grouped for each species. These data were analyzed using BUTTAZ.BAS, a


26
computer program designed by Walker (1985a) to calculate mean direction (M.D.), length of mean vector (r) and the 95% confidence interval of M.D. (Batschelet, 1981, Zar, 1984).
Mark-Release-Recapture
A number of individuals of Phoebis sennae were also marked and released during the fall migration season of 1987. This mark-release-recapture took place at two sites in Gainesville, Florida, from September through December. Both sites had gardens with a number of flowering plants that the adults used as nectar sources. The most northern site was Kanapaha Botanical Gardens and the second site was a Gainesville residence located eight miles south of Kanapaha Gardens on Williston Rd. Between my visits, observations of marked individuals were recorded by the resident, Dr. Lincoln Brower, resulting in more complete information at this site. A total of 800 cloudless sulphurs were netted over a period of three days from 10-12 September and given a number on the upper and lower surfaces of the right wing with a permanent felt tip marker, color coded for each site. Sites were visited three times per week until 23 November 1987 and all marked individuals were noted. Observations at the residential site were continued by Dr. Brower throughout December.
Results and Discussion
The mean flight directions during fall migration 1985-1988 through various sites in Florida for Phoebis sennae are shown in Fig. 2-1. The length of the line represents


A=Arcadia n=13,r=.58,161 ,p<.01 AP= Alligator Point n=18,r=.26,68 B=Bartow n=27,r=.88,146,p<.001 BG=Belle Glade n=0 C=Cross City n=51 j=.89,171,p<.001 CB=Crescent Beach n=0 CL=Clermontn=8,r=.45,19I CW=aewiston n=4,r=.77,123 E=Everglades City n=2.r=.51,32 E P= East point n=3,r=.9,46 F=Frostproof n=3,r=.72,92 FC=Fishcreek n=4.r=.53,320 G=Gainesville 141 (Walker, 1985a) GS=Glen SL Mary n=5j=.59,126 H=Hawthome n=61,r=.81.152,pc001 HA=Hastings n=9,r=.33,175 I=Interlachen n=34,r=.94,168,p< 001 TM=Immokalee n=12,r=.62,170 ,p<.01
J=Jasper n=4,r=. 12,162 KB=Keeton Beach n=6,r=.29,330 L=Lake City n=4,r=.80,146
LE=Leesburg n=8,r=.96,133,p<.001 LA=Lake Alfred n=7,r=.74,165,p<.02 LB=LaBeUe n=13,r=.66,153.p<.005 LP= Lake Placid n=4,r=.79.71 N=Newberry n=41x=.80,146,p<.001 OOlga n=13,r=.84,178 JK.001 OK=Okeechobee n=12,r=.63.77.p<.005 P=Palatka n=17,r=.72,143 ,p<.001 PE=Perry n=18,r=.93.144,p<.001 S=Steinhatchee n=46.r-.42,86,p<001 SG=SL George n=2,r=l,54 T=Trenton n=25,r=.72,151,p<.001 TA=Tallahassee n=9j=.32,94 W=White Springs n=2,r=.88,167 YJ=YeehawJunction n=l,r=l,159
Fig. 2-1.--Summary of mean flight directions at various sites throughout Florida for Phoebis sennae during fall migrations, 1985-1988 (direction of arrow=mean flight direction, value given in key for each site; length of line=mean vector r, bold letters=significant r value). ^


28
the mean vector (r) with sites labeled in bold letters indicating statistically significant vector values for those sites with n >5 using Rayleigh's z test (Zar, 1984). Gainesville data is, as previously reported by Walker (1985a), 141. The observations made in this study along the latitude of Gainesville also show a similar pattern of movement. The inland sites south of Gainesville correspond to the expected flight direction if migrants were continuing along the same track observed at Gainesville. Although not statistically significant, due to small sample sizes at many sites, the movement along coastlines seemed confused and quite variable. Along the Gulf coast, butterflies usually followed the coastline even in inappropriate directions. Previous reports of Phoebis sennae flight directions are southeastward inland and coastwise when near the coast, as reported by a number of different observers over many years (Shannon, 1916; Williams, 1930,1958; Clark and Clark, 1951; Lambremont, 1968; Howe, 1975; Urquhart and Urquhart, 1976; Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981; Walker, 1985a).
Phoebis sennae is particularly sparse on the east coast considering the southeast bearings of inland migrants. It has been observed by Walker (1985a) that even 90 km from the coast, butterflies in south Georgia fly more southerly than further inland, as though they can detect the coast. This trend is also seen here and no individuals of any migrant species were observed at Crescent Beach, the easternmost site of the transect. The reason for an absence of migrants along the east coast portion of the transect is not clear. Migrating butterflies do tend to follow coastlines possibly using a different orientation mechanism used from inland flight. If north or south coastline flight is not distinguished, then by avoiding the east coast altogether, northward flight during the fall,


29
possibly resulting in freezing, could be prevented. Dave Baggett has reported (pers.comm.) large numbers of long-tailed skippers flying south along the St. John's River near Jacksonville during some years. This geographical feature may serve as a path when found and, if it can be detected at a distance, may create a funnel drawing migrants away from the east coast areas. Calhoun ej a] (1990) reported some colonization by cloudless sulphurs along rivers that appeared to serve as "dispersal corridors." It has also been observed (see Chapter 3) that this "shadow" or absence of migration along the east coast seemed to widen inland with distance travelled south. Other possible reasons for the absence of these migrating butterflies along the Florida east coast may be the lack of an appropriate habitat or some other adverse environmental conditions found there.
Throughout the state, significant migratory flight directions for Phoebis sennae are usually south or southeast. It is noteworthy that in this species, there appears to be a shift in mean direction immediately north of Lake Okeechobee, as though the migrants "detect" the lake. At the town of Okeechobee, the mean migratory track shifted to the east and most individuals (79%) avoided flying toward the large expanse of water. No cloudless sulphurs were observed along the southern boundary of the lake at Belle Glade, just a short time after seeing many individuals flying through Okeechobee. Again, this seemed to be consistent with observations that the butterflies at the northern border of the lake were going around, rather than flying directly over water. A distinct southward migratory flight was detected as far as Immokalee, just southwest of Lake Okeechobee. This is also the northern limit for much of south Florida's more tropical fauna and flora


30
and the southern boundary for many northern species. Historically, this is about the point at which damage from a hard freeze is unlikely. It is important to consider the number of visits to a site and the pattern of individual flight directions shown in Fig. 2-2, rather than the mean vector alone. For example, although the mean vector at Steinhatchee (Fig. 2-1) is approximately due east, from Fig. 2-2, it is evident that the individuals had no consensus in migratory direction. The same is true at Alligator Point during the only visit there. The eastward vector in Fig. 2-1 really is a result of south and north flying individuals as shown in Fig. 2-2.
The mean flight direction patterns for the gulf fritillary throughout the state are similar to those described for the cloudless sulphur and are shown in Fig. 2-3. Significantly directed movement for the gulf fritillary continued south to Immokalee, but fewer individuals were observed. Whereas Phoebis sennae is not common along the gulf coast, Agraulis vanillae masses along the panhandle coast in large numbers (pers. obs.). Also, unlike Phoebis sennae. the gulf fritillary showed no evidence of avoiding Lake Okeechobee. Three individuals were observed flying southeast from Okeechobee and others were apparent at the opposite shore in Belle Glade, heading in the same direction. However, this species too avoided Crescent Beach, the most eastern site along the Atlantic Coast.
Whereas some cloudless sulphurs do overwinter in north Florida and buckeyes have also been reported to do so (Walker, 1978; Scott, 1986), this is not usual for the gulf fritillary. The more highly directed flight of the gulf fritillary may be necessary to ensure survival before freezing temperatures occur. The individual flight directions


ARCADIA ALLIGATOR POINT BARTOW
i
PHP GLADE CROSS CITY
O
CLERMONT CRESCENT BEACH CLEWISTON EVERGLADES CTTY EASTPOINT
0 o

G

FROSTPROOF F1SHCREEK GLEN ST. MARY HAWTHORNE HASTINGS


INTERLACHEN IMMQ
JASPER KEETON BEACH LAKE CITY



LEESBURG LAKE ALFRED LABRI .IF, LAKE PLACID NEWBERRY
Q
0
OLGA
OKEECHOBEE PALATKA
PERRY STFJNHATCHEE

ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
0
Or
Q
31
Fig. 2-2.--Azimuths of Phoebis sennae flight directions taken at various locations in Florida during fall migration 1985-1988. The diameter of the circle represents the number of visits (5 mm = l visit); the length and width of the line represents the number of individuals (2 mm = l, in bold 2 mm =4); the orientation of the line represents the flight direction.


A=Arcadia n=6,r=.98,170 ,p<.001 AP=Alligator Point n=13,r=.17,164 B=Bartown=10,r=.84,174 ,p<.001 BG=BeUe Gladen=0 C=Cross Cityn=21 j=.89,156 .rx.001 CB=Crescent Beach n=o CL=aermont n=8,r=.6,142 CW=Clewiston n=5,r=.92,146 E=Everglades Cityn=6j=.62,109 EP=Eastpoint n=2,r=.99,277 F=Frostproof n=1 ,r= 1,163 FC=Fishcreek n=0 G=Gainesville (Walker, 1985a) GS=Glen St. Mary n=0 H=Hawthome n=27,r=.84,160 ,p<.001 HA=Hastings n=4,r=.84,132 I=Interlachen n=14,r=.96,179 JX-001 IM=Immokalee n=16j=.72,164 ,p<.001
J=Jasper n=4,r=.92 KB=Keeton Beach n=4,r=.7,290
L=Lake Cityn=4,r=.66,154 LE=Leesburg n=5j=.54,163 LA=Lake Alfred n=8,r=.96,155 .rx.001 LB=LaBelle n=2j=.97,164 LP=Lake Placid n=3,r=.99,175 N=Newberryn=16,r=.97,148 jx.OOl 0=01ga n=2j=. 17,253 OK=Okeechobec n=4,r=.96,143 P=Palatka n=12j=.84,172 ,p<001 PE=Perry n=3,r=.94,151 S=Steinhatchee n=13j=.80,137,p<001 SG=St. George n=3,r=.56,141 T=Trenton n=6,r=.34,147 TA=TaJJahassee n=9j=.32,146 W=White Springs n=5,r=.58,174 YJ=Yeehaw Junction n=lj=l,159
Fig. 2.3.--Summary of mean flight directions at various sites throughout Florida of Agraulis vanillae during fall migrations, 1985-1988 (direction of arrow=mean flight direction, value given in key for each site; length of line=mean vector r, bold letters=significant r value).
to


33
observed throughout Florida during fall 1985-1988 are shown in Fig. 2-4. Like P. sennae. Agraulis vanillae showed a great deal of variation in flight direction along the Florida panhandle coast at sites such as Alligator Point. However, at Steinhatchee, which is further south along the west coast, there was no evidence of the confused flight that P.. sennae had shown and movement was significantly biased southeast.
The summary of mean flight directions of Urbanus proteus throughout Florida are shown in Fig. 2-5. Its migratory flight patterns are similar to those of the gulf fntillary. Also common along the Gulf coast, this species exhibited a great deal of variation in flight at coastal sites. This is evident in the individual flight directions from various sites in Florida shown in Fig. 2-6. For example, at Keeton Beach, movement was in every direction. Oddly, a similar variation in flight was noted at Yeehaw Junction, an inland location just north of Lake Okeechobee. However, a few miles directly south, at Okeechobee, the flight of individuals was very consistently due south. It appeared they were heading out across the water, but just south of the lake, at Clewiston and Belle Glade, there were no sightings of incoming individuals. It is possible that the long tailed skipper makes shorter, more localized flights than the other two species discussed at inland sites. This may also account for a great deal of variability in flight directions.
Very few individuals of Precis coenia were recorded throughout the state during fall migrations. This is consistent with Walker's (1991) findings from flight traps in Gainesville that most of the movement is northward in the spring. In this study, of the


34
ARCADIA ALLIGATOR POINT BARTOW BELLE GLADE CROSS CITY

CLERMONT CRESCENT BEACH CLEWISTON EVERGLADES CITY EASTPO


O
FROSTPROOF FISHCREEK GLEN ST. MARY HAWTHORNE HASTINGS
o o
INTERLACHEN JASPER KEETON BEACH LAKE CITY
LEESBURG LAKE ALFRED LAB ELLE LAKE PLACID NEWBERRY

OLGA OKEECHOBEE PALATKA
O O
PERRY STEINHAT


ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
9

Fig. 2-4.Azimuths of Agraulis vanillae flight directions taken at various locations in Florida during fall migration 1985-1988. The circle diameter represents the number of visits (5 mm = l); the length and width of the line represents the number of individuals observed (2 mm = l); the orientation of the line represents the flight direction.


A=Arcadia n=9,r=.81,168 AP=Alligator Point n=0 B=Bartow n=8,r=.66,156 BG=Belle Glade n=0 C=CrossCity n=12,r=.82,147 CB=Crescent Beach n=0 CL=aermont n=3,r=.82,147 CW=Clewiston n=0 E=Everglades City n=6j=.48.121 EP=Eastpoint n=0 F=Frostproof n=2,r=.94,143 FC=Fishcreek n=2,r=.64,318 G=Gainesville (Walker, 1985a) GS=Glen St. Mary n=0 H=Hawthome n=6,r=.89,157 HA=Hastings n=3j=.31,338 l=Interlachen n=3,r=.67,135 IM=Immokalee n=llj=.68,177
J=Jasper n=l,r= 1,152 KB=Keeton Beach n=22,r=.28,168 L=Lake City n=0
LE=Leesburg n=12j=.89,145 LA=Lake Alfred n=2j=l,347 LB=LaBeUe =7,r=.53,215 LP=Lake Placid n=lj= 1,277 N=Newberry n=l,r= 1,280 0=01ga n=5,r-87,168 OK=Okeechobee n=6j=.97,186 P=Palatlca n=o PE=Perry n=2j=l,162 S=Steinhatchee n=23j=.91,120 SG=St George n=0 T=Trenton n=3,r=.98,152 TA=Tallahassee n=0 W=White Springs n=0 YJ=Yeehaw Junction n=13,r=.55,144
Fig. 2-5.-Summary of mean flight directions at various sites throughout Florida of Urbanus proteus during fall migrations of 1985-1988. (The direction of the arrow=mean flight direction, value given in key for each site; the length of line=mean vector r; bold letters=significant r value).


ARCADIA ALLIGATOR POINT BARTOW BELLE GLADE CROSS CITY
off
o
CLERMONT CRESCENT BEACH CLEWISTON EVERGLADES CITY EASTPOINT
O O
o
O o
FROSTPROOF FISHCREEK GLEN ST. MARY HAWTHORNE HASTINGS
Q 0 O
IMMOKA I.I EE INTERLACHEN JASPER KEETON BEACH LAKE CITY
o
O
o
LEESBURG LAKE ALFRED LAB ELLE LAKE PLACID NEWBERRY
o o
OLGA OKEECHOBEE PALATKA PERRY STEINHATCHEE
Q
Q
ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
O G) O O
Fig. 2-6.~Azimuths of Urbanus proteus flight directions during fall migrations, 1985-1988. The circle diameter represents the number of visits (5 mm=l); the length and width of the line represents the number of individuals (2 mm=l); the orientation of the Itine represents the flight direction.


37
two sites where there were more than two individuals recorded, the movement was southward. The only location where buckeyes were sighted south of Gainesville was in Bartow, central Florida and those few individuals were headed north.
Mark-Release-Recapture
In two instances, cloudless sulphurs were recaptured at some distance from the mark-release sites in Gainesville. These are the first instances of an individual Phoebis sennae being tracked over a relatively long distance and time. The first individual had been marked at Kanapaha Gardens and was seen again feeding at flowers at the Williston Rd. residence, eight miles south of there, ten days later. The second individual had been marked at the Williston Rd. residence and was recaptured in Bronson, Florida, 51 km southwest of Gainesville, 14 days later. This cloudless sulphur had evidently taken refuge in some potted plants during a cold evening and had been brought into the house. In both instances, the butterflies travelled a much shorter distance than would be expected from the elapsed time. It has been calculated (Arbogast, 1966; Balciunas and Knopf, 1977) that migrating cloudless sulphurs or gulf fritillaries go 15-20 km/hr under ideal conditions. Thus, such a journey could easily be completed in one day.


CHAPTER 3 NUMBERS OF MIGRANTS
Introduction
Each fall, a migratory mass of butterflies comprised of at least eight species from four families, passes through north central Florida. The most conspicuous of these migrants is a pierid, Phoebis sennae (L.), commonly known as the cloudless sulphur. This butterfly has been reported throughout the southeast during fall migratory flights which are usually toward the Florida peninsula (Walker, 1985a). The three other principal migrant species are Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long-tailed skipper; and Precis coenia (Hiibner) (Nymphalidae), the buckeye. Large numbers of these species have been noted along the Florida Gulf coast (Urquhart and Urquhart, 1976) as well as passing through north Florida regularly (Walker 1985ab, 1978, 1980, 1991).
It has been difficult to estimate the size of this migratory flight. Walker (1991) has operated traps for more than ten years in Gainesville, Florida, but the magnitude of migration had not been monitored at any other location. In order to accurately estimate the total numbers of butterflies moving south into central Florida each fall, the density of migratory flights must be monitored at other locations along the latitude of Gainesville (29.65). Walker's (1991) long term trapping results from Gainesville could then be
38


39
projected to estimate the absolute numbers moving south each fall. The goal of this study was to construct a "migration profile" of density estimates for the two principle migrants, Phoebis sennae and Agraulis vanillae. along a transect across north central Florida at the latitude of Gainesville. This cross section through the migratory stream will then be used to estimate the total fall migration of Phoebis sennae and Agraulis vanillae into peninsular Florida.
Materials and Methods
Latitudinal Pole Counts
In the fall of 1986, a counting system previously used by Walker (1985a) was used to estimate the numbers of migrants moving through various sites along the Gainesville, Florida, latitude (29.65). Three 3-meter lengths of PVC pipe were used with the second pipe set in the ground 15 meters from the first, and the third, 30 meters beyond the second. All three pipes were along a straight line at 51-231, perpendicular to the average flight track of 141 established in Gainesville by Walker (1985a). All observations were made in large, clear areas (i.e. athletic fields or lawns) to avoid any vegetative or structural influence on the paths taken by migrants. The air temperature (C), percentage of clear sky and sun exposure (b= bright, h=hazy, o=obscured by clouds, p=disk obscured, but position easily determined 5) were noted. Wind speed was measured with a hand held pith ball anemometer and direction was noted. The numbers of individuals of the four principal migrant species were counted crossing between two poles for three five-minute periods with a one minute break between each


40
observation period. Cloudless sulphurs were easily observed at 45 meters and therefore individuals flying between the first and third poles, were counted. Long-tailed skippers, gulf fritillaries and buckeyes were less visible and were counted if within 15 meters (between the first and second poles). Counts were made at least once every three weeks from 1 September to 2 November 1986 at most of the ten sites (see Table 1.1 for assigned weeks and Fig. 1-1 for location of sites). This period has been shown by Walker (1991) to include >95% of the total migration through Gainesville.
Observations were adjusted for time of day differences by using a "time-of-day" factor, T, from Walker's (1985a) observations of each species' flight variation throughout the day. The average for each half hour was calculated from three days of observation (3 October 1982 and 4, 11 October 1983) by Walker of percent individuals flying throughout the day. This average percent of individuals flying was then multiplied by the T factor which would adjust up to 100% for that time period. A 15 minute adjustment to T was made by averaging two adjacent half hourly observations. For example, if 6.4% of the day's total was the average seen flying at 1000 HRS (EDT) (counts made between 0945-1014) then T would be 15.6. An average of 5.9% flying at 1030 HRS (counts between 1015-1044) results in a T of 16.9. To adjust for 15 minute periods, the average percent flying for both half hourly observations was averaged with a result of 6.15, the resulting T factor for observations at 1015 HRS would then be 16.3. Observations were grouped according to their closest time period in 15 minute intervals and multiplied by the appropriate T factor. The west-east transect sites were approximately 23 kms apart and extended the width of the state from Steinhatchee to


41
Crescent Beach, a distance of about 200 km. All results of pole counts were summarized by averaging the net numbers of individuals/m/minute from samples made at each site throughout the season.
Migration Profiles
During the fall migrations of 1986, another technique was devised to estimate the net migration across west/east transects. Counts of north and south flying butterflies were made while driving along selected roads running west to east. This method covered entire transects rather than the IS or 45 m samples made during the pole counts. Transects with low vegetation and therefore good visibility were selected. The major transect was at the latitude of Gainesville, through the same sites as the pole counts. Ideally, the entire transect was travelled roundtrip once per week, with half the counts completed on a particular day. Only cloudless sulphurs and gulf fritillaries were included in the drive counts because they were easily identifiable at a distance. To be counted, a butterfly had to at least have reached the centerline of the road before the car passed. A speed of approximately 97 kph was maintained and counts were recorded for every 16 km of transect from which the net number of indiv/min was calculated. Standard time was noted at the beginning and end of each segment of transect, as was weather. Weather conditions recorded were the percentage of blue sky visible, the appearance of the sun (b=bright, h=hazy, o=obscured, p=disk obscured, but position easily determined 5), air temperature (C), wind direction and speed (m/sec with a hand held anemometer).


42
During this study, the analysis of the data presented several problems incurred by the distance and time required for sampling and daily and seasonal changes in the density of migration. Walker and Riordan (1981) have shown that by sampling only on days that meet certain weather criteria, day-to-day fluctuations in numbers can be minimized. Driving periods were limited to weather conditions favorable to migration and were terminated if more than 50% of the sky was obscured by clouds, the temperature dropped below 21 C or wind speed exceeded 3 m/sec. There are however, hourly differences in migration throughout the day with peak flight times usually from 1000 to 1400 HRS EDT (Walker, 1985). Since it was not possible to sample simultaneously along the transect sections, it was necessary to accommodate the daily changes in peak flight periods. Adjustments for time of day differences were made to the actual net numbers observed during drive counts (D) by multiplying with the "time of day" factor, T (see explanation for pole counts above).
As a result of time limitations, unpredictable weather changes and other occurrences during sampling periods, not all segments along a transect were always visited on a particular sampling trip. To equalize for the differences in sample number from each transect segment, the time adjusted counts (DnT) were then transformed relative to a time corrected "base" segment (DjT) which was always sampled during each trip. For the western portion of the transect (Gainesville-Steinhatchee), this base was the Gainesville-Newberry segment. For the eastern portion of the transect (Gainesville-Crescent Beach), the base was the Gainesville-Hawthorne segment. The net number of


43
individuals observed along each transect segment was converted to a proportion of the net number observed along the base segment = DnT/Dj.
Another important influence on the amount of migratory movement observed is a seasonal factor (Walker, 1991). Since the drive counts were made over a period of several months, numbers had to be adjusted for season. In addition, partial profiles from different drive counts had to be combined to estimate the entire trans-state profile. Walker (1991) has collected phenological information on the migration pattern in Gainesville, Florida for more than ten years. His two permanent traps sample continuously throughout the season and have been shown to reflect migration patterns reliably. To eliminate the effects of season and to unify the east and west transects, Walker's trap catches were used to quantify the "strength" of migration for each day that drive counts were made. The net numbers of gulf fritillaries and cloudless sulphurs captured in Walker's two traps were summed for the period during which drive counts were made that year and a mean net number was calculated. The daily net number caught in his traps was then divided by the mean net number to produce a factor that would adjust to the mean net number. The reciprocal of this factor was called the migration index, I, for that day. If the migration was particularly strong, the migration index was correspondingly low. The adjusted data from drive counts was then multiplied by the migration index appropriate to the day that sampling occurred. In this way, the high numbers observed during the peak migratory season were reduced and the more sparse, late or early migration was augmented, in effect eliminating the seasonal


44
difference. The results of these calculations yielded the estimated migration profiles for 1986, 1987 and 1988.
Dn = NET # INDIVIDUALS/MINUTE FROM DRIVE COUNTS T = TIME OF DAY ADJUSTMENT
D, = NET # INDIVIDUAI^/MINUTE AT BASE SITE (TIME CORRECTED) I = MIGRATION INDEX
ESTIMATED MIGRATION PROFILE = I (DnT/D,).
The summary migration profiles were then derived from the median segment-by-segment values of the yearly estimated profiles for each species. These cross sections of migration density through north central Florida were used to estimate the annual number of P. sennae and A. vanillae butterflies flying into central Florida by applying the migration profiles to Walker's (1991) Gainesville permanent trap data.
Results and Discussion
Latitudinal Pole Counts Cloudless sulphur
Pole counts for the fall 1986 cloudless sulphur migration are summarized in Fig. 3-1. The two coastal sites showed no net movement. At Steinhatchee, this was due to equal numbers of individuals flying north and south, and, at Crescent Beach it was due to a lack of migrants. All net movement southward was statistically significant, except for Palatka, also near the east coast, where not many individuals were seen. Peak


45
Fig. 3-1.--Summary of pole counts showing Phoebis sennae migration (sum of time adjusted net number of individuals/meter/minute/number of visits) in north central Florida along the latitude of Gainesville (29.65). Solid bars represent statistically significant (chi-square, p < .05) southward movement and open bars were sites with no statistical bias (chi-square, p <.05) in net movement. Counts were made during 1 September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number of visits to each site are shown above the bars.


46
numbers of individuals were counted through the central sites of Newberry, Gainesville and Hawthorne, averaging about .45 indiv/m/min. The highest count was through Gainesville at .57 indiv/m/min and numbers declined toward both the Gulf and Atlantic coasts.
Gulf fritillary
Pole counts taken of the gulf fritillary migration during fall 1986 are shown in Fig. 3-2. All net movement was southward and the peak migration was concentrated slightly more to the west, passing through Trenton, Newberry and Gainesville. Peak numbers were recorded in Trenton at .58 indiv/m/min followed by Newberry at .48 indiv/m/min. Whereas the cloudless sulphur had little or no migration on either coast, the west coast migration for the fritillary was relatively strong (.3 indiv/m/min). Few individuals were counted at Hastings, which was also the only site lacking statistical significance to the southward net movement. At Crescent Beach however, directly along the coast, movement was recorded at .13 indiv/m/min. The greater presence along the coastlines of this species corresponds to the gulf fritillary buildup along the panhandle coast.
Long tailed skipper
The summary for fall 1986 pole counts of long tailed skipper net movement across the west-east transect at the latitude of Gainesville is shown in Fig. 3-3. The greatest number of individuals move south through the central sites of Gainesville-Hawthorne (.21 indiv/m/min), but migration of Urbanus proteus is also substantial along both the Gulf and Atlantic coastlines (. 14 and 13 indiv/m/min, respectively). Centrally


47
n i i i i l i-1-1-1-1--tj-1
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-2.--Summary of pole counts showing Agraulis vanillae migration (sum of time adjusted net number of individuals/meter/minute/number of visits) in north central Florida along the latitude of Gainesville (29.65). Solid bars represent statistically significant (chi-square, p < .05) southward movement and open bars were sites with no statistical bias (chi-square, p <.05) in net movement. Counts were made during 1 September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB = Newberry, GNV=Gainesville, HAW=Hawthorne, rTL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number of visits to each site are shown above the bars.


48
STN CRC TRN NWB GNV HAW ITL PLK HST CRB
0 16 32 48 64 80 96 112 1 28 144 160 176 192
KILOMETERS
Fig. 3-3.--Summary of pole counts showing Urbanus proteus migration (sum of time adjusted net number of individuals/meter/minute/number of visits) in north central Florida along the latitude of Gainesville (29.65). Solid bars represent statistically significant (chi-square, p < .05) southward movement and open bars were sites with no statistical bias (chi-square, p <.05) in net movement. Counts were made during 1 September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number of visits to each site are shown above the bars.


49
located Gainesville, as well as the two coastal areas were the only locations with statistically significant movement (chi-square, p < .05). When the sampling period was subdivided to early, middle and late migration, a single, statistically southward biased peak shifted eastward from Steinhatchee to Gainesville to Crescent Beach, respectively. The total fall migration resolved into three peaks, which may represent three pathways into central Florida. Large numbers of long tailed skippers were counted near the west coast, but the highest counts were in the central region, passing through Gainesville. A third peak in migration began at the St. John's River in Palatka and extended to the east coast. Dave Baggett (pers. comm.) has reported large numbers of JJ. proteus migrating along the St. John's River during the fall of some years. This pattern may reflect some geographical obstacles, visual attractions, habitat preferences or source areas, such as bean fields.
Migration Profiles Cloudless sulphur
Profile estimates of Phoebis sennae migration in 1986 across the west-east transect at the Gainesville latitude are shown in Fig. 3-4 a,b,c,d,e,f,g,h. The peak net southward movement along the western half of the transect (=STN-GNV portion) occurred through the TRN-NWB or NWB-GNV segments on 14 October, during week 7 of sampling. Throughout the first five weeks, the movement along the western transect remained fairly constant and relatively low, <. 2 indiv/min. During week 6, this rate doubled and in week 7, counts in the western segments (Steinhatchee-Newberry) increased dramatically. Along the peak transect segment (CRC-TRN, Fig. 3-4g, west)


Fig. 3-4 a,b,c,d,e,f,g,h.Profile estimates of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 13 September to 28 October 1986.Counts were taken on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


ADJ NEI # INDIVMIN S ADJ NET # INDIVMIN
A DJ NET # INDIVMIN
MO Ol to M
Ol
ti
?
Q
g
A

s
< z
i
-1
ADJ NET t INDIVMIN ijjg MO u
11 m 1 t'
3K
a *
SI g -
E I
H
w
i-i
i
i
ADJ NET INDIVMIN 3
M M u
m
ADJ NET INDI V/MIN 3 ?5 ? oa <7> *> w u
M
Si
m o


52
at this time, the net number of individuals flying south/minute was nearly 40, 8 times greater than seen previously. Unfortunately, that was the only drive count through CRB-TRN during the entire fall. During week 9, the GNV-HAW and ITL-PLK segments along the eastern transect also increased, but only to about a third of what was seen on the Gulf coast. On 16 October, the migration through the two Gulf coast segments (STN-CRC, CRC-TRN) was southward and twenty times higher than that of any other segment, at any other time, throughout the season. It is not known if large numbers of migrants regularly pass through these segments. The one additional sample through STN-CRC (Fig. 3-4c, west) and four samples at TRN-NWB (Fig.3-4 a,c,d,e, west) seem to indicate that at least during the early season it is not common. It is possible that such large scale movements result from synchronous emergence of a locally produced generation. Large fields observed in the Trenton area had extensive stands of Cassia obtusifolia. a primary hostplant of cloudless sulphurs in this area.
Similar profile estimates of the cloudless sulphur migration during fall 1987 are shown in Fig. 3-5 a,b. All net movement was southward and greatest along the western transect through GNV-HAW during week 6. This segment had more than 300 indiv/min passing south, but the net catch for Gainesville that day was only one cloudless sulphur, flying south! This resulted in a high migration index that perhaps inflated what was an average day in the HAW-ITL segment. Along the eastern transect, the numbers were highest through the TRN-NWB segment during week 7, with 4.5 indiv/min. This transect was sampled on 17 October, almost one year to the day when extremely large


53
WEST
WEEK 4
WEEK 6
EAST
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-5 a,b.-Profile estimates of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 26 September to 18 October 1987. Counts were taken on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW = Hawthorne, ITL = Interlachen, PLK = Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


54
numbers had been noted there in 1986. The peak numbers along this western transect in 1987 were only one tenth of what they were the previous year.
The profile estimates showing Phoebis sennae migration along this transect in fall of 1988 are given in Fig. 3-6 a,b,c,d. Each transect consists of a roundtrip made during the same day and close in time (trip 1 and 2). Each trip consistently showed the peak net movement occurred southward through the central, NWB-INT, or western STN-TRN segment and was lowest in the east.
All profile estimates made of Phoebis sennae net movement during the fall of
1986, 1987 and 1988 are summarized as yearly migration profiles in Fig. 3-7 a,b,c. Most transect segments showed a statistically significant (chi-square test, p < .05) net southward movement throughout the sampling period. The exceptions were PLK-HST, where in 1986 and 1988, net movement was not significantly biased and HST-CRB, in
1987, where no individuals were sighted. During 1986 (Fig. 3-7a), the peak southward movement was through the western segment of CRC-TRN, 37 indiv/min, followed by STN-CRC with 14 indiv/min. As was pointed out in the individual 1986 profile estimates (Fig. 3-4g), this was a result of one sample taken on 16 October. If this trip were not included, the central transect segment from Trenton to Palatka would contain most of the southward movement. Peak movement south for the fall of 1987 (Fig. 3-7b) was through the HAW-INT segment at nearly 200 indiv/min, the highest rate recorded through three years of sampling. The secondary peak was through the adjacent segment of GNV-HAW with about 11 indiv/min. Both of these segments showed an average


a)
WEST
WEEK 3 :18 SEP
b)
EAST
WEEK 4 :25 SEP
1
2
TRIP 1
MM'
1NVHAW m PLK MST CRapNV HAW (TL PLK HST era
t-r
TRIP 2
I
c)
.5
z
s o
>
Q
Z 1 Z 2
3 <
0 16 32 48 64 80 06 0 16 32 48 64 80 96 KILOMETERS
WEST
WEEK 5 : 29 SEP
4 112120136152168184200112128144160176192
TRIP 1
T
iTN CRC TIM NWI OMV (TN CMC TUN NWI ONV
TRIP 2
T
-1-t------1-r-
KILOMETERS
EAST
WEEK 6 :12 OCT
3 > Q Z
*
tu
z
a <
0 16 32 48 64 80 96 0 16 32 48 64 80 96 KILOMETERS
TRIP 1 TRIP 2
III"! 1 V
112 120138 152 188184 201 M12128144160176192
KILOMETERS
Fig. 3-6 a,b,c,d.-Profile estimates of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 18 September to 12 October 1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.
m


Fig. 3-7 a,b,c.~Migration profiles summarizing estimates of Phoebis sennae migration across a west-east transect in north central Florida along the latitude of Gainesville (29.65). Yearly migration profiles were derived from the median estimate (net number of individuals/minute, adjusted for time of day, sample size and season). Counts were made from 13 September-28 October 1986, 26 September-18 October 1987 and 18 September-12 October 1988 on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically significant (chi-square, p < .05) southward movement and open bars were sites with no statistical bias (chi-square, p < .05) in net movement. Total number of visits made to each site are shown above the bars.


57
a)
_1986_
|STN CRC TRN NWB QNV HAW ITL PIX HST CRB
iii
b)
-210 160 O110 60
1"
5 1.5
0 16 32 48 64 0 S6 112 12S 144 160 17 102
KILOMETERS 1987
pTN CRC TRN NWB QNV HAW
.5
D
u
ITL PIK HST CRB
0 16 32
I I-r
60 M 112 128 144 160 176 162
c)
KILOMETERS 1988
0 16 32 46 64 80 86 112 12S 144 160 176 182
KILOMETERS


58
migration rate, but traps in Gainesville caught very little. As a result, the high migration index for that day has increased numbers for these segments. If these days were not included, the GNV-HAW segment would still have the peak southward movement at 2.3 indiv/min, followed by the TRN-NWB segment at 1.4 indiv/min. There were no cloudless sulphurs observed during the one trip made through the HST-CRB segment along the Atlantic coast in 1987. The migration profile for fall of 1988 (Fig. 3-7c) showed a peak through HAW-INT of 2.7 indiv/min, nearly twice that of the next highest peak through NWB-GNV (1.4 indiv/min). Movement along either coast was low, with both east and west segments below .5 indiv/min. For all three years, the Atlantic coastal segment of transect, Palatka to Crescent Beach, consistently had the least number of migrants, averaging about .18 indiv/min. Gulf fritillary
The profile estimates of Agraulis vanillae migration during fall 1986 are shown in Fig. 3-8 a,b,c,d,e,f,g,h. On the western transect, movement mostly occurred through the TRN-NWB segment and peaked during week 6 (Fig. 3-8e) with 20 indiv/min. The Gulf coast segment of STN-CRC, was not sampled during this time, but the following week (Fig. 3-8g), it showed peak movement south at 8 indiv/min. Numbers were low along the eastern transect during all the weeks sampled, except for large numbers (23 indiv/min) during week 9 (trip 2, Fig. 3-8h) through INT-PLK. All peak movement through the season occurred within this segment, except for week 9 (trip 1, Fig. 3-8h) where GNV-HAW had nearly 15 indiv/min.


Fig. 3-8 a,b,c,d,e,f,g,h.-Profile estimates of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 13 September to 28 October 1986. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


ADJ NET f INDIV/MIN S
oi
fi
zi 5s
i-m N

g
fi
S
aal
a
al
S
3

i s

ADJ NET INDIV/MIN
S ui t* fJ
fi %
-
S8
ADJ NET f INDIV/MIN
e m m m
i-1-1-1-
I
M
i
_ADJ NET f INDIV/MIN e m m
23 ~ ? 8 *
fif
al
S
25
* i s
-1-1---
M i
M 13 SE


:
,
1 w a W
EP
*
>
i
S!
(
AOJ NET # INDI VMIN
tt S M MO
ADJ NET INDIV/MIN
is =18
m -
S
i S
S
ad
a*.
to
09
fi
Ol
= 5
a
s 10 S M
qs-
m ->
(
M
3 fi
Ol
3
1 5
fi
* s
SI
W M 0>

fi
Ol
ADJ NET f INDIV/MIN m m m
—i-1-1-1-
_A0J NET INDIV/MIN
m m M
5
(
ss
(


61
Profile estimates for the gulf fritillary migration during fall of 1987 are shown in Fig. 3-9 a,b. All net movement was southward and the migratory peak was recorded during week 7 (Fig. 3-9b, west) along the TRN-NWB segment at nearly 15 indiv/min. This increase in numbers was seen one year and a day after the extraordinarily large numbers of Phoebis sennae were noted along the Gulf coast in 1986 (Fig. 3-4g). Numbers of individuals seen along the eastern transect remained low and the only segments with significant movement southward were the central areas of GNV-HAW and HAW-INT.
Profile estimates of gulf fritillary migration during fall of 1988 are shown in Fig. 3-10 a,b,c,d. Each weekly sample is a round trip during the same day and most of the patterns were similar between trips 1 and 2, except during week 3 of sampling (Fig. 3-10a). Although trip 1 of week 3 showed fairly uniform movement across the western transect (all < 2 indiv/min), trip 2 peaked through TRN-NWB with > 10 indiv/min. Other than the TRN-NWB segment on this day, STN-CRC had the next largest number of individuals, during week 3 and week 5 (Fig. 3-10c, trips 1 and 2). The Atlantic coast had relatively low numbers of migrants (< 2 indiv/min) during week 4 (Fig. 3-10b), but in week 6 showed an increase with peaks at HAW-INT (trip 1) and INT-PLK (trip 2).
Yearly migration profiles of Agraulis vanillae movement during the fall of 1986, 1987 and 1988 are shown in Fig. 3-11 a,b,c. The profile for 1986 (Fig. 3-1 la) peaked through the INT-PLK segment (7 indiv/min) with the next peak at TRN-NWB (4.8 indiv/min). During 1987 (Fig. 3-1 lb), the TRN-NWB transect segment also contained the migration peak at 7.5 indiv/min. This rate was nearly three times the number


62
a)
WEST
WEEK 4
WEEK 6
EAST
>
z
LU
Q <
0
1 -
2 -3 4
26 SEP
~4
5TN CRC TRN
NWB
4 OCT
XXX HAW ITL PLK HST CRB
-1-1-1-1-1-1-1-1-1-1-1-r-
16 32 46 64 80 96 112 128 144 180 176 192
AN
ts
b) WEST
KILOMETERS
WEEK 7
17 OCT
WEEK 8
EAST
18 OCT
STN CRC TRN NWB GNV HAW ITL PLK HST CRB
\-1-1-1-1-1-1-1-1-1-1-1-1
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-9 a,b.--Profile estimates of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 26 September to 18 October 1987. Counts were made on nine transect segments defined by 10 sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


2 0
2 4 6
e
10
c)
WEST
WEEK 3:18 SEP
TRIP 1 I I
TRIP 2
STN CRC TRN NWB tlNV 5TN CRC TRN IWB ONV
0 16 32 48 64 80 96 0 16 32 48 64 80 96 KILOMETERS
WEST
WEEK 5 : 29 SEP
o
2 4 6 8
10
TRIP 1 TRIP 2
J 1 1 B STN crc TRN NWBONV STN crc TRN NWB ONV
N
f
S
b)
i
2 0
2 4 6 8 10
EAST
WEEK 4 : 25 SEP
TRIP 1 TRIP 2
I INV HAW OL PLK HST CRB wr~mr- 1NV HAW (TL PLK HST CRE
112120136152168184200112128144160176192
KILOMETERS
EAST
WEEK 6:12 OCT
TRIP 1 TRIP 2
II1 1 INV HAW ITL PLK HST CRB >l||-MV HAW ITL PLK HST CRB
KILOMETERS
j )M *M 111 im
KILOMETERS
Fig. 3-10 a,b,c,d. Profile estimates of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day, sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 18 September to 12 October 1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


Fig. 3-11 a,b,c.--Yearly migration profiles of Agraulis vanillae movement across a west-east transect at the latitude of Gainesville (29.65). The yearly migration profiles are from the median estimates (net number of individuals/minute, adjusted for time of day, sample size and season) during 13 September-28 October 1986,26 September-18 October 1987 and 18 September-12 October 1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically significant (chi-square, p < .05)'southward movement and open bars were sites with no statistical bias (chi-square, p < .05) in net movement. Number of visits to each site is above bars.


65


66
recorded at the next highest peak, through CRC-TRN (2.7 indiv/min). All southward movement was statistically significant (chi-square, p < .05) except at PLK-HST and HST-CRB during 1987, where numbers were small. During 1988, the migration peak occurred along the Gulf coast through STN-CRC, but at lower numbers than previous years (approx. 4 indiv/min). During all three years, the lowest numbers of individuals moved southward along the east coast.
Numbers of Migrants
A migration profile summarizing the results of 1986, 1987 and 1988 profiles of Phoebis sennae migration is shown in Fig. 3-12. All net movement was southward throughout the fall sampling period (13 Sept-28 Oct). The transect segment with the greatest numbers of cloudless sulphurs moving south was HAW-INT at 2.7 indiv/min, more than double any other segment. The migration rate was fairly even across the other transect segments, averaging about 1 indiv/min, except for PLK-HST (.34 indiv/min) and HST-CRB (.18 indiv/min), where numbers declined approaching the Gulf coast.
The migration profile summarizing Agraulis vanillae movement for 1986, 1987 and 1988 is given in Fig. 3-13. Peak migratory movement for the gulf fritillary was through the TRN-NWB segment at 4.8 indiv/min. The other peak was along the west coast segment, STN-CRC at 3.9 indiv/min. In the case of the gulf fritillary, all three segments with peak movement were along the west coast. Migration east of Newberry was low in general, but the transect segments with the least movement were PLK-HST and HST-CRB, at .81 and.84 indiv/min respectively.


67
Fig. 3-12.A migration profile summarizing Phoebis sennae migration across a west-east transect at the latitude of Gainesville, Florida (29.65). The profile is derived from a median estimate (net number of individuals/minute, adjusted for time of day, sample size and season) at each site during 13 September-28 October 1986, 26 September-18 October 1987 and 18 September-12 October 1988. Estimates were made along nine segments of a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.


68
QOn I I I I I I i .....
< 0 16 32 48 64 80 96 112 128 144 160 176 192 (KM)
1986-1988
Fig. 3-13.-A migration profile summarizing Agraulis vanillae migration across a west-east transect at the latitude of Gainesville, Florida (29.65). The profile is derived from a median estimate (net number of individuals/minute, adjusted for time of day, sample size and season) at each site during 13 September-28 October 1986, 26 September-18 October 1987 and 18 September-12 October 1988. Estimates were made along nine segments of a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry, GNV = Gainesville, HAW=Hawthorne, rTL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.


69
Walker (1991) estimated the size of the seasonally recurring migration through peninsular Florida by extrapolating his flight trap catches from Gainesville using a preliminary estimate of the migration profile from this study. His estimate of the net movement of migrating butterflies each fall was an average of 22.2 million individual cloudless sulphurs and 41.3 million gulf fritillaries. The combined figure of 63.5 million butterflies approaches the estimated number of monarchs reported to overwinter at Mexican sites (Brower, 1985; Calvert el al, 1979). The average rate of net movement for Phoebis sennae and Agraulis vanillae (1986-1988) from the migration profile of each species was used to recalculate Walker's (1991) data (Table 3.1). Walker's (1991)
Table 3.1 Estimates of the Phoebis sennae and Agraulis vanillae fall migration into the Florida peninsula (Lenczewski, 1992 and Walker, 1991). Average absolute net numbers of individuals flying across a north central Florida transect at the latitude of Gainesville (29,65),_
E
Mean profile height (Lenczewski, 1992) 1.03 2.10
NWB-GNV profile height 1.03 1.43
Profile correction factor 1.00 1.47
Kilometers across state (Lenczewski, 1992) 190.00 190.00
Migrants trapped/6m (Walker, 1991) 800.00 868.00
Est. migrants into cen. Fla. (millions) 25.30 40.40
Est. trapping efficiency (Walker, 1985b) 0.60 0.35
Adj. migrants trapped/6m 1333.00 2480.00
Est. migrants into C. Fla. (millions) 42.20 115.00


70
calculations were based on an assumption that the average rate of migration across the transect was no less than 60% of the rate at the longitude of his trapping site in Gainesville. In actuality, for Phoebis sennae. his site fell within the NWB-GNV segment of the transect which had a rate of 1.03 indiv/min, identical to the transect average. Walker also used a conservative transect length of 170 km, whereas the actual transect length used in this study is 190 km. Walker's (1991) Gainesville trap yields of 800 individuals/6 m for Phoebis sennae were applied to these data resulting in an estimated influx into central Florida of 25.3 million individuals. With further adjustment for Walker's (1985b) 60% trap efficiency, the size of the cloudless sulphur fall migration into central Florida was estimated as 42.2 million individuals, nearly double the figure given by Walker (1991).
These calculations were repeated using the migration profile information for Agraulis vanillae. also shown in Table 3.1. The average migration along the latitude of Gainesville was 2.10 indiv/min, as compared to the NWB-GNV segment, which had a rate of 1.43 indiv/min, resulting in a profile correction factor of 1.47. After applying Walker's trap catch of 868 individuals/6m and an efficiency of 35%, the final results were 115 million gulf fritillaries, more than double Walker's (1991) estimate of 41.3 million. The total number of these two species of butterflies migrating south through north central Florida in the fall was estimated at 157 million individuals annually. This number does not include the buckeyes or long tailed skippers, which were estimated by Walker as numbering 3.7 and 14.6 million, respectively. Since profiles were not established for these species from drive counts, Walker's estimates cannot be evaluated.


71
Using his numbers for these other two species, the total migration of these four principle migrants is given as 175 million individuals annually.
Fall 1987 drive counts of Phoebis sennae net migration across west-east transects south of Gainesville, Florida, during fall 1987 are given in Fig. 3-14 a,b,c,d,e,f. The net numbers observed in these drive counts were adjusted only for time of day. Visits to the same transects were approximately 4 weeks apart and week 9 was the only time that the full north-south range of transects was travelled. Statistically significant southward movement is evident south to the Crystal River-Leesburg transect (3) during week 4 (Fig. 3-14a) and continued through week 9 (Fig. 3-14c). It is not known whether individuals are continually moving southward throughout this time or if this is a secondary wave of migrants from subsequent generations. From the pattern observed at Gainesville, there was probably a continual, and gradually diminishing, net southward movement throughout this period. By week 15 (Fig. 3-14f), this northern area no longer showed any significant migration. The only time statistically significant migration was noted south of Crystal River-Leesburg was during week 7 at the Zolfo Springs-Sebring transect. As is seen with results of trap catches (discussed in Chapter 4), there was no evidence of large scale migratory movement through south central Florida. Unless the migratory track deviates from the mean flight direction observed at Gainesville, the migration appears to diminish south of Leesburg. Small numbers of individuals may be moving farther south, perhaps with a second generation augmenting populations around the Zolfo Springs-Sebring area. Drive counts made during fall 1987 depicting the net migration of Agraulis vanillae net migration across various west-east transects south of


Fig. 3-14 a,b,c,d,e.f.--Drive counts of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day) across various west-east transects south of Gainesville, Florida during fall 1987. Transects with a southward or northward net migration of statistical significance (chi-square test, p < .05), are shown with half filled circles weighted in the appropriate direction. Net movement, either north or south, that is not statistically significant (p < .05), is depicted with open circles. Lines with no circles represent transects where there were no individuals seen. Number of individuals and percentage flying south are given in parentheses after each transect.


1. GULF HAMMOCK-KENDRICK
2. INGUS-BELLEVIEW
3. CRYSTAL RIVEfl-LEESBURG
4. W1LDWOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
6. BROOKSV1LLE-HILL N DALE
7. BAYONET POINT-SAN ANTONK) 8 WESLEY CHAPEL-ZEPHRHILLS
8. LAKELAND-HAINES CITY
10. BARTOW-LAKE WALES
11. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIDGE-LAKE PLACID
q-Q south bias
- north bias
o-o no bias
GNV LAT (28.65) NONE SEEN
7 MM =20 KM


74
Gainesville, Florida are shown in Fig. 3-15 a,b,c,d,e,f. A pattern similar to P. sennae is seen for this species which also demonstrated a statistically significant southward movement from Inglis-Belleview through Crystal River-Leesburg during week 4. Subsequently, the farthest significant southward movement seen occurred during week 7 through Bartow-Lake Wales. Again, since the more northern transects were not sampled at this time, it is not known if southward migration continued there during this period, but presumably it did. Significant migration was still seen through the north central transects during week 9 and, from the patterns seen along the Gainesville transect (Fig. 3-11), it is likely that this continued. No individuals were observed on any trips south of Gainesville (Fig. 3-15 d,e,f) after this time and Bee Ridge-Lake Placid also showed no significant migration throughout the season for this species.
The area known as the Green Swamp is located between transects 5 and 7. It is a large area maintained as wilderness, with some grazing areas, for cattle. The wet areas along the Withlahochee River support a variety of flowering plants, especially in the disturbed habitats where cattle are allowed to graze. Large numbers of migrants have been sighted there during some years, particularly in late November (H. Nigg, pers. com.). This largely unexplored area may contain many resources and provide an ideal environment for adult winter survival. There is evidence from Gainesville (Chapter 4), that some cloudless sulphur individuals will remain in an area with nectar sources for several weeks.


Fig. 3-15 a,b,c,d,e.f.-Drive counts of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day) across various west-east transects south of Gainesville, Florida during fall 1987. Transects with a southward or northward net migration of statistical significance (chi-squared test, p < .05), are shown with half filled circles weighted in the appropriate direction. Net movement, either north or south, that is not statistically significant (p < .05), is depicted with open circles. Lines with no circles represent transects where there were no individuals seen. Number of individuals and percentage flying south are given in parentheses after each transect.


WEEK 13
27 NOV
13 t
TRANSECTS
1. GULF HAMMOCX-K END RICK
2. INGUS-BELLEVIEW
3. CRYSTAL RIVER-LEESBURG
4. VflLDWOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
. BROOKSVILLE-WLL N DALE 7. BAYONET POINT-SAN ANTOMO
I. WESLEY CHAPEL-ZEPHRHJLLS LAKELAND-HAINES CITY
10. BARTOW-LAKE WALES
II. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIDGE-LAKE PLACtD
WEEK 15
9-10 DEC
o
Q-Q SOUTH BIAS
- NORTH BIAS
O-O "O BIAS
GNVLAT (29.85) NONE SEEN
7 MM-20 KM


CHAPTER 4 PHENOLOGY OF MOVEMENT
Introduction
Many migratory insects, such as aphids and locusts, are greatly dependent on wind and weather conditions to direct their flight once they are airborne (Pedgeley, 1982; Taylor, 1958). These insects, as well as nocturnal moths, fly long distances at high altitudes requiring the use of sophisticated, expensive technology, such as radar, for study (Rainey, 1951; Schaefer, 1976) or extensive mark-release-recapture programs (Li t aj, 1964; Showers s al, 1989). Butterflies are particularly suitable as subjects for the study of insect migration because they are daytime fliers and move within a variable boundary layer, usually within 4-6 meters of the ground (Edwards and Richman, 1977). Wind speeds within this layer are low enough to allow individual control of flight direction (Johnson, 1969). As the butterflies move near ground level, an observer can identify species as well as note behavior and preferred flight tracks. Migrating butterflies typically continue their flight on a linear track and will fly up and over obstacles rather than around (Williams, 1930). This behavioral characteristic can be exploited to selectively capture migrants for research purposes and to quantify movement by using flight traps. Butterflies also have large wing surfaces which make it relatively easy to conspicuously mark individuals for recapture (Walker and Wineriter, 1981).
77


78
Despite all these advantages, and, probably because they are not agriculturally significant pests, the often spectacular movements of these insects have been largely ignored. A few workers (Arbogast, 1965, 1966; Balciunas and Knopf, 1977) have made observations over short periods of time and space, but it has been difficult to simultaneously monitor migrations over distance. Through a massive tagging campaign, Urquhart and Urquhart (1976) succeeded in determining the destination of Danaus plexippus (L.), the monarch butterfly, in the Mexican mountain ranges. The migrations of the great southern white, Ascia monuste L., along the Florida east coast were detailed by Nielsen and Nielsen (1952; Nielsen, 1961) over a period of at least 20 years. In Germany, Roer (1959, 1961a, 1961b, 1962, 1968, 1969, 1970) has described the movements of Aglais urticae L., Inachis io., Nymphalis antiopa and other European butterflies, while Baker (1968ab, 1969) has studied Pieris rapae in England. For more than ten years, Walker (1978, 1980, 1985ab, 1991) has maintained directional flight traps in Gainesville, Florida that continuously monitor the annual migrations of eight species of butterflies through this area. This type of data collection eliminates observer bias, time of day or seasonal influences and permits correlation with environmental factors (Walker and Riordan, 1981). A continuous, long term documentation of migration in this manner is unique and has provided reliable information about the species composition, phenology and size of the migration, at least through Gainesville.
There are three species that comprise most of the migration through north central Florida in the fall (Walker, 1991): Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; and Urbanus proteus


79
(L.) (Hesperiidae), the long-tailed skipper. Another important migrant in Florida is Precis coenia Hubner (Nymphalidae), the buckeye, but most of its migration occurs northward in the spring with the southward fall movement being relatively small. This pattern is the reverse of what is seen for the other species listed. Together, these four species make up more than 85% of Walker's (1991) total trap catch. Walker (1991) captured at least 100 individuals of five other species in his ten years of trapping. Pieris rapae (L.) (Pieridae) and Vanessa virginiensis (Drury) (Nymphalidae) showed significant northward movement in the spring, but were seldom caught in the fall. Eurema lisa (Boisduval & LeConte) (Pieridae) had significant southward movement through Gainesville in the fall, but was never caught in the spring. Another pierid, Eurema daira (Godart) moved south in the fall of one year, but showed no statistical bias in either direction over all years combined. Finally, the pierid, Eurema nicippe (Cramer) was found in some years to move northward in the fall, a seemingly inappropriate direction. The monarch passes primarily along the gulf coast to Texas in its migrations south (Brower gt al, 1985) although, a few individuals do move through north central Florida (Walker, 1991). Since this species flies at a higher altitude than others (Gibo, 1986), they are seldom captured in flight traps.
Although the movements of these species have been well documented in Gainesville by Walker, migratory activities in the rest of the state are virtually unknown. In particular, the timing, or phenology, of the migration south of Gainesville is not known, nor are the characteristics of the migratory front, as far as density and species composition. Arbogast (1966) and Balciunas and Knopf (1977) determined the flight


80
speed of migrating gulf fritillaries and long tailed skippers to be approximately 12-22 kph under ideal conditions. Weather permitting, peak migration times are usually 0900-1600 HRS LMT (Walker, 1985a) daily, suggesting that individuals may travel as much as 100 km/day. If the same individuals that pass through Gainesville during the peak of migration move directly to lower latitudes, a similar pattern of peaks would be expected at sites to the south. However, if peaks lag to the south by about 3 weeks, the approximate generation time for gulf fritillaries (Arbogast, 1965, 1966), movement south could be in a stepwise, generational manner.
Materials and Methods
Longitudinal Pole Counts
During 1986, pole counts (see description of methods in Chapter 3) were initially made at Gainesville and five sites south, Clermont, Bartow, Arcadia, LaBelle and Immokalee (see Chapter 1, Fig 1-1 for location of all sample sites). The primary purpose of these counts was to get some information on when migration began and how far it extended south of Gainesville before planning more extensive data collection with traps. Counts were made from 13 September to 2 November 1986, at approximately two week intervals (see Chapter 1, Table 1-1 for dates of all sampling). This period was figured by Walker (1991) to include 95% of the total seasonal migration through Gainesville. The six sites along the north-south transect were spaced approximately 62 km apart, a total distance of 372 km. All net numbers were adjusted as explained previously (Chapter 3) for time of day differences (T) using Walker's (1985a)


81
observations of daily flight periodicity. These adjusted net numbers were averaged to yield a seasonal summary for each site along the north-south transect.
Portable Flight Traps
For the purposes of this study, a portable flight trap was designed (Walker and Lenczewski, 1989) to quantify the phenology of butterfly migration along the Florida peninsula. Walker (1985a) has determined that the migratory track through Gainesville, Florida remained relatively constant at 141 annually. A migratory route was projected north and south (321 and 141) from Gainesville and pairs of these flight traps were erected on it at five sites (an average of 105 km apart) from southern Georgia to south central Florida: Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida. Migration was monitored with these traps through the fall of 1987 and 1988. The traps were placed in pairs along an ENE-WSW line (67.5-247.5), one trap facing north and one facing south. These traps, erected approximately perpendicular to the axis of the Florida peninsula and to the mean direction of migration, were left in place from 13 September through 12 December 1987. Butterflies captured by the south-facing trap were flying 321 90; those captured by the north-facing trap were heading 157.5 90. All traps were situated in large, open fields to eliminate the influence of vegetation or other obstacles on the preferred flight tracks of migrants. As the migrant butterflies tried to fly over a main wall of polyester screening, they passed through a narrow slot into a duct that led to two holding cages made of 1/4 inch hardware cloth. Captured butterflies were removed from the holding cages semiweekly throughout the trapping


82
period. Walker and Lenczewski (1989) estimated that semi-weekly service recovered at least 90% of all individuals entering the traps' collecting cages. After testing other construction materials in the spring of 1988, the traps were modified, primarily to increase durability. The backs of eight polyester traps were replaced with monofilament shrimp netting. Two of the polyester netting traps were left intact. An additional four traps were built of the same design but with all monofilament shrimp netting. All these traps were found to be similar in efficiency (Walker and Lenczewski, 1989), but the monofilament shrimp netting was more durable and easier to work with. In 1988, traps were in place from 28 August through 26 November.
The sites were 157, 100, 70 and 92 km apart, respectively, a total of 419 km from northernmost to southernmost site. The types of traps used each year, at each site, are as follows:
1. Valdosta State College, Valdosta, Georgia (30.85 lat, 83.29 long). 1987 two all polyester traps, 1988 four polyester traps with shrimp net backing.
2. Green Acres Farm, University of Florida, Gainesville, Florida (29.65 lat, 82.44 long). 1987 two all polyester traps. 1988 no traps at this location, data used from Walker (1991).
3. IFAS/AREC, University of Florida, Leesburg, Florida (28.90 lat, 81.54 long). 1987 two all polyester traps, 1988 four polyester traps with shrimp net backing.
4. IFAS/CREC, University of Florida, Lake Alfred, Florida (28.00 lat,


83
81.44 long). 1987 two all polyester traps, 1988 four all shrimp net traps.
5. Archbold Biological Station, Lake Placid, Florida (27.45 lat, 81.23 long). 1987 and 1988 two all polyester traps. During the fall of 1987, two all polyester traps were in operation at all five sites, one facing north and one facing south. In fall of 1988, four polyester traps with shrimp net backing were used in Valdosta and Leesburg. Four all shrimp net traps were placed at Lake Alfred, and the Lake Placid station had two all polyester traps similar to those used in 1987. During 1988, all stations had two traps facing north and two south, in an alternating pattern, except for Lake Placid, where one trap was positioned north and one south. There were no traps placed at the Gainesville site in 1988 since similar data were available from Walker's (1985b) permanent flight traps. The dates for each week sampled are as given in Chapter 1, Table 1-1.
Results and Discyssipn
Cloudless Sulphur
Longitudinal pole counts
Pole counts of Phoebis sennae net movement along a north-south transect from Gainesville to Immokalee during 1986 are shown in Fig. 4-1 a,b,c,d,e,f. All statistically significant (p < .05) net movement at the six sites was southward throughout the period sampled. All three weeks sampled at Gainesville (a) showed significant southward movement with the peak occurring during week 2. Of all the sites sampled, Gainesville


84
a)
c)
.2 0 J A M
i *
e)
GAINESVILLE
b)
1 24710
BARTOW

x. ft
12t4*7t10
WEEKS
LABELLE
g)
X X_XXX
1 214I47I4 10
WEEKS
CLERMONT
WEEKS IMMOKALEE
XXX
12S4I4710
WEEKS
Fig. 4-1 a,b,c,d,e,f.-Pole counts of Phoebis sennae movement (net number of individuals/meter/minute, adjusted for time of day) along a north-south transect through Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11 September through 7 November 1986. Solid bars represent a statistically significant net movement (chi-square test, p > .05), shaded bars represent net movement which was not significantly biased (chi-square test, p < .05). x=no counts made.


8;
had the greatest numbers moving south (1 indiv/m/min) during peak migration, well over three times that of the next largest movement at Clermont (.25 indiv/m/min). At Clermont (b), which is located approximately 136 kms south of Gainesville, the cloudless sulphur did not achieve a statistically significant southward movement until five weeks later, during week 7. The rest of the sites had no significantly directed movement, except for LaBelle, where cloudless sulphurs also migrated southward (.2 indiv/m/min) during week 7.
The summary of Phoebis sennae pole counts along the transect from Gainesville south is shown in Fig. 4-2. Gainesville had the highest counts with an average of about .6 indiv/m/min passing southward. Clermont and LaBelle, the only other sites with significant southward movement, had a rate of about 1 indiv/m/min. Flight trap catches
Weekly percentages of total net trap catch of Phoebis sennae during the fall of 1987 at five sites along a north-south transect through Gainesville are shown in Fig. 4-3 a,b,c,d,e. All statistically significant net movement was southward for this species, falling between weeks 4 and 12. The early weeks were not sampled completely at any of the sites, and the week in which half of the total net catch was achieved is probably not a reliable estimate of mid-migration point. The Gainesville data (b) were corrected for this by using information taken from Walker's (1991) permanent traps near the same location, shown in Fig. 4-4c. In the case of the cloudless sulphur, the percent total net catches for weeks 5-13 were reduced by 33% (Walker's trap catch during the first 4 weeks). The appropriate adjustment was applied in a similar manner for the other three


86
Fig. 4-2.Summary of pole counts of Phoebis sennae (average net number of individuals/meter/minute, adjusted for time of day) through Gainesville, Clermont, Arcadia, LaBelle, and Immokalee, Florida during 11 September through 8 November 1986. Numbers below the bars represent total number of counts at that site.


87
VALDOSTA
b)
GAINESVILLE
GAINESVILLE PTP#&5 WALKER(1991)
LEESBURG
5
r
m\ 90
V
n-80
10
)
1 > I 4 I I 7 10 tl U It 14 11
WEEKS LAKE ALFRED
1 2 7 10 11 12 It 14 II
WEEKS
LAKE PLACID
1 2 3 4 6 7 10 1112 13 14 16
WEEKS
1 2 3 4 7 10 11 12 13 14 16
WEEKS
Fig. 4-3 a,b,c,d,e,f. Weekly percentage of seasonal total net trap catch of Phoebis sennae at Valdosta, Georgia, Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida in flight traps during 13 September through 12 December 1987. Solid bars represent statistically significant southward biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p >.05) southward movement. Clear bars show nonsignificant (p > .05) northward movement. The total net catch (southbound minus northbound) for the period is shown in the upper right corner. The midpoint of migration is designated by an asterisk, x=trap not operating. Total percent migration in b) was reduced by 33% on the basis of data in c).


88
species. The estimate for mid-migration was either week 5 or 6 at all sites. The close coordination of mid-migration seems to indicate that the same wave of individuals passed through all five areas. The most southern site, Lake Placid, had no net movement, and even though several individuals were seen and two were trapped, there was no indication of migration. In fact, the numbers of cloudless sulphurs were surprisingly low at all other sites, as compared to Gainesville. The reason for low numbers at Valdosta at least could be that since the first three weeks were not sampled, most of the migrants may have passed southward earlier.
The percentage of total net catch of Phoebis sennae captured at the five sites in 1988 are shown in Fig. 4-4 a,b,c,d,e,f,g,h,i. Gainesville data is taken from Walker (1991). All statistically significant movement was southward biased. At Valdosta (Fig. 4-4a), for the cloudless sulphur, catches in traps 1 and 2 were not well coordinated. Trap 2 went out of commission after week 6 and the percentage of total catch was adjusted in relation to the first six weeks of trap 1 (i.e. decreased by 40%). During 1988, traps were in place by week 1, at which time there was no evidence of migration. The mid-migration point for trap 1 was during week 6, and the only other significant southward movement was seen during week 7. Southward migration occurred during the second week for trap 2, but sample size was small and the trap may already have been damaged. Walker's (1991) data from the two permanent traps located in Gainesville (Fig. 4-4 c,d) both show mid-migration occurred during week 7. The two traps at Leesburg and the average of the two traps at Lake Alfred all suggest a midpoint during week 8 at those sites. The only significant southward movement seen from catches at


Fig. 4-4 a,b,c,d,e,f,g,h,i.--Weekly percentage of seasonal total of net trap catch of Phoebis sennae at Valdosta, Georgia, Leesburg, Lake Alfred and Lake Placid, Florida during 28 August-26 November 1988. Data for Gainesville, Florida are from Walker (1991). Solid bars represent statistically significant southward biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p > .05) southward movement. Clear bars depict nonsignificant (p > .05) northward movement. The total net catch (southbound minus northbound) for the period is shown in the upper right corner. The midpoint of migration is designated by an asterisk, x=trap not operating. Total percent migration in b) was reduced by 40% on the basis of data in a).


90
a)
60
a40
I"
P 10
VALDOSTA TRAP 1
n. 2
11
c)
I I I I I T I I 11 11
WEEKS
GAINESVILLE PTP3 (WALKER. 18911 60
g)
WEEKS LAKE ALFRED TRAP 1
b)
60 40 30
*
VALDOSTA TRAP 2
N-l
X X X X X X X
d)
1 1 I I t I ) t mi 11 II WEEKS
GAINESVILLE PTP6 [WALKER. 1891) 60
XXX
12346678S10111213 WEEKS LAKE ALFRED TRAP 2
1 1 I 4 I 7 I I M 11 11 11
WEEKS
LAKE PLACID
100
i)
3
I
80 80 40
20 0
N-1
1 > I 4 I 7 I I 10 11 1] 11 WEEKS


91
Lake Alfred trap 2 was during week 11. At these four sites, from Valdosta to Lake Alfred, the mid-point of migration for traps 1 and 2 fell between weeks 6-9 and it is found almost uniformly (except for Lake Alfred, trap 2) at a one week difference between sites. Lake Placid had no statistically significant bias in movement at any time during the trapping period, although five individuals were trapped and others were seen.
The percentage of total net trap catches of cloudless sulphurs are summarized for both the 1987 and 1988 fall migration seasons in Fig. 4-5 a,b,c,d,e. The 1988 trap 1 and 2 data were tested for homogeneity and combined before calculating the mean percent total net catch for both years. The first four weeks of Gainesville 1987 and all of 1988 data used in this summary are from Walker's (1991) permanent trap catches as shown in Figs. 4-3c and 4-4 c,d. The migration midpoint occurred between weeks 5 through 7 at all of the four main sites. Lake Placid had no evidence of migration with a total net of only one individual heading south in two years. The net numbers captured at Valdosta (Fig. 4-5a) were surprisingly low, totaling only 70 individuals in two years, and similar to catches at Lake Alfred (Fig. 4-5d). During 1987 alone, Gainesville traps (Fig. 4-3b) captured nearly three times that number. The total net number at Leesburg for the two years was 275. This was unexpected since it was assumed that most individuals begin travelling south through Valdosta and then move through Gainesville. During 1987, the earlier part of the migration may have been missed, but in 1988 the pattern was similar with comparable numbers at Gainesville-Leesburg and Valdosta-Lake Alfred. Subsequent breeding could augment the migratory stream between Leesburg and Gainesville, accounting for the larger numbers caught at these sites. The sites south of


12346678 8 10 1112131416 WEEKS
2 3 4 6 6 7 8 8 10 1112131416 WEEKS
LAKE ALFRED
1 2 3 4 6 8 7 8 8 101112131416
WEEKS
8-77
XXX
2 3 4 6 8 7 8 8 10 111213 1416
WEEKS
e)
LAKE PLACID
123456789101112131415 WEEKS
Fig. 4-5 a,b,c,d,e.--Summary of percent total net trap catches of Phoebis sennae at five sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida) during 30 August-12 December 1987 and 28 August-26 November 1988. Solid bars represent net southward movement and clear bars show net northward movement. Total net catch (southbound minus northbound) is shown in the upper right corner and x=unsampled weeks. Data for the first 4 weeks of 1987 and all of 1988 Gainesville are from Walker (1991).


Full Text
87
a) b)
i a a 4 a 7 a to n u m C) WEEKS
GAINESVILLE PTP#&5 WALKER! 1991)
i a a 4 a a 7 a a io ii ia ia 14 it
GAINESVILLE
z
o
S40
hr *
N-202
V
V- 10
*
X X X x X X
d)
WEEKS
LEESBURG
s
r
N-ao
til 90
i"
K
*
0

*ix, ill,
1 a a 4 a a 7 a a io ii ia ia 14 ia
WEEKS
1 2 3 4 6 7 I 10 11 12 13 14 16
WEEKS
f)
WEEKS
LAKE PLACID
2 3 4 7 3 10 11 12 13 14 16
WEEKS
Fig. 4-3 a,b,c,d,e,f.--Weekly percentage of seasonal total net trap catch of Phoebis
sennae at Valdosta, Georgia, Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida in flight traps during 13 September through 12 December 1987. Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05),
shaded bars show nonsignificant (p > .05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right corner. The midpoint of
migration is designated by an asterisk, x=trap not operating. Total percent migration
in b) was reduced by 33% on the basis of data in c).


83
81.44 long). 1987 two all polyester traps, 1988 four all shrimp net
traps.
5. Archbold Biological Station, Lake Placid, Florida (27.45 lat, 81.23
long). 1987 and 1988 two all polyester traps.
During the fall of 1987, two all polyester traps were in operation at all five
sites, one facing north and one facing south. In fall of 1988, four polyester traps with
shrimp net backing were used in Valdosta and Leesburg. Four all shrimp net traps were
placed at Lake Alfred, and the Lake Placid station had two all polyester traps similar to
those used in 1987. During 1988, all stations had two traps facing north and two south,
in an alternating pattern, except for Lake Placid, where one trap was positioned north
and one south. There were no traps placed at the Gainesville site in 1988 since similar
data were available from Walkers (1985b) permanent flight traps. The dates for each
week sampled are as given in Chapter 1, Table 1-1.
Results and Discussion
Cloudless Sulphur
Longitudinal pole counts
Pole counts of Phoebis sennae net movement along a north-south transect from
Gainesville to Immokalee during 1986 are shown in Fig. 4-1 a,b,c,d,e,f. All statistically
significant (p < .05) net movement at the six sites was southward throughout the period
sampled. All three weeks sampled at Gainesville (a) showed significant southward
movement with the peak occurring during week 2. Of all the sites sampled, Gainesville


)
WEST
WEEK 2
EAST
|o
i*
4
h c
III D
z
3 20
<30
40
13 SEP
14 SEP

X X
BTNCftC TUN NWB Of

X X
VtTNCRC TUN NWB ONV
M m 1M 144 iao 171 i
KILOMETERS
C)
WEST
WEEK 4
EAST
5 2
5 4
K 6
Z 8
3io
*12
28 SEP
26 SEP
1 "
X
STN CRC TRN NWB

x X
QNV HAW m. PLKH8TCRB
0 16 32 48 64 BO 96 112128144160176192
KILOMETERS
b)
I
S 2
z
5 4
6
! 10
12
EAST
WEEK 3
EAST
20 SEP
rwpi
20 SEP trip]
Tn -
OMVHAWITL PLK MST CUB G
IV HAW ITL PLK HST CRB
96 112126144160176192 96 112126144160176192
KILOMETERS
WEST
WEEK 5 : OCT 2
EAST
KILOMETERS
EAST
WEEK 7
EAST
lo
2
4
U¡ 6
z
3 8
< 10
12
140CT
TWPI
14 OCT
TRIP 2
-
X X
ONV HAW ITL PLK MST CRB ON
" 1
X X
r HAW ITL PLKHST CRB
KILOMETERS
9611212814416017619296112128144160176192
KILOMETERS
WEST WEEK 7 WEEK 8 EAST
KILOMETERS
h)
EAST WEEK 9 EAST
TRIP 1 28 OCT
28 OCT TRIP 2
I'1
X X
JNVHAW ITL PLKH8TCRI
i fl
NV HAW ITL
X X
PLKHST CRB
96 11212814416017619296 112128144160176192
KILOMETERS


49
located Gainesville, as well as the two coastal areas were the only locations with
statistically significant movement (chi-square, p < .05). When the sampling period was
subdivided to early, middle and late migration, a single, statistically southward biased
peak shifted eastward from Steinhatchee to Gainesville to Crescent Beach, respectively.
The total fall migration resolved into three peaks, which may represent three pathways
into central Florida. Large numbers of long tailed skippers were counted near the west
coast, but the highest counts were in the central region, passing through Gainesville.
A third peak in migration began at the St. Johns River in Palatka and extended to the
east coast. Dave Baggett (pers. comm.) has reported large numbers of II. proteus
migrating along the St. Johns River during the fall of some years. This pattern may
reflect some geographical obstacles, visual attractions, habitat preferences or source
areas, such as bean fields.
Migration Profiles
Cloudless sulphur
Profile estimates of Phoebis sennae migration in 1986 across the west-east
transect at the Gainesville latitude are shown in Fig. 3-4 a,b,c,d,e,f,g,h. The peak net
southward movement along the western half of the transect (=STN-GNV portion)
occurred through the TRN-NWB or NWB-GNV segments on 14 October, during week
7 of sampling. Throughout the first five weeks, the movement along the western transect
remained fairly constant and relatively low, <_ 2 indiv/min. During week 6, this rate
doubled and in week 7, counts in the western segments (Steinhatchee-Newberry)
increased dramatically. Along the peak transect segment (CRC-TRN, Fig. 3-4g, west)


26
computer program designed by Walker (1985a) to calculate mean direction (M.D.),
length of mean vector (r) and the 95% confidence interval of M.D. (Batschelet, 1981,
Zar, 1984).
Mark-Release-Recapture
A number of individuals of Phoebis sennae were also marked and released
during the fall migration season of 1987. This mark-release-recapture took place at two
sites in Gainesville, Florida, from September through December. Both sites had gardens
with a number of flowering plants that the adults used as nectar sources. The most
northern site was Kanapaha Botanical Gardens and the second site was a Gainesville
residence located eight miles south of Kanapaha Gardens on Williston Rd. Between my
visits, observations of marked individuals were recorded by the resident, Dr. Lincoln
Brower, resulting in more complete information at this site. A total of 800 cloudless
sulphurs were netted over a period of three days from 10-12 September and given a
number on the upper and lower surfaces of the right wing with a permanent felt tip
marker, color coded for each site. Sites were visited three times per week until 23
November 1987 and all marked individuals were noted. Observations at the residential
site were continued by Dr. Brower throughout December.
Results and Discussion
The mean flight directions during fall migration 1985-1988 through various sites
in Florida for Phoebis sennae are shown in Fig. 2-1. The length of the line represents


5
been determined (Meier and Fivizzani, 1980). Beall (1948) reported that monarchs
captured in Ontario, before flight, were heavier than those caught after flight in
Louisiana. Fat stores are conserved carefully by thermoregulation through the over
wintering period and must remain sufficient to at least fuel the return journey to new
feeding or breeding grounds (Masters, 1988).
There has been very little experimental investigation of migration in the
laboratory. Most studies in this area have been done with insects in which flight activity
and/or orientation can be easily measured, and usually in tethered flight (Dingle, 1985).
So far, researchers have been unable to measure a butterflys level of flight activity and
preferred direction in the laboratory. Flight traps (Walker, 1985b; Walker and
Lenczewski, 1989), that exploit the behavioral characteristic of migrants to go up and
over obstacles, rather than around, can segregate these individuals in the field, but the
problems in the laboratory remain.
Florida Butterfly Migration
In a long term study using permanent flight traps in Gainesville, Walker (1991)
has identified at least eight species of butterflies that migrate southward each fall through
the Florida peninsula. The four principal migrants represent four families of
Lepidoptera: Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae (L.)
(Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long-tailed
skipper; and Precis coenia (Hbner) (Nymphalidae), the buckeye. These species
comprised 85% of Walkers (1991) total trap catch and they also exhibited bidirectional


CHAPTER 3
NUMBERS OF MIGRANTS
Introduction
Each fall, a migratory mass of butterflies comprised of at least eight species
from four families, passes through north central Florida. The most conspicuous of these
migrants is a pierid, Phoebis sennae (L.), commonly known as the cloudless sulphur.
This butterfly has been reported throughout the southeast during fall migratory flights
which are usually toward the Florida peninsula (Walker, 1985a). The three other
principal migrant species are Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary;
Urbanus proteus (L.) (Hesperiidae), the long-tailed skipper; and Precis coenia (Hiibner)
(Nymphalidae), the buckeye. Large numbers of these species have been noted along the
Florida Gulf coast (Urquhart and Urquhart, 1976) as well as passing through north
Florida regularly (Walker 1985ab, 1978, 1980, 1991).
It has been difficult to estimate the size of this migratory flight. Walker (1991)
has operated traps for more than ten years in Gainesville, Florida, but the magnitude of
migration had not been monitored at any other location. In order to accurately estimate
the total numbers of butterflies moving south into central Florida each fall, the density
of migratory flights must be monitored at other locations along the latitude of Gainesville
(29.65). Walkers (1991) long term trapping results from Gainesville could then be
38


Ill
b)
GAINESVILLE
z 60
O
H 50
t 40
z 30
£ 20
10
123466789 101112131416
WEEKS
1 2 3 4 6 7 S 9 10 11 12 13 14 16
WEEKS
LAKE ALFRED
12346678 9 10 1112131416
WEEKS
e)
O
<
o
<
-
o
I-
#
WEEKS
50
40
30
20
10
0
123466789 101112131416
LAKE PLACID
N-0
X
WEEKS
Fig. 4-15 a,b,c,d,e.-Summary of percent total net trap catches of Urbanus proteus at
five sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake
Alfred and Lake Placid, Florida) during 13 September-12 December 1987 and 28
August-26 November 1988. Solid bars represent net southward movement and clear bars
show net northward movement. Total net catch (southbound minus northbound) is shown
in the upper right comer, x=trap not operating. Data from the first 4 weeks of 1987,
and all of 1988, collected in Gainesville are from Walker (1991).


14
recorded in Mississippi and Alabama (Mather and Mather, 1958; Lambremont, 1968),
apparently heading southeast toward the Florida peninsula. Walkers (1985a) observation
of systematic changes in P. sennaes flight direction at many locations in the southeast
suggests that these migrants are navigating toward peninsular Florida. The cloudless
sulphur does not exhibit the coastal build up of other species mentioned and, as Walker
has suggested (pers. comm.), may detect the "proper" time to change its flight direction
before hitting the Gulf or Atlantic coastlines from up to 60 miles away! Gaddy and
Laurie (1983) note some consistent discrepancies in flight direction along the South
Carolina coastline for Phoebis sennae. In August, September and early October of 1978-
1980, they noted northeastern migrations along the immediate coast and, what seemed
to be random movements, inland. As previously discussed, flights adjacent to coastlines
can be erratic. Migrating butterflies often seem "confused" and fly in what appear to be
inappropriate directions, more often than not following the coastline in either or both
directions.
Adult males are yellow above and unmarked. Females are a deeper, more
orange yellow, fringed with dark brown marginal spots. The adults are particularly
attracted to red flowers, many of which flower in the fall and are an important source
of nectar for migrating butterflies during that time (pers. obs.). Larvae are a pale
yellowish green with a yellow lateral stripe along each side. They are most often found
at the young shoots of their leguminous hostplants and sometimes web leaves together
for shelter (Howe, 1975). The larvae of this butterfly feed on a wide range of
leguminous hostplants. There are at least 50 species of hostplants recorded and adults


128
Meier, A.H. and A.J. Fivizzani, 1980. Physiology of migration. Pp. 225-282 in
Animal Migration, Orientation, and Navigation, S.A. Gauthreaux, Jr. (ed.),
Academic Press, New York. 387 pp.
Muller, J. 1977. Notes on migrating Phoebis sennae eubule Linn. (Pieridae). News
Lep. Soc. 4.
National Geographic. 1979. Bird Migration. 156:154A (Suppl.).
Nielsen, A. 1961. On the habits of the migratory butterfly Ascia monuste L. in Florida.
Biol. Meddel. 23:1-81.
Nielsen, A. and E.T. Nielsen, 1952. Migrations of the pierid butterfly Ascia monuste
L. Florida. Entomol. Mededeelingen 26:386-391.
Pedgley, D. 1982. Windbome Pests and Diseases, Meteorology of Airborne Organisms.
John Wiley and Sons, New York. 250 pp.
Penn, G.H. 1955. Mass flight of ocola skippers (Panoquina ocola Edwards). Lep. News
Soc. 9:79.
Pyle, R.M. 1981. The Audobon Field Guide to North American Butterflies. Alfred A.
Knopf, New York. 917 pp.
Rainey, R. C. 1951. Weather and the movements of locust swarms: a new hypothesis.
Nature (Lond.) 168:1057-1060.
Rankin, M.A. 1978. Hormonal control of insect migratory behavior. Pp. 5-32 in
Evolution of Insect Migration and Diapause, H. Dingle (ed.), Springer-Verlag,
Berlin and New York. 284 pp.
Riley, N.D. 1975. A Field Guide to the Butterflies of the West Indies. Collins,
London. 224 pp.
Roer, H. 1959. Uber Flug- und Wandergewohnheiten von Pieris brassicae L. Z.
Angew. Entomol. 44:272-309.
Roer, H. 1961a. Zur Kenntnis der Populationsdynamik und des
Migrationsverhaltens von Vanessa atalanta L. im palaarktischen Raum. Beitr.
Entomol. 11:594:613.
Roer, H. 1961b. Ergebnisse mehijahriger Markierungsversuche zur Erforschung der
Flug- und Wandergewohnheiten europaischer Schmetterlinge. Zool. Anz.
167:456-463.


3
migratory movement. Although not necessarily so, a cyclic, or seasonal pattern is often
seen in these migratory movements. The same individual may return, at least part way,
to re-colonize old habitats, or, the return may consist of a step-wise expansion of new
generations into the former breeding range during favorable conditions (Walker, 1985a).
Baker (1969) has speculated on the evolution of migration and the maintenance
of migratory behaviors in populations. The factors that initiate migration are varied and
have been found to be finely attuned by genetic programming (Lamb and McKay, 1983)
to changes in habitat quality over time and space (Southwood, 1962) and other
environmental cues (Dingle, 1972). The influence of genetic factors on migratory
tendency in members of a population has been investigated empirically by Dingle gt a!
(1977) and Istock (1978), and theoretically by Roff (1975). It is known that individual
dispersal tendencies can be highly variable within populations (Baker, 1978) and these
characters have been influenced by experimental selection to some degree (Dingle, 1968;
Rankin, 1978). There are a number of possible life history strategies available to an
insect and how natural selection operates on these genetic factors in the evolution of the
migratory habit is still unclear.
How migrants find their destinations, that is, their navigational techniques,
(Schmidt-Koenig, 1975), also remain a mystery. Monarch butterflies are highly specific
in their annual migration to overwintering sites in Mexico (Calvert and Brower, 1980).
Walker (1985a) suggested that cloudless sulphurs from the southeast navigate toward the
Florida peninsula.


117
a)
VALDOSTA
1 2 3 4 6 6 7 8 6 10 111213 1416
WEEKS
e)
b)
GAINESVILLE
WEEKS
LAKE ALFRED

N = 7

N

N
rfiri..
u
XXX
1 2 3 4 6 6 7 8 9 10 11 12131416
WEEKS
LAKE PLACID
WEEKS
Fig. 4-18 a,b,c,d,e.Summary of percent total net trap catches of Precis coenia at five
sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred
and Lake Placid, Florida) during 13 September-12 December 1987 and 28 August-26
November 1988. Solid bars represent net southward movement and open bars show net
northward movement. Total net catch (southbound minus northbound) is shown in the
upper right comer, x=traps not operating. Data for the first 4 weeks of 1987 and all of
1988 Gainesville data are from Walker (1991).


80
speed of migrating gulf fritillaries and long tailed skippers to be approximately 12-22 kph
under ideal conditions. Weather permitting, peak migration times are usually 0900-1600
HRS LMT (Walker, 1985a) daily, suggesting that individuals may travel as much as 100
km/day. If the same individuals that pass through Gainesville during the peak of
migration move directly to lower latitudes, a similar pattern of peaks would be expected
at sites to the south. However, if peaks lag to the south by about 3 weeks, the
approximate generation time for gulf fritillaries (Arbogast, 1965, 1966), movement south
could be in a stepwise, generational manner.
Materials and Methods
Longitudinal Pole Counts
During 1986, pole counts (see description of methods in Chapter 3) were
initially made at Gainesville and five sites south, Clermont, Bartow, Arcadia, LaBelle
and Immokalee (see Chapter 1, Fig 1-1 for location of all sample sites). The primary
purpose of these counts was to get some information on when migration began and how
far it extended south of Gainesville before planning more extensive data collection with
traps. Counts were made from 13 September to 2 November 1986, at approximately two
week intervals (see Chapter 1, Table 1-1 for dates of all sampling). This period was
figured by Walker (1991) to include 95% of the total seasonal migration through
Gainesville. The six sites along the north-south transect were spaced approximately 62
km apart, a total distance of 372 km. All net numbers were adjusted as explained
previously (Chapter 3) for time of day differences (T) using Walkers (1985a)


53
WEST
WEEK 4
WEEK 6
EAST
KILOMETERS
b) WEST wee|( 7 WEEK 8 EAST
KILOMETERS
Fig. 3-5 a,b.--Profile estimates of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65) from 26 September to 18 October
1987. Counts were taken on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW = Hawthorne, ITL = Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


108
period, but the net movement for the season was significantly southward. There were
no long tailed skippers captured in flight traps at Lake Placid (Fig. 4-13 e), although
several adults were observed feeding at flowers during each visit.
The percent total net trap catch of Urbanus proteus at five north-south sites
during 1988 is shown in Fig. 4-14 a,b,c,d,e,f,g,h,i. All statistically significant (chi-
square, p < .05) net movement was southward. Numbers of individuals captured at
Valdosta were small, with a total net catch of 16 for traps 1 and 2 (Fig. 4-14 a,b).
There was no statistically biased movement revealed by trap 1, but trap 2 had significant
(p <.05) net southward movement during week 5. Walkers (1991) Gainesville data
indicates the mid-migration point at that site was during week 9 (Fig. 4-14 c,d). This
is also true for Leesburg, both trap 1 and 2 (Fig. 4-14 e,f). The average midpoint of the
two Lake Alfred traps (Fig. 4-14 g,h) was during week 10. South of Valdosta, where
significant net movement south ended after week 5, all sites sampled (except Lake Placid)
showed net southward movement through at least week 12. No long tailed skippers were
captured at Lake Placid (Fig. 4-14i).
The summary of percent total net trap catches of Urbanus proteus for 1987 and
1988 is shown in Fig. 4-15 a,b,c,d,e. All statistically significant (chi-square, p <.05)
net movement is southward. The average midpoint of migration for these two years in
Valdosta was during week 5. Gainesville migration midpoint occurred during week 7.
All movement stopped at Valdosta by week 9, but continued at low levels in Gainesville
through at least week 13. Leesburg migration midpoint was also during week 7 and
movement there ended by week 13. Lake Alfred had a bimodal pattern showing strong


125
Beall, G. 1946. Seasonal variation in sex proportion and wing length in the monarch
butterfly, Danaus plexippus L. Trans. R. Entomol. Soc. (Lond.) 97:337-353.
Beall, G. 1948. The fat content of a butterfly, Danaus plexippus L., as affected by
migration. Ecology 29:80-94.
Beall, G. and C.B. Williams, 1945. Geographical variation in the wing length of Danaus
plexippus. Proc. R. Entomol. Soc. (Lond. A) 20:65-76.
Beenakkers, A.M.T. 1965. Transport of fatty acids in Locusta migratoria during
sustained flight. I. Insect Physiol. 11:879-888.
Brower, L.P. 1961. Studies on the migration of the monarch butterfly. I. Breeding
populations of Danaus plexippus and D. gilippus berenice in south central
Florida. Ecology 42:76-83.
Brower, L.P. 1962. Biology of the monarch butterfly. Ecology 43:181-182.
Brower, L.P. 1977. Monarch migration. Nat. Hist. 86:40-53.
Brower, L.P. and Calvert, W.H. 1985. Foraging dynamics of bird predators on
overwintering monarch butterflies in Mexico. Evolution 39:852-868.
Calhoun, J.V., T.J. Allen and D.C. Iftner, 1990. Temporary breeding populations of
Phoebis sennae eubule L. (Lepidoptera: Pieridae) in Ohio and West Virginia.
J. Res. Lepid. 28:123-125.
Calvert, W.H. and L.P. Brower, 1986. The location of monarch butterfly (Danaus
plexippus! overwintering colonies in Mexico. J. Lep. Soc. 40:164-187.
Calvert, W.H., L.E. Hedrick and L.P. Brower, 1979. Mortality of the monarch
butterfly (Danaus plexippus L.): avian predation at five overwintering sites in
Mexico. Science 204:847-851.
Clark, A.H. and L.F. Clark, 1951. The Butterflies of Virginia. Smithsonian Institution,
Washington. 239 pp.
Danthanarayana, W. 1986. Introductory chapter. Pp. 1-10 in Insect Flight: Dispersal
and Migration, W. Danthanarayana (ed.), Springer-Verlag, New York. 289 pp.
Dingle, H. 1968. Life history and population consequences of density, photo-period, and
temperature in a migrant insect, the milkweed bug, Oncopeltus. Am. Nat.
102:149-163.


/ : & 6 !
BUTTERFLY MIGRATION
THROUGH THE FLORIDA PENINSULA
By
BARBARA LENCZEWSKI
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
1992


122
latitudinal gradient. This, in conjunction with a mark-release-recapture and trapping
program for adult populations would explore the relationship between resource
availability and migratory movements. Breeding patterns need to be described along a
north-south gradient in order to identify generational movements. A large scale tagging
program throughout the southeast is the most definitive way to identify the long distance
movements of individuals. The recent development of an inexpensive, and very portable
flight trap (Walker, pers. comm.), makes this type of program more feasible, and one
has been undertaken by J. Whitesell (pers. comm.) in south Georgia with the
participation of local high schools over the past several years.
How northern regions are recolonized is also intriguing and particularly
important with regard to planning future conservation practices. Even with so many
millions of individuals moving through the peninsula, no specific "overwintering" area
has been identified. From the patterns seen during the three years this study was
conducted, it appears that the western half of south central Florida is the limit of
significant movement south. There are large areas of wilderness, much of this within
the Southwest Florida Water Management District, along the Withlahoochee River which
should be investigated as potentially vital winter survival sites. It is possible that a very
small percentage of individuals survive scattered along the southern part of the migratory
route, and these slowly repopulate the northern range the following year. This also
appears to fit the gradual population buildup which seems to occur throughout the
summer in north Florida. If this is true, then a very large genetic pool is drastically
reduced each year and any large scale catastrophic incident may significantly effect the


39
projected to estimate the absolute numbers moving south each fall. The goal of this
study was to construct a "migration profile" of density estimates for the two principle
migrants, Phoebis sennae and Agraulis vanillae. along a transect across north central
Florida at the latitude of Gainesville. This cross section through the migratory stream
will then be used to estimate the total fall migration of Phoebis sennae and Agraulis
vanillae into peninsular Florida.
Materials Methods
Latitudinal Pole Counts
In the fall of 1986, a counting system previously used by Walker (1985a) was
used to estimate the numbers of migrants moving through various sites along the
Gainesville, Florida, latitude (29.65). Three 3-meter lengths of PVC pipe were used
with the second pipe set in the ground 15 meters from the first, and the third, 30 meters
beyond the second. All three pipes were along a straight line at 51-231, perpendicular
to the average flight track of 141 established in Gainesville by Walker (1985a). All
observations were made in large, clear areas (i.e. athletic fields or lawns) to avoid any
vegetative or structural influence on the paths taken by migrants. The air temperature
(C), percentage of clear sky and sun exposure (b=bright, h=hazy, o=obscured by
clouds, p=disk obscured, but position easily determined 5) were noted. Wind speed
was measured with a hand held pith ball anemometer and direction was noted. The
numbers of individuals of the four principal migrant species were counted crossing
between two poles for three five-minute periods with a one minute break between each


81
observations of daily flight periodicity. These adjusted net numbers were averaged to
yield a seasonal summary for each site along the north-south transect.
Portable Flight Traps
For the purposes of this study, a portable flight trap was designed (Walker and
Lenczewski, 1989) to quantify the phenology of butterfly migration along the Florida
peninsula. Walker (1985a) has determined that the migratory track through Gainesville,
Florida remained relatively constant at 141 annually. A migratory route was projected
north and south (321 and 141) from Gainesville and pairs of these flight traps were
erected on it at five sites (an average of 105 km apart) from southern Georgia to south
central Florida: Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida. Migration was monitored with these traps through the fall of 1987 and 1988.
The traps were placed in pairs along an ENE-WSW line (67.5-247.5), one trap facing
north and one facing south. These traps, erected approximately perpendicular to the axis
of the Florida peninsula and to the mean direction of migration, were left in place from
13 September through 12 December 1987. Butterflies captured by the south-facing trap
were flying 321 90; those captured by the north-facing trap were heading 157.5
90. All traps were situated in large, open fields to eliminate the influence of vegetation
or other obstacles on the preferred flight tracks of migrants. As the migrant butterflies
tried to fly over a main wall of polyester screening, they passed through a narrow slot
into a duct that led to two holding cages made of 1/4 inch hardware cloth. Captured
butterflies were removed from the holding cages semiweekly throughout the trapping


37
two sites where there were more than two individuals recorded, the movement was
southward. The only location where buckeyes were sighted south of Gainesville was in
Bartow, central Florida and those few individuals were headed north.
Mark-Release-Recapture
In two instances, cloudless sulphurs were recaptured at some distance from the
mark-release sites in Gainesville. These are the first instances of an individual Phoebis
sennae being tracked over a relatively long distance and time. The first individual had
been marked at Kanapaha Gardens and was seen again feeding at flowers at the Williston
Rd. residence, eight miles south of there, ten days later. The second individual had been
marked at the Williston Rd. residence and was recaptured in Bronson, Florida, 51 km
southwest of Gainesville, 14 days later. This cloudless sulphur had evidently taken
refuge in some potted plants during a cold evening and had been brought into the house.
In both instances, the butterflies travelled a much shorter distance than would be expected
from the elapsed time. It has been calculated (Arbogast, 1966; Balciunas and Knopf,
1977) that migrating cloudless sulphurs or gulf fritillaries go 15-20 km/hr under ideal
conditions. Thus, such a journey could easily be completed in one day.


98
bidirectional movement throughout the season. The midpoint of migration at Leesburg
was also during week 5 and this was the only week of statistically significant net
movement (biased southward). During visits to this site throughout the collecting period,
Passiflora incamata. a larval hostplant, was noted as abundant in the surrounding
cultivated crop fields. Gulf fritillaries, especially females, were observed visiting and
ovipositing on these plants. The presence of these hostplants may have affected flight
trap catches by diverting individuals to those sites and the butterflies may have remained
longer, taking advantage of breeding sites. It was seen from captures at Lake Alfred
(Fig. 4-8d), some 70 kilometers south of Leesburg, that the southward migration
continued well into week 12 of the season although the median was in week 8. The
favorable breeding conditions at Leesburg may have caused the 2-3 week delay seen at
Lake Alfred. There also may have been a high population of residents (non-migratory)
at Leesburg which could have been travelling short distances and were also trapped.
This may have resulted in the trapping pattern seen at Lake Alfred. From these data,
it appears that migration through Valdosta, Gainesville and Leesburg (a distance of 265
km) occurs during weeks 4 or 5, but the movement does not reach Lake Alfred, only 90
km south of Leesburg until 2-3 weeks later. Using an extremely conservative estimate
of travel time, individuals that fly south through Gainesville would be expected to reach
Lake Alfred within two weeks. According to estimates of adult longevity for the gulf
fritillary by Arbogast (1965), unmated adults had a mean life span of 18.4 3.0 days
in captivity. It seems reasonable to assume that in nature, subjected to various stresses,
the normal life span may be shorter. Arbogasts (1965) estimates of the developmental


This work is lovingly dedicated to my parents, Lucjan and Stanislawa
Lenczewski, who have waited half their lives to see it and to my husband, Joseph
lowers, for believing that he would.


This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy. )
December 1992
cT Dean, College of Agriculture
Dean, Graduate School


6
seasonal movement -- i.e. north in the spring and south in the fall. The total size of the
fall migration of these species has been estimated using preliminary data from this study
by Walker (1991) at about 80 million individuals. This is comparable to the estimates for
monarchs at Mexican overwintering sites (Calvert ei al, 1979). The most significant and
noticeable movements through the Florida peninsula are exhibited by the cloudless
sulphur and the gulf fritillary, which have been estimated by Walker (1991) to comprise
nearly 80% of the fall migration, or 64 million individuals annually.
A number of other species also show less directed or irregular migratory
movements (Walker, 1991). These include four pierids, Pieris rapae (L.), the cabbage
white; Eurema daira (Godart), the dainty sulphur; Eurema lisa (Boisduval & LeConte),
the little sulphur; Eurema nicippe (Cramer), the sleepy orange, as well as a nymphalid,
Vanessa virginiensis (Drury), the painted lady. Two hesperiid skippers are known to
migrate southward in the fall through north Florida, although there is very little known
of their movements since they are too small to be retained by Walkers (1991) traps.
They are Lerema accius (Smith), the clouded skipper and Panoquina ocola (Edwards),
the ocola skipper. Walker (1991) also reported some individuals of the monarch
butterfly, Danaus plexippus (L.), as passing through Gainesville. The majority of the
monarch migration, however, is southwest toward Texas and Mexico (Brower, 1977).
The recurring spectacle of huge butterfly migrations along the Gulf coast
(Urquhart and Urquhart, 1976) and through north central Florida (Walker, 1978, 1980,
1985ab, 1991) is locally well known, but is inadequately represented in the literature.
The Florida Gulf coast is a renowned stopover along the major North American


10
At the very least, understanding why, how and where these butterflies are going
presents a fascinating puzzle. As stated previously, outside of Gainesville, Florida,
virtually nothing is known of the phenology or extent of the migratory movements. The
goal of this study was therefore to contribute to general knowledge of the migration,
primarily of Phoebis sennae and Agraulis vanillae. elsewhere in peninsular Florida.
Research Plan
The investigations in this study focused on the migratory patterns in peninsular
Florida of the two principal migrant butterflies, Agraulis vanillae and Phoebis sennae.
Some attention was also given Urbanus proteus and Precis coenia. The research was
designed as a four part study, carried out during the falls of 1986 through 1988 and
addressed the following questions:
1. Flight direction Is the mean flight direction of migrants throughout Florida the same
as in Gainesville?
2. Numbers of migrants Is the migration density across Florida along the Gainesville
latitude constant? How many migrants pass into central Florida annually?
3. Phenology of movement Does the peak migration at southern sites occur at
approximately the same time as it does in Gainesville? Where does the
migration stop?
Estimates of migration density for the cloudless sulphur and gulf fritillary were
made along the Gainesville latitude (29.65). From these data, a "migration profile" or
cross section, through the stream of migration for each species was constructed. This


west of Gainesville, respectively. Both species avoided the Atlantic coast. Using
previous data from Gainesville flight trap catches in conjunction with density estimates
from this study, the total number of cloudless sulphurs and gulf fritillaries moving south
each fall into central Florida was estimated at 42 and 115 million individuals,
respectively. When previous estimates for the long tailed skipper and buckeye were
included, the total number of butterflies migrating south through the Florida peninsula
was estimated to be 175 million individuals annually. Observations of flight azimuths
for three species (excluding the buckeye) were made at various locations south of
Gainesville during fall 1986. Most significant mean flight directions were consistent with
the 141 previously described for Gainesville and the limit of significant southward
movement for that year was observed just south of Lake Ockeechobee for all three
species. The phenology of the migration was studied by trapping the four major migrant
species using directional flight traps at five stations from south Georgia to south central
Florida during fall 1987-1988. All significant net movement was southward, extending
to the Lake Alfred site, about 192 km south of Gainesville. There was no evidence from
trap catches or drive counts at the Lake Placid latitude (27.45) that large numbers of
migrants move into extreme south Florida to overwinter.
Vlll


47
STN CRC TRN NWBGNV HAWITL PLK HST CRB
O.D r
i I T I I I ~T1
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-2.--Summary of pole counts showing Agraulis vanillae migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p C.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB = Newberry, GNV=Gainesville, HAW=Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.


44
difference. The results of these calculations yielded the estimated migration profiles for
1986, 1987 and 1988.
D = NET # INDIVIDUALS/MINUTE FROM DRIVE COUNTS
T = TIME OF DAY ADJUSTMENT
D, = NET # INDIVIDUALS/MINUTE AT BASE SITE (TIME CORRECTED)
I = MIGRATION INDEX
ESTIMATED MIGRATION PROFILE = I (D^T/Dj).
The summary migration profiles were then derived from the median segment-by
segment values of the yearly estimated profiles for each species. These cross sections
of migration density through north central Florida were used to estimate the annual
number of P. sennae and A. vanillae butterflies flying into central Florida by applying
the migration profiles to Walkers (1991) Gainesville permanent trap data.
Results and Discussion
Latitudinal Pole Counts
Cloudless sulphur
Pole counts for the fall 1986 cloudless sulphur migration are summarized in Fig.
3-1. The two coastal sites showed no net movement. At Steinhatchee, this was due to
equal numbers of individuals flying north and south, and, at Crescent Beach it was due
to a lack of migrants. All net movement southward was statistically significant, except
for Palatka, also near the east coast, where not many individuals were seen. Peak


Fig. 3-11 a,b,c.--Yearly migration profiles of Agraulis vanillae movement across a west-
east transect at the latitude of Gainesville (29.65). The yearly migration profiles are
from the median estimates (net number of individuals/minute, adjusted for time of day,
sample size and season) during 13 September-28 October 1986,26 September-18 October
1987 and 18 September-12 October 1988. Counts were made on nine transect segments
defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen,
PLK=Palatka, HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically
significant (chi-square, p < .05)southward movement and open bars were sites with no
statistical bias (chi-square, p < .05) in net movement. Number of visits to each site is
above bars.


107
a)
b)
VALDOSTA
GAINESVILLE
1 2 3 4 B 6 7 8 9 101112131416
WEEKS
C)
d)
GAINESVILLE PTP#3&5 (WALKER,1991)
2 3 4 5 7 8 9 10 11 12 13 14 16
WEEKS
LEESBURG
12 3 4
7 8 a 10 11 12 13 14 16
WEEKS
a)
LAKE ALFRED
1 2 3 4 6 6 7 8 9 101112131416
WEEKS
LAKE PLACID
WEEKS
1 2 3 4 6 8 7 8 a 10 11 12 13 14 16
WEEKS
Fig. 4-13 a,b,c,d,e,f.-Weekly percentage of seasonal total net trap catch of Urbanus
proteus at Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida during 13 September-12 December 1987. Solid bars represent statistically
significant southward biased movement (chi-square test, p < .05), shaded bars show
nonsignificant (p > .05) southward movement. Clear bars show nonsignificant (p > .05)
northward movement. The total net catch for the period is shown in the upper right
comer. The midpoint of migration is designated by an asterisk. x=trap not operating.


CHAPTER 4
PHENOLOGY OF MOVEMENT
Introduction
Many migratory insects, such as aphids and locusts, are greatly dependent on
wind and weather conditions to direct their flight once they are airborne (Pedgeley, 1982;
Taylor, 1958). These insects, as well as nocturnal moths, fly long distances at high
altitudes requiring the use of sophisticated, expensive technology, such as radar, for
study (Rainey, 1951; Schaefer, 1976) or extensive mark-release-recapture programs (Li
et ai, 1964; Showers e ai, 1989). Butterflies are particularly suitable as subjects for the
study of insect migration because they are daytime fliers and move within a variable
boundary layer, usually within 4-6 meters of the ground (Edwards and Richman, 1977).
Wind speeds within this layer are low enough to allow individual control of flight
direction (Johnson, 1969). As the butterflies move near ground level, an observer can
identify species as well as note behavior and preferred flight tracks. Migrating butterflies
typically continue their flight on a linear track and will fly up and over obstacles rather
than around (Williams, 1930). This behavioral characteristic can be exploited to
selectively capture migrants for research purposes and to quantify movement by using
flight traps. Butterflies also have large wing surfaces which make it relatively easy to
conspicuously mark individuals for recapture (Walker and Wineriter, 1981).
77


69
Walker (1991) estimated the size of the seasonally recurring migration through
peninsular Florida by extrapolating his flight trap catches from Gainesville using a
preliminary estimate of the migration profile from this study. His estimate of the net
movement of migrating butterflies each fall was an average of 22.2 million individual
cloudless sulphurs and 41.3 million gulf fritillaries. The combined figure of 63.5 million
butterflies approaches the estimated number of monarchs reported to overwinter at
Mexican sites (Brower, 1985; Calvert gt ai, 1979). The average rate of net movement
for Phoebis sennae and Agraulis vanillae (1986-1988) from the migration profile of each
species was used to recalculate Walkers (1991) data (Table 3.1). Walkers (1991)
Table 3.1 Estimates of the Phoebis sennae and Agraulis vanillae fall migration into the
Florida peninsula (Lenczewski, 1992 and Walker, 1991). Average absolute net numbers
of individuals flying across a north central Florida transect at the latitude of Gainesville
(29.65).
Mean profile height (Lenczewski, 1992)
££
1.03
£E
2.10
NWB-GNV profile height
1.03
1.43
Profile correction factor
1.00
1.47
Kilometers across state (Lenczewski, 1992)
190.00
190.00
Migrants trapped/6m (Walker, 1991)
800.00
868.00
Est. migrants into cen. Fla. (millions)
25.30
40.40
Est. trapping efficiency (Walker, 1985b)
0.60
0.35
Adj. migrants trapped/6m
1333.00
2480.00
Est. migrants into C. Fla. (millions)
42.20
115.00


116
Valdosta showed any significant net movement during the season. Walkers (1991) data
from Gainesville revealed a median at weeks 6 and 7 for traps 3 and 5 (Fig. 4-17 c,d),
respectively. Leesburg trap 1 (Fig. 4-17e) netted the most individuals of any site (n = 17)
sampled and was the only site with significant net southward movement (week 4). There
were no individuals caught in trap 2 (Fig. 4-17f) at Leesburg. Both traps at Lake Alfred
recorded slight net movement northward, and Lake Placid traps caught only one
individual during the entire season.
A summary of percent total net trap catches of Precis coenia for 1987 and 1988
is shown in Fig. 4-18 a,b,c,d,e. Overall, the only statistically significant (chi-square,
p C.05) net movement for this species during 1987-1988 sampling periods was
southward. Mid-migration at all sites occurred between weeks 3-5. All significant
movement ceased after week 7 in Valdosta, but continued at low levels through at least
week 13 in Gainesville. Movement south continued at Leesburg and Lake Alfred through
week 11. Only one individual was trapped in Lake Placid. As was shown for the other
three migrant species, the lack of substantial trap catches or large numbers of individuals
suggests there is no significant migration through Lake Placid. These indications are
further substantiated by Dr. Mark Deyrup, insect ecologist, Archbold Biological Station,
Lake Placid, Florida. Gulf fritillaries were seen ovipositing on Passiflora during the fall
of 1988 and occasionally large flights of Eurema daira were observed, but no other
conspicuous migratory flights were noted (M. Deyrup, pers. comm.).


34
FROSTPROOF FISHCREEK
O o
DMTERLACHEN
OLGA OKEECHOBEE
o
ST. GEORGE TRENTON
GLEN ST. MARY HAWTHORNE
HASTINGS
LAKE CITY

LABELLE LAKE PLACID NEWBERRY
TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION

Fig. 2-4.Azimuths of Agraulis vanillae flight directions taken at various locations in
Florida during fall migration 1985-1988. The circle diameter represents the number of
visits (5 mm = l); the length and width of the line represents the number of individuals
observed (2 mm = l); the orientation of the line represents the flight direction.


Fig. 4-9 a,b,c,d,e,f,g,h,i.-Weekly percentage of total net trap catch of Agraulis vanillae
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August
through 26 November 1988. Data for Gainesville, Florida are from Walker (1991).
Solid bars represent statistically significant southward biased movement (chi-square test,
p < .05). Shaded bars show nonsignificant (p > .05) southward movement. Clear bars
show nonsignificant (p > .05) northward movement. The total net catch for the period
is shown in the upper right comer. The midpoint of migration is designated by an
asterisk, x=traps not operating.


33
observed throughout Florida during fall 1985-1988 are shown in Fig. 2-4. Like P.
sennae. Agraulis vanillae showed a great deal of variation in flight direction along the
Florida panhandle coast at sites such as Alligator Point. However, at
Steinhatchee, which is further south along the west coast, there was no evidence of the
confused flight that 2. sennae had shown and movement was significantly biased
southeast.
The summary of mean flight directions of Urbanus proteus throughout Florida
are shown in Fig. 2-5. Its migratory flight patterns are similar to those of the gulf
fritillary. Also common along the Gulf coast, this species exhibited a great deal of
variation in flight at coastal sites. This is evident in the individual flight directions from
various sites in Florida shown in Fig. 2-6. For example, at Keeton Beach, movement
was in every direction. Oddly, a similar variation in flight was noted at Yeehaw
Junction, an inland location just north of Lake Okeechobee. However, a few miles
directly south, at Okeechobee, the flight of individuals was very consistently due south.
It appeared they were heading out across the water, but just south of the lake, at
Clewiston and Belle Glade, there were no sightings of incoming individuals. It is
possible that the long tailed skipper makes shorter, more localized flights than the other
two species discussed at inland sites. This may also account for a great deal of
variability in flight directions.
Very few individuals of Precis coenia were recorded throughout the state during
fall migrations. This is consistent with Walkers (1991) findings from flight traps in
Gainesville that most of the movement is northward in the spring. In this study, of the


29
possibly resulting in freezing, could be prevented. Dave Baggett has reported
(pers.comm.) large numbers of long-tailed skippers flying south along the St. Johns
River near Jacksonville during some years. This geographical feature may serve as a
path when found and, if it can be detected at a distance, may create a funnel drawing
migrants away from the east coast areas. Calhoun el al (1990) reported some
colonization by cloudless sulphurs along rivers that appeared to serve as "dispersal
corridors." It has also been observed (see Chapter 3) that this "shadow" or absence of
migration along the east coast seemed to widen inland with distance travelled south.
Other possible reasons for the absence of these migrating butterflies along the Florida
east coast may be the lack of an appropriate habitat or some other adverse environmental
conditions found there.
Throughout the state, significant migratory flight directions for Phoebis sennae
are usually south or southeast. It is noteworthy that in this species, there appears to be
a shift in mean direction immediately north of Lake Okeechobee, as though the migrants
"detect" the lake. At the town of Okeechobee, the mean migratory track shifted to the
east and most individuals (79%) avoided flying toward the large expanse of water. No
cloudless sulphurs were observed along the southern boundary of the lake at Belle Glade,
just a short time after seeing many individuals flying through Okeechobee. Again, this
seemed to be consistent with observations that the butterflies at the northern border of
the lake were going around, rather than flying directly over water. A distinct southward
migratory flight was detected as far as Immokalee, just southwest of Lake Okeechobee.
This is also the northern limit for much of south Floridas more tropical fauna and flora


CO
CD
o
0
% TOTAL NET CATCH % TOTAL NET CATCH
% TOTAL CATCH
% TOTAL NET CATCH
% TOTAL NET CATCH
% TOTAL NET CATCH ~ % TOTAL NET CATCH
A TRAP 1 D) VALDOSTA TRAP 2


67
1986-1988
Fig. 3-12.--A migration profile summarizing Phoebis sennae migration across a west-east
transect at the latitude of Gainesville, Florida (29.65). The profile is derived from a
median estimate (net number of individuals/minute, adjusted for time of day, sample size
and season) at each site during 13 September-28 October 1986, 26 September-18 October
1987 and 18 September-12 October 1988. Estimates were made along nine segments of
a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen,
PLK=Palatka, HST=Hastings, CRB=Crescent Beach.


93
Valdosta have milder winter temperatures and the population build up during summer
months may be much larger than at more northern sites, resulting in more winter
survival. Cloudless sulphurs are commonly seen in Gainesville on warm days throughout
the winter, but do seem affected by freezing temperatures which would be more likely
in Valdosta. Following a particularly severe freeze in Gainesville on 23 December 1989,
this species was not seen again until 18 February the following year. This was despite
unusually high temperatures after the hard freeze.
Gulf Fritillarv
Lpngtiydinai p Pole counts of Agraulis vanillae net movement during 1986 along six north-south
sites from Gainesville to Immokalee, Florida are shown in Fig. 4-6 a,b,c,d,e,f. All net
movement at the sites was southward during the period sampled. Of the weeks sampled,
the peak statistically significant migration (chi-square, p < .05) for this species occurred
during week 6 in Gainesville at approximately .55 indiv/m/min. The only other site
with significant movement was the next site, Clermont, about 180 kms south of
Gainesville, with .42 indiv/m/min. There were no individuals of Agraulis vanillae
counted at either LaBelle or Immokalee during the sampling period.
The pole count summaries of gulf fritillary net movement at six sites from
Gainesville to Immokalee is shown in Fig. 4-7. Like the cloudless sulphur, all net
movement was southward for the gulf fritillary. The greatest numbers of individuals
travelled south through Gainesville at an average rate of nearly .4 indiv/m/min.
Numbers of individuals moving through the more southern sites of Clermont and Bartow


Fig. 4-17 a,b,c,d,e,f,g,h,i.--\Veekly percentage of total net trap catch of Precis coenia
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August-
26 November 1988. Data for Gainesville, Florida are from Walker (1991). Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05).
Shaded bars show nonsignificant (p >.05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right comer. The midpoint of
migration is designated by an asterisk, x=traps not operating.


119
peak numbers recorded just east of Gainesville. The greatest density for the gulf
fritillary migration was observed along the western sections of the state, with peak
numbers noted just west of Gainesville. Flight azimuth observations, drive counts and
pole counts made across north central Florida have all demonstrated that the numbers of
cloudless sulphurs and gulf fritillaries are very low along the Atlantic coast. This is
consistent with Walkers (pers. comm.) reports that these species gradually changed their
migratory tracks, from southeast to south, along an east-west transect from the central
to Atlantic coastal regions in south Georgia. Long tailed skippers were not included in
the migration profiles, but pole counts at sites along the same transect suggest that
migration is more common for this species along both the Atlantic and Gulf coasts, as
well as through Gainesville. The four species are alike in having a summer breeding
range that includes much of the eastern United States, and although the overwintering
ranges are not adequately described, they appear to be restricted to peninsular Florida
or to peninsular Florida and the adjacent coastal plain. The buckeye is the only one of
these species that has a large northward spring migration. Since the fall migration is
comparatively small (Walker, 1991), very few individuals of this species were seen
during these fall studies.
Using the density estimates from migration profiles and previous data from
permanent traps in Gainesville (Walker, 1991), the total number of cloudless sulphurs
and gulf fritillaries moving south each fall into central Florida was estimated at 42 and
115 million individuals, respectively. When Walkers (1991) previous estimates for the


9
generations that will also presumably join the migration south and perhaps colonize
northern areas in which they have had no experience, the following spring. As Baker
(1978) has suggested, the behavior and ecology of these species may demonstrate a
sequence in evolutionary development along the spectrum of the migratory habit and can
provide new insights into these behaviors.
Florida provides an ideal geographical location for the study of butterfly
migration. The peculiar configuration of the state results in a number of migratory
species from the southeastern United States being funneled through the peninsula
presumably to the southern latitudes where they join resident populations and overwinter.
Although there are reports of occasional butterflies some distance at sea (Baust s ai,
1981), the difficulty of long oceanic trips (Brower, 1962) most likely precludes the
majority of butterflies from flying across the Gulf of Mexico, and certainly, the Atlantic.
The peninsula of Florida, then, acts as a trap with a concentration of individuals
contained by extensive water masses on three sides. The butterfly migration through
Florida perhaps ranks among the worlds phenomenal insect movements in that it
involves a large number of unrelated species from a wide area within a short period of
time. Larsen (1988ab) has reported a similar mixed butterfly migration during spring
and fall in southern India. The movement involves more than twenty species, numbering
approximately four million spring migrants and as many as 100 million fall migrants.
The width of the migratory front reported by Larsen (1988ab) was, at most, only 12 km
in the spring. It may also be true that the majority of the Florida spring return migration
does not pass through Gainesville.


61
Profile estimates for the gulf fritillary migration during fall of 1987 are shown
in Fig. 3-9 a,b. All net movement was southward and the migratory peak was recorded
during week 7 (Fig. 3-9b, west) along the TRN-NWB segment at nearly 15 indiv/min.
This increase in numbers was seen one year and a day after the extraordinarily large
numbers of Phoebis sennae were noted along the Gulf coast in 1986 (Fig. 3-4g).
Numbers of individuals seen along the eastern transect remained low and the only
segments with significant movement southward were the central areas of GNV-HAW and
HAW-INT.
Profile estimates of gulf fritillary migration during fall of 1988 are shown in
Fig. 3-10 a,b,c,d. Each weekly sample is a round trip during the same day and most of
the patterns were similar between trips 1 and 2, except during week 3 of sampling (Fig.
3-10a). Although trip 1 of week 3 showed fairly uniform movement across the western
transect (all < 2 indiv/min), trip 2 peaked through TRN-NWB with > 10 indiv/min.
Other than the TRN-NWB segment on this day, STN-CRC had the next largest number
of individuals, during week 3 and week 5 (Fig. 3-10c, trips 1 and 2). The Atlantic coast
had relatively low numbers of migrants (< 2 indiv/min) during week 4 (Fig. 3-10b), but
in week 6 showed an increase with peaks at HAW-INT (trip 1) and INT-PLK (trip 2).
Yearly migration profiles of Agraulis vanillae movement during the fall of 1986,
1987 and 1988 are shown in Fig. 3-11 a,b,c. The profile for 1986 (Fig. 3-1 la) peaked
through the INT-PLK segment (7 indiv/min) with the next peak at TRN-NWB (4.8
indiv/min). During 1987 (Fig. 3-1 lb), the TRN-NWB transect segment also contained
the migration peak at 7.5 indiv/min. This rate was nearly three times the number


25
Materials and Methods
Flight Azimuths
During the fall migratory season (September-November) of 1985-1988, flight
azimuths for the four migrant species were recorded at a number of locations throughout
the Florida peninsula. Sites selected for observations were large, open areas, such as
high school football fields or large lawns. Hours for observation were peak flight
periods established by Walker (1985a) from 0730 to 1430 HRS LMT. Observation
periods were also limited to weather conditions that are favorable to migration and were
terminated if more than 50% of the sky was obscured by clouds, temperatures dropped
below 21 C or wind speed exceeded 3 m/sec. For each observation, these conditions
were noted as described in detail by Walker (1985a): civil time, species, appearance of
sun, percentage of blue sky visible, wind speed, wind direction and air temperature. The
butterfly passing closest to the observer was selected and a sighting of its flight path was
then made from the point where it had passed, to where it disappeared on the horizon,
noting the magnetic bearing with a Suunto KB 14 compass. Such observations continued
with the next individual spotted and so forth, also as described by Walker (1985a).
Magnetic bearings were adjusted for magnetic declination, resulting in true bearings and
civil time was converted to local mean time. Walker (1985a) found that mean migratory
directions at Gainesville remained approximately 141 regardless of species, season, time
of day or wind. For the purposes of this study, migration data for all dates at a locality
were grouped for each species. These data were analyzed using BUTTAZ.BAS, a


Fig. 3-4 a,b,c,d,e,f,g,h.Profile estimates of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65) from 13 September to 28 October
1986. Counts were taken on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC = Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


103
a)
VALDOSTA
C)
WEEKS
LEESBURG
WEEKS
b)
d)
GAINESVILLE
,x x
123466789 101112131416
WEEKS
LAKE ALFRED
12346878 9 10 1112131415
WEEKS
e)
12346678 9 10 1112131416
WEEKS
Fig. 4-10 a,b,c,d,e. Summary of percent total net trap catches of Agraulis vanillae at
five sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake
Alfred and Lake Placid, Florida) during 13 September-12 December 1987 and 28
August-26 November 1988. Solid bars represent net southward movement and clear bars
show net northward movement. Total net catch is shown in the upper right comer and
x=trap not operational. Data for the first 4 weeks and all of 1988 Gainesville are is
from Walker (1991).


18
June and July in south Florida (Dade or Monroe Cos.). Instead, the recently introduced,
nonmigratory species, II. dorantes, with very similar habits is common during this period
(Lenczewski, 1980).
Precis coenia
Commonly known as the buckeye, this butterfly is the only Florida migrant
which demonstrates relatively large scale spring northward migration (Walker, 1991).
The genus Precis is largely confined to the tropical regions of the world, only two
species occur in North America. There has been considerable confusion concerning the
systematics of this species, which can be quite variable in size and color. P. coenia is
found from southern Canada, west to California, Arizona and south through tropical
America and Cuba. Clark and Clark (1951) reported two seasonal forms in Virginia.
The spring/autumn form had an average forewing length of 27 mm in females and 24
mm in males. The summer form averaged a slightly greater wing length and darker
color, characterized by an irregular reddish band on the underside of the hind wings
which the spring/autumn form lacks. Clark and Clark (1951) noted that the wings in
nearly all of the spring/autumn specimens they collected in Virginia were damaged.
These worn wings may indicate long distance travel and this corresponds well to what
has been observed of migration periods for this species. Mather (1967) found light and
dark forms to be correlated with seasonal changes. He found the major seasonal shift
from dark to light occurs between March and April and from light to dark between
August and September. The shifts were found to coincide with a change in the mean


95
z
.6
1
.4
(/>
.2
<
D
O
0
>
o
.2
z
.4
h*
LU
Z
.6
GNV CLM BRT ARC LBL IMK
LOCATION
Fig. 4-7.-A summary of pole counts of Agraulis vanillae (average net number of
individuals/meter/minute, adjusted for time of day) through Gainesville, Clermont,
Bartow, Arcadia, Labelle and Immokalee, Florida during 11 September-8 November
1986. Numbers below the bars represent total number of visits to that site.


76
4. WILDWOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
6. BROOKSVILLE-HILL N DALE
7. BAYONET POINT-SAN ANTON
B. WESLEY CHAPEL-ZEPHRHILLS
>. LAKELAND-HAINES CITY
10. BARTOW-LAKE WALES
11. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIOGE-LAKE PLACID
d d SOUTH BIAS
NORTH BIAS
O O NO AS
GNVLAT (29.65)
NONE SEEN
7 MM =20 KM


Fig. 4-4 a,b,c,d,e,f,g,h,i.--Weekly percentage of seasonal total of net trap catch of
Phoebis sennae at Valdosta, Georgia, Leesburg, Lake Alfred and Lake Placid, Florida
during 28 August-26 November 1988. Data for Gainesville, Florida are from Walker
(1991). Solid bars represent statistically significant southward biased movement (chi-
square test, p < .05), shaded bars show nonsignificant (p > .05) southward movement.
Clear bars depict nonsignificant (p > .05) northward movement. The total net catch
(southbound minus northbound) for the period is shown in the upper right comer. The
midpoint of migration is designated by an asterisk, x=trap not operating. Total percent
migration in b) was reduced by 40% on the basis of data in a).


15
will opportunistically oviposit on plants according to availability. The preferred larval
hostplants in Florida for this species include various species of Cassia, particularly
obtusifolia (L.) (pers. obs.), and as Howe (1975) reported in Louisiana, the partridge
pea, Chamaecrista cinrea (L.). Two generations reportedly occur in the northern part
of the range and although Howe (1975) reports that breeding is continuous in the Gulf
region, including Florida, there are no records of such activity during the winter months,
at least in the latter area. During 1987, Calhoun e (1990) reported cloudless sulphur
individuals sighted in Wisconsin and New York. Numbers of individuals seen that
summer were unusually large and sightings extended the northern range significantly.
The species was common in Mississippi, Illinois, Kentucky, Indiana, Ohio and West
Virginia. He described temporary breeding colonies of Phoebis sennae in Ohio and West
Virginia, where apparently, due to the unusually warm winter of 1986-1987, individuals
overwintered and were available in early spring to utilize newly emerged hostplants. The
larval hostplants of this species usually senesce in late fall and are absent through the
winter months. Interestingly, these breeding populations in Ohio and West Virginia were
located along river banks, which Calhoun (1990) considered to be primary corridors of
dispersal.
Agraulis vanillae
This butterfly, also known as the gulf fritillary, is commonly found in
temporary, disturbed habitats such as roadsides and fields, usually near its larval
foodplant, during the late summer and fall months in Florida. The species is widespread,


20
and Acuba (Verbenaceae). The Scrophulariaceae appear to be the most important larval
hostplants in Florida.
Other Migrants
Pieris rapae (L.l. An introduction from England into Quebec in 1860, this well-
known migrant has spread rapidly throughout most of North America. Also known as
the cabbage white, it can be a serious agricultural pest on cruciferous crops. Although
very common in agricultural fields in north central Florida, this species is now rather
rare in south Florida (Lenczewski, 1980). Walker (1991) reported northward movement
in the spring in Florida, but it seems this species is also capable of overwintering as far
north as Canada (Scott, 1986).
Vanessa virginienis (Drurv). Although the American painted lady is a common
species in north Florida, this butterfly has been reported by Lenczewski (1980) as
infrequent in south Florida (Dade and Broward Cos). The range of this species extends
from coast to coast in the U.S., from southern Canada to Colombia, the Canary Islands
and Hawaii. It is rare in the Antilles and an occasional vagrant in Europe. This species
was not previously known to be migratory, at least to the extent of V. cardui. As found
for P. rapae in Gainesville, Florida, V. virginiensis showed northward movement in the
spring but individuals were seldom caught in the fall (Walker, 1991). Possibly their
movements are not related to temperature since Scott (1986) reports this species as also
overwintering in Canada.


106
Z .6
2
GNV CLM BRT ARC LBL ¡MK
LOCATION
Fig. 4-12. A summary of pole counts showing net movement of Urbanus proteus
(average net number of individuals/meter/minute, adjusted for time of day) through
Gainesville, Clermont, Bartow, Arcadia, Labelle and Immokalee, Florida during 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p < .05), x=no counts made. Numbers below the
bars represent total number of visits to that site.


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Thomas J. W
Professor of
Nematology
er, Chair
tomology and
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
>
ames E. Lloyd
Professor of Entomology and
Nematology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Professor of Entomology and
Nematology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Lincoln P. Brower
Professor of Zoology


BIOGRAPHICAL SKETCH
Barbara Lenczewski was bom in New Haven, Connecticut, in 1952. She
received a Bachelor of Science degree from the University of Connecticut, Storrs, in
1978. The following year, she collected mammals in the Paraguayan Chaco, as part of
a University of Connecticut research team. Upon returning to the U.S., Barbara was
employed as a fire ecology technician with the South Florida Research Center,
Everglades National Park. Her work documenting the butterflies of the region resulted
in the publication of Butterflies of Everglades National Park in 1980. That same year,
she began graduate studies in insect ecology and a teaching assistantship at Florida State
University. In 1985, she was awarded the Master of Science degree for her study of
Trachymvrmex septentrionalis. a northern leafcutting ant. That fall, she began doctoral
studies in entomology at the University of Florida, pursuing a long time interest in
butterfly migration. During 1985-1988, Barbara was employed as a research assistant
in the Department of Entomology and Nematology through the mole cricket biological
control project. From 1988-1992, her lepidopterological horizons were further
broadened to include moth mating behavior, and, she worked as a research associate at
the U.S.D.A., Insect Attractants, Behavior and Basic Biology Research Laboratory in
Gainesville, Florida until 1992. She received her Ph.D., in 1992, from the University
of Florida.
132


66
recorded at the next highest peak, through CRC-TRN (2.7 indiv/min). All southward
movement was statistically significant (chi-square, p < .05) except at PLK-HST and
HST-CRB during 1987, where numbers were small. During 1988, the migration peak
occurred along the Gulf coast through STN-CRC, but at lower numbers than previous
years (approx. 4 indiv/min). During all three years, the lowest numbers of individuals
moved southward along the east coast.
Numbers of Migrants
A migration profile summarizing the results of 1986, 1987 and 1988 profiles of
Phoebis sennae migration is shown in Fig. 3-12. All net movement was southward
throughout the fall sampling period (13 Sept-28 Oct). The transect segment with the
greatest numbers of cloudless sulphurs moving south was HAW-INT at 2.7 indiv/min,
more than double any other segment. The migration rate was fairly even across the other
transect segments, averaging about 1 indiv/min, except for PLK-HST (.34 indiv/min) and
HST-CRB (.18 indiv/min), where numbers declined approaching the Gulf coast.
The migration profile summarizing Agraulis vanillae movement for 1986, 1987
and 1988 is given in Fig. 3-13. Peak migratory movement for the gulf fritillary was
through the TRN-NWB segment at 4.8 indiv/min. The other peak was along the west
coast segment, STN-CRC at 3.9 indiv/min. In the case of the gulf fritillary, all three
segments with peak movement were along the west coast. Migration east of Newberry
was low in general, but the transect segments with the least movement were PLK-HST
and HST-CRB, at .81 and.84 indiv/min respectively.


30
and the southern boundary for many northern species. Historically, this is about the
point at which damage from a hard freeze is unlikely. It is important to consider the
number of visits to a site and the pattern of individual flight directions shown in Fig. 2-2,
rather than the mean vector alone. For example, although the mean vector at
Steinhatchee (Fig. 2-1) is approximately due east, from Fig. 2-2, it is evident that the
individuals had no consensus in migratory direction. The same is true at Alligator Point
during the only visit there. The eastward vector in Fig. 2-1 really is a result of south
and north flying individuals as shown in Fig. 2-2.
The mean flight direction patterns for the gulf fritillary throughout the state are
similar to those described for the cloudless sulphur and are shown in Fig. 2-3.
Significantly directed movement for the gulf fritillary continued south to Immokalee, but
fewer individuals were observed. Whereas Phoebis sennae is not common along the gulf
coast, Agraulis vanillae masses along the panhandle coast in large numbers (pers. obs.).
Also, unlike Phoebis sennae. the gulf fritillary showed no evidence of avoiding Lake
Okeechobee. Three individuals were observed flying southeast from Okeechobee and
others were apparent at the opposite shore in Belle Glade, heading in the same direction.
However, this species too avoided Crescent Beach, the most eastern site along the
Atlantic Coast.
Whereas some cloudless sulphurs do overwinter in north Florida and buckeyes
have also been reported to do so (Walker, 1978; Scott, 1986), this is not usual for the
gulf fritillary. The more highly directed flight of the gulf fritillary may be necessary to
ensure survival before freezing temperatures occur. The individual flight directions


85
had the greatest numbers moving south (1 indiv/m/min) during peak migration, well over
three times that of the next largest movement at Clermont (.25 indiv/m/min). At
Clermont (b), which is located approximately 136 kms south of Gainesville, the cloudless
sulphur did not achieve a statistically significant southward movement until five weeks
later, during week 7. The rest of the sites had no significantly directed movement,
except for LaBelle, where cloudless sulphurs also migrated southward (.2 indiv/m/min)
during week 7.
The summary of Phoebis sennae pole counts along the transect from Gainesville
south is shown in Fig. 4-2. Gainesville had the highest counts with an average of about
.6 indiv/m/min passing southward. Clermont and LaBelle, the only other sites with
significant southward movement, had a rate of about .1 indiv/m/min.
Flight trap catches
Weekly percentages of total net trap catch of Phoebis sennae during the fall of
1987 at five sites along a north-south transect through Gainesville are shown in Fig. 4-3
a,b,c,d,e. All statistically significant net movement was southward for this species,
falling between weeks 4 and 12. The early weeks were not sampled completely at any
of the sites, and the week in which half of the total net catch was achieved is probably
not a reliable estimate of mid-migration point. The Gainesville data (b) were corrected
for this by using information taken from Walkers (1991) permanent traps near the same
location, shown in Fig. 4-4c. In the case of the cloudless sulphur, the percent total net
catches for weeks 5-13 were reduced by 33% (Walkers trap catch during the first 4
weeks). The appropriate adjustment was applied in a similar manner for the other three


8
migrant species seem more common in south Florida during this time (Lenczewski,
1980), there has only been one report of Agraulis vanillae and Urbanus proteus migration
in Melbourne (Williams, 1958). In addition, there are no data reporting any large
aggregations of these species in south Florida, which is lepidopterologically very well
explored, especially during the fall and winter months (Lenczewski, 1980). Perhaps, as
suggested by Walker (1985a), the densities become insignificant when the migrants settle
into the large areas that are available.
The only spring migrant trapped at Gainesville in large numbers has been Precis
coenia. the buckeye. Although there must be some method of northern recolonization,
there is only minimal direct evidence of spring return migrations by the long tailed
skipper, Urbanus proteus. Spring migrations of Phoebis sennae and Agraulis vanillae are
also small and little noticed except for significant flight trap catches (Walker, 1991). The
total northern movement during spring of the four species has been estimated using
preliminary data from this study as eight million individuals (Walker, 1991).
As stated previously, of the insect migrants, selection for precision in destination
is perhaps most evident in the monarch butterfly (Brower, 1977), where large scale
movements to several precisely located overwintering sites are coordinated for millions
of individuals. The monarch completes the southward journey in reproductive diapause
and the same individuals return, at least part of the way, during spring to repopulate the
northern range (Brower, 1962). For the migrant butterflies observed moving through
north Florida, it is clearly evident that the butterflies are mating, ovipositing and stopping
to feed at nectar sources enroute. This reproductive activity must result in subsequent


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 EIBUVQCHO_WFD7AX INGEST_TIME 2012-07-25T23:13:22Z PACKAGE AA00011452_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


Fig. 3-14 a,b,c,d,e.f.Drive counts of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day) across various west-east transects south of
Gainesville, Florida during fall 1987. Transects with a southward or northward net
migration of statistical significance (chi-square test, p < .05), are shown with half filled
circles weighted in the appropriate direction. Net movement, either north or south, that
is not statistically significant (p < .05), is depicted with open circles. Lines with no
circles represent transects where there were no individuals seen. Number of individuals
and percentage flying south are given in parentheses after each transect.


70
calculations were based on an assumption that the average rate of migration across the
transect was no less than 60% of the rate at the longitude of his trapping site in
Gainesville. In actuality, for Phoebis sennae. his site fell within the NWB-GNV segment
of the transect which had a rate of 1.03 indiv/min, identical to the transect average.
Walker also used a conservative transect length of 170 km, whereas the actual transect
length used in this study is 190 km. Walkers (1991) Gainesville trap yields of 800
individuals/6 m for Phoebis sennae were applied to these data resulting in an estimated
influx into central Florida of 25.3 million individuals. With further adjustment for
Walkers (1985b) 60% trap efficiency, the size of the cloudless sulphur fall migration
into central Florida was estimated as 42.2 million individuals, nearly double the figure
given by Walker (1991).
These calculations were repeated using the migration profile information for
Agraulis vanillae. also shown in Table 3.1. The average migration along the latitude of
Gainesville was 2.10 indiv/min, as compared to the NWB-GNV segment, which had a
rate of 1.43 indiv/min, resulting in a profile correction factor of 1.47. After applying
Walkers trap catch of 868 individuals/6m and an efficiency of 35%, the final results
were 115 million gulf fritillaries, more than double Walkers (1991) estimate of 41.3
million. The total number of these two species of butterflies migrating south through
north central Florida in the fall was estimated at 157 million individuals annually. This
number does not include the buckeyes or long tailed skippers, which were estimated by
Walker as numbering 3.7 and 14.6 million, respectively. Since profiles were not
established for these species from drive counts, Walkers estimates cannot be evaluated.


a) WEST b) east
KILOMETERS KILOMETERS
WEST EAST
Fig. 3-6 a,b,c,d.--Profile estimates of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day,
sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 18 September to 12
October 1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB=Newberry, GNV =Gainesville, HAW = Hawthome, ITL=Interlachen, PLK=Palatka, HST=Hastings,
CRB=Crescent Beach. x=unsampled sites.


12346678 8 101112131416
WEEKS
1 2 3 4 6 8 7 8 8 10 1112131416
WEEKS
C)
LEESBURG
WEEKS
d)
LAKE ALFRED
WEEKS
e)
LAKE PLACID
WEEKS
Fig. 4-5 a,b,c,d,e.--Summary of percent total net trap catches of Phoebis sennae at five
sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred
and Lake Placid, Florida) during 30 August-12 December 1987 and 28 August-26
November 1988. Solid bars represent net southward movement and clear bars show net
northward movement. Total net catch (southbound minus northbound) is shown in the
upper right comer and x=unsampled weeks. Data for the first 4 weeks of 1987 and all
of 1988 Gainesville are from Walker (1991).


21
Eurema lisa (Boisduval and Le Conte). Scott (1986) reports this species as
overwintering in the southeastern U.S. and extending its range north to Canada in
summer. At Gainesville, Walker (1991) trapped £. lisa only in the fall when it was
moving south. This species is a well-known migrant and has been reported on occasions
to fly over the Caribbean, the Atlantic and the Gulf of Mexico in huge swarms (Howe,
1975; Klots, 1951; Lenczewski, 1980).
Eurema daira (Godart). This species overwinters in the southeastern Coastal
Plain and spreads north a few hundred kilometers in summer (Scott, 1986). Walker
(1991) found no significant movement south in the fall through Gainesville except in the
fall of 1985.
Eurema nicippe (Cramer). This species sometimes moves in a seemingly
inappropriate direction, northward during fall flights (Walker, 1991) and reportedly
overwinters throughout the southeastern U.S. (Howe, 1975). However, it has not been
reported in December, January or March in south Florida Dade or Monroe Cos.,
although it is very common there in the fall (Lenczewski, 1980).
Danaus plexippus (L.). The monarch is undoubtedly the best studied migrant
butterfly. Large numbers of monarchs aggregate along the Florida panhandle coast each
fall. Most of this eastern population flies west along the Florida coast and down through
Texas to overwintering sites in Mexico (Brower, 1977). The migrating monarchs
generally fly at altitudes (Gibo, 1986) where flight traps are of little use, although
Walker (1991) did trap 15 in ten years and reported a significant southward bias in the
fall. Those individuals that enter Florida become physiologically "trapped" when warm


22
temperatures end their reproductive diapause, and subsequently, migratory tendency
(Brower, 1961; T. Van Hook, pers. comm.). There are reports of several small
overwintering colonies along the west coast of Florida (Brower, 1961; and pers. com.)
but the breeding status of these is unknown. About 20% of female monarchs along the
Florida Gulf coast during fall migrations are mated and will oviposit on available
milkweeds in Gainesville also during that time (T. Van Hook, pers. comm.). This
species is also known to breed in south Florida during the winter months, but the
subsequent migratory tendencies of the offspring are not known (Lenczewski, 1980).
Panoquina ocola (Edwards!. This small skipper is often found concentrated in
large numbers along the Gulf coast with other migrants in the fall (pers. obs.) and has
been reported by Walker (1978) as migrating through Gainesville. A mass movement
of these skippers in Louisiana was described by Penn (1955). Found through most of
the southeast, from Virginia to Florida, west to Texas, Kentucky and Arkansas, it has
also been reported as far north as New Jersey (Howe, 1975). There are no records of
this species in south Florida (Dade and Monroe Cos.) in January (Lenczewski, 1980).
Lerema accius CSmithL The range of this skipper is from New England to
Florida, west to Illinois, Arkansas, Texas and south to northern South America. Howe
(1975) reports it as scarce northward but common in the southern states, with records
from February to November. Walker (1978) reported some fall southward movement
for this species and it is present during every month of the year in south Florida
(Lenczewski, 1980).


48
STN CRC TRN NWB GNV HAW ITL PLK HST CRB
0.6
0.5
z
5
2 0.4
>
Q
2 0.3
**
h
LU
z 0.2
-a
O
<
0.1
0
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-3.--Summary of pole counts showing Urbanus proteus migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p <.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN = Trenton, NWB = Newberry, GNV = Gainesville, HAW = Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.


TABLE OF CONTENTS
ACKNOWLEDGEMENTS ni
ABSTRACT vii
CHAPTER 1. INTRODUCTION 1
Literature Review 1
General Migration 1
Florida Butterfly Migration 5
Research Plan 10
Biology of the Butterflies 11
Phwfris samas n
Agraulis vanillae 15
Urbanus Proteus 17
Precis wenia 18
Other Migrants 20
CHAPTER 2. FLIGHT DIRECTION 23
Introduction 23
Materials and Methods 25
Flight Azimuths 25
Mark-Release-Recapture 26
Results and Discussion 26
Mark-Release-Recapture 37
CHAPTER 3. NUMBERS OF MIGRANTS 38
Introduction 38
Materials and Methods 39
Latitudinal Pole Counts 39
Migration Profiles 41
Results and Discussion 44
Latitudinal Pole Counts 44
v


43
individuals observed along each transect segment was converted to a proportion of the
net number observed along the base segment = DnT/D,.
Another important influence on the amount of migratory movement observed is
a seasonal factor (Walker, 1991). Since the drive counts were made over a period of
several months, numbers had to be adjusted for season. In addition, partial profiles from
different drive counts had to be combined to estimate the entire trans-state profile.
Walker (1991) has collected phenological information on the migration pattern in
Gainesville, Florida for more than ten years. His two permanent traps sample
continuously throughout the season and have been shown to reflect migration patterns
reliably. To eliminate the effects of season and to unify the east and west transects,
Walkers trap catches were used to quantify the "strength" of migration for each day that
drive counts were made. The net numbers of gulf fritillaries and cloudless sulphurs
captured in Walkers two traps were summed for the period during which drive counts
were made that year and a mean net number was calculated. The daily net number
caught in his traps was then divided by the mean net number to produce a factor that
would adjust to the mean net number. The reciprocal of this factor was called the
migration index, I, for that day. If the migration was particularly strong, the migration
index was correspondingly low. The adjusted data from drive counts was then multiplied
by the migration index appropriate to the day that sampling occurred. In this way, the
high numbers observed during the peak migratory season were reduced and the more
sparse, late or early migration was augmented, in effect eliminating the seasonal


102
migration waves could represent generational movement. There was no evidence of
significant movement through Lake Placid (Fig. 4-9i).
The summary of percent total net trap catches of Agraulis vanillae individuals
along the north-south transect during fall 1987 and 1988 is shown in Fig. 4-10 a,b,c,d,e.
All statistically significant (chi-square test, p < .05) net movement was southward.
There was a time span of three weeks in the migration midpoints from Valdosta, in south
Georgia, to the south central Florida site of Lake Alfred. This lag fits more closely the
hypothesis of synchronous initiation of movement, or travel by the same individuals,
rather than successive generations. Migration midpoints occurred in Valdosta and
Gainesville during week 5, Leesburg during week 6 and Lake Alfred during week 8 and
the migration lasted at least 9 weeks at all sites. Significant net movement ended at
Valdosta by week 12 and at Leesburg by week 13. Movement south at Lake Alfred and
Gainesville continued at least through weeks 12 and 13, respectively. There was no
evidence of significant net movement of gulf fritillaries through Lake Placid. Although
only seven individuals were caught throughout the two years of trapping, all were headed
south. Over the two year period, this would constitute a significant southward
movement.
Long Tailed Skipper
Longitudinal pole counts
The pole counts of net movement Urban us proteus net movement along a north-
south transect from Gainesville to Immokalee, Florida during 1986 are shown in Fig. 4-
11 a,b,c,d,e,f. The southward movement for this species at Gainesville was similar to


11
information was used, together with Walkers (1991) data from Gainesville, to estimate
the size of the annual fall migration of these two species moving through north central
Florida. The phenology of the migration was studied by trapping the four principal
migrant species using directional flight traps at five stations from south Georgia to south
central Florida. Counts of north and south flying Phoebis sennae and Agraulis vanillae
also were made on other west-east transects along the Florida peninsula.
The individual chapters have been written as drafts of separate publications.
This has unavoidably resulted in some repetition. The numerical assignment of weeks
for calendar dates during which sampling was done with all methods is given in Table
1-1. The location of all sites used for pole counts and flight trap stations are shown in
Fig. 1-1.
Biology of the Butterflies
Phoebis sennae
Commonly known as the cloudless sulphur, this is a fast flying butterfly of
temporary, disturbed habitats. It is found in the eastern and southern United States,
rarely straying to Canada, through the West Indies and south to Argentina (Lenczewski,
1980). In both the temperate and tropical zones, the subspecies of sennae exhibit strong
migratory habits (Howe, 1975) and C.B. Williams (1938) nicknamed this "The
Travelling Butterfly." The subspecies eubule occurs commonly in the southeastern
United States, straying as far north as Canada (Klots, 1951) during the summer and fall


131
Walker, T.J. and A.J. Riordan, 1981. Butterfly migration: are synoptic-scale wind
systems important? Ecol. Entomol. 6:433-440.
Walker, T.J. and S.A. Wineriter, 1981. Marking techniques for recognizing individual
insects. Fla. Entomol. 64:18-29
Williams, C.B. 1930. The Migration of Butterflies. Oliver and Boyd, Edinburgh. 473
pp.
Williams, C.B. 1937. Recent progress in the study of some North American migrant
butterflies. Ann. Entomol. Soc. Am. 31:211-230.
Williams, C.B. 1958. Insect Migration. Collins, London. 235 pp.
Zar, J.H. 1984. Biostatistical Analysis, 2nd ed., Prentice-Hall, New York. 718 pp.


88
species. The estimate for mid-migration was either week 5 or 6 at all sites. The close
coordination of mid-migration seems to indicate that the same wave of individuals passed
through all five areas. The most southern site, Lake Placid, had no net movement, and
even though several individuals were seen and two were trapped, there was no indication
of migration. In fact, the numbers of cloudless sulphurs were surprisingly low at all
other sites, as compared to Gainesville. The reason for low numbers at Valdosta at least
could be that since the first three weeks were not sampled, most of the migrants may
have passed southward earlier.
The percentage of total net catch of Phoebis sennae captured at the five sites in
1988 are shown in Fig. 4-4 a,b,c,d,e,f,g,h,i. Gainesville data is taken from Walker
(1991). All statistically significant movement was southward biased. At Valdosta (Fig.
4-4a), for the cloudless sulphur, catches in traps 1 and 2 were not well coordinated.
Trap 2 went out of commission after week 6 and the percentage of total catch was
adjusted in relation to the first six weeks of trap 1 (i.e. decreased by 40%). During
1988, traps were in place by week 1, at which time there was no evidence of migration.
The mid-migration point for trap 1 was during week 6, and the only other significant
southward movement was seen during week 7. Southward migration occurred during the
second week for trap 2, but sample size was small and the trap may already have been
damaged. Walkers (1991) data from the two permanent traps located in Gainesville
(Fig. 4-4 c,d) both show mid-migration occurred during week 7. The two traps at
Leesburg and the average of the two traps at Lake Alfred all suggest a midpoint during
week 8 at those sites. The only significant southward movement seen from catches at


/ : ^ 6 í/
BUTTERFLY MIGRATION
THROUGH THE FLORIDA PENINSULA
By
BARBARA LENCZEWSKI
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
1992

This work is lovingly dedicated to my parents, Lucjan and Stanislawa
Lenczewski, who have waited half their lives to see it and to my husband, Joseph
lowers, for believing that he would.

ACKNOWLEDGEMENTS
A number of people were helpful during this research. Foremost, I would like
to thank my major professor, Dr. Thomas J. Walker, for supplying the materials and
workshop used in construction of the portable traps and for providing expertise and labor
during their construction. The funds for travel and other expenses were arranged by Dr.
Walker and he provided a great deal of enthusiasm and encouragement for this project
throughout many years. I am also indebted to Dr. J. J. H. Frank, who, through the
Mole Cricket Biocontrol Project, provided the opportunity for a graduate assistantship
and facilitated the loan of travel vehicles. Dr. Lincoln P. Brower provided many
opportunities for "butterfly" socializing, and the use of his property for mark-release-
recapture studies, during which he kindly made many valuable observations. Dr. Brower
and Dr. Frank Slansky generously allowed me the use of their laboratories and
equipment. I also thank the other members of my committee, Drs. J. E. Lloyd and J. L.
Nation, for their helpful suggestions and a critical review of the manuscript.
The following people made the trapping study feasible by providing locations
for, and assistance with, flight traps. I am particularly grateful to Dr. J. Whitesell,
Valdosta State College; Dr. G. W. Ellstrom and his assistant, Ms. Annette Chandler,
IFAS/AREC Leesburg; Dr. Mohammed Abou-Setta and Dr. H. N. Nigg, IFAS/CREC;
and Dr. Mark Deyrup, Archbold Biological Station. Special thanks go to Dr. Whitesell
m

for field assistance with traps, Dr. Nigg for a tour of the Green Swamp and Dr. Deyrup
for help during hurricanes and otherwise.
My first interest in butterflies and their migration through the Florida Peninsula
was inspired by the late Dr. Dennis Leston, in whose memory this work is completed.
His enthusiasm for nature and love for insects will remain with me always. I am greatly
indebted to my parents, Lucjan and Stanislawa Lenczewski, who have encouraged my
education. My sister, Eva Lenczewski, despite her dislike of insects, has sometimes
even helped with this perversity of mine and she deserves recognition for her courage.
Finally, and most importantly, I thank my husband, Joseph Jowers, for a great deal of
assistance and companionship in the field. He and my son, Justin, held down the home
front to really make my work possible.
IV

TABLE OF CONTENTS
ACKNOWLEDGEMENTS ni
ABSTRACT vii
CHAPTER 1. INTRODUCTION 1
Literature Review 1
General Migration 1
Florida Butterfly Migration 5
Research Plan 10
Biology of the Butterflies 11
Phwfris samas n
Agraulis vanillae 15
Urbanus Proteus 17
Precis wenia 18
Other Migrants 20
CHAPTER 2. FLIGHT DIRECTION 23
Introduction 23
Materials and Methods 25
Flight Azimuths 25
Mark-Release-Recapture 26
Results and Discussion 26
Mark-Release-Recapture 37
CHAPTER 3. NUMBERS OF MIGRANTS 38
Introduction 38
Materials and Methods 39
Latitudinal Pole Counts 39
Migration Profiles 41
Results and Discussion 44
Latitudinal Pole Counts 44
v

Cloudless sulphur 44
Gulf fritillary 46
Long tailed skipper 46
Migration Profiles 49
Cloudless sulphur 49
Gulf fritillary 58
Numbers of Migrants 66
CHAPTER 4. PHENOLOGY OF MOVEMENT 77
Introduction 77
Materials and Methods 80
Longitudinal Pole Counts 80
Portable Flight Traps 81
Results and Discussion 83
Cloudless Sulphur 83
Longitudinal pole counts 83
Flight trap catches 85
Gulf Fritillary 93
Longtiudinal pole counts 93
Flight trap catches 96
Long Tailed Skipper 102
Longitudinal pole counts 102
Flight trap catches 105
Buckeye 112
Flight trap catches 112
CHAPTER 5. SUMMARY AND DISCUSSION 118
REFERENCES 124
BIOGRAPHICAL SKETCH 132
vi

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BUTTERFLY MIGRATION THROUGH THE FLORIDA PENINSULA
By
Barbara Lenczewski
December 1992
Chairman: Dr. Thomas J. Walker
Major Department: Entomology and Nematology
At least eight species of butterflies migrate southward during fall of each year
through the Florida peninsula. The four major migrants, representing four families of
Lepidoptera, are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae
(L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long tailed
skipper; and Precis coenia (Hübner) (Nymphalidae), the buckeye. Drive counts of the
net southward movement of the cloudless sulphur and gulf fritillary were made across
north central Florida at the latitude of Gainesville (29.65°) during fall 1986-1988. This
migration profile, or cross section through the migratory stream, was used to estimate
the density of individuals moving south into central Florida annually. The highest
density of the cloudless sulphur was through the central parts of the state and that of the
gulf fritillary, along the western sections. Peak numbers were recorded just east and just
vii

west of Gainesville, respectively. Both species avoided the Atlantic coast. Using
previous data from Gainesville flight trap catches in conjunction with density estimates
from this study, the total number of cloudless sulphurs and gulf fritillaries moving south
each fall into central Florida was estimated at 42 and 115 million individuals,
respectively. When previous estimates for the long tailed skipper and buckeye were
included, the total number of butterflies migrating south through the Florida peninsula
was estimated to be 175 million individuals annually. Observations of flight azimuths
for three species (excluding the buckeye) were made at various locations south of
Gainesville during fall 1986. Most significant mean flight directions were consistent with
the 141° previously described for Gainesville and the limit of significant southward
movement for that year was observed just south of Lake Ockeechobee for all three
species. The phenology of the migration was studied by trapping the four major migrant
species using directional flight traps at five stations from south Georgia to south central
Florida during fall 1987-1988. All significant net movement was southward, extending
to the Lake Alfred site, about 192 km south of Gainesville. There was no evidence from
trap catches or drive counts at the Lake Placid latitude (27.45°) that large numbers of
migrants move into extreme south Florida to overwinter.
Vlll

CHAPTER 1
INTRODUCTION
Literature Review
General Migration
Migration has been recognized in recent years as an important element in the
dynamics of insect populations (Solbreck, 1985). Long distance movements of butterflies
(Williams, 1930), locusts (Johnson, 1969), aphids (Kennedy, 1961), moths (Stinner si
al, 1983), dragonflies (Dumont and Hinnekint, 1973) and insects of many other groups
have been recorded by observers across practically every continent (Williams, 1958), at
high altitudes (Rainey, 1951), and many miles out to sea (Baust el al, 1981). Although
some insect flights seem to function as random dispersal by forces beyond the
individual’s control (Johnson, 1969), closer behavioral studies reveal that active initiation
of flight occurs in many species and correlates with favorable wind conditions and
appropriate seasons (Baker, 1978; Taylor and Reling, 1986). In other words, the
migrants will "choose" the appropriate winds, during the appropriate times of year, by
which they may be carried.
It was perhaps due to his interest in butterflies that Williams (1930) stressed the
concept of individual control over flight direction in his early definition of migration.
This individual control over flight direction is particularly evident in butterfly migrations.
1

2
These highly visible insects travel within the boundary layer, the layer of air near the
ground where wind velocity is less than the insect’s air speed, and flight direction can
be controlled by the individual. The thickness of this boundary layer is variable,
determined by wind velocity and the air speed maintained by the individual (Taylor,
1958; Pedgley, 1982). This determined choice of flight track is perhaps easiest to
observe in butterflies and demonstrates that migration can be more under the individual
control of these, and perhaps other, insects than previously suspected.
There has been considerable controversy in theoretically extricating dispersion
from migration. Baker (1978) has suggested that there is a continuum in the expression
of the migratory habit, ranging from random movement to the evolution of highly
specific destinations with correspondingly adaptive physiological changes. In a recent
summary, Dan than arayana (1986, p. 1) claimed a consensus has emerged on the
terminology used to describe insect movements. His definition of migration will be
accepted for the purposes of this study:
Non-migratory movements involve travel within the habitat associated with such
activities as feeding, mating and opposition ... In contrast, migratory
movements take insects beyond the habitat for the purposes of colonizing new
habitats, re-colonizing old ones, aestivation or hibernation ....
The defining criterion is travel outside what is, or had previously served, as the
sustaining "habitat" to more favorable conditions. The behavior and physiology involved
in getting to these new habitats fall somewhere on a continuum in the evolution of the
migratory habit. Some insect migrants, such as the monarch butterfly, may be
reproductively inactive during these movements and maintain long distance flights on
stored body fats, but in others, mating and feeding occur in conjunction with the

3
migratory movement. Although not necessarily so, a cyclic, or seasonal pattern is often
seen in these migratory movements. The same individual may return, at least part way,
to re-colonize old habitats, or, the return may consist of a step-wise expansion of new
generations into the former breeding range during favorable conditions (Walker, 1985a).
Baker (1969) has speculated on the evolution of migration and the maintenance
of migratory behaviors in populations. The factors that initiate migration are varied and
have been found to be finely attuned by genetic programming (Lamb and McKay, 1983)
to changes in habitat quality over time and space (Southwood, 1962) and other
environmental cues (Dingle, 1972). The influence of genetic factors on migratory
tendency in members of a population has been investigated empirically by Dingle gt a!
(1977) and Istock (1978), and theoretically by Roff (1975). It is known that individual
dispersal tendencies can be highly variable within populations (Baker, 1978) and these
characters have been influenced by experimental selection to some degree (Dingle, 1968;
Rankin, 1978). There are a number of possible life history strategies available to an
insect and how natural selection operates on these genetic factors in the evolution of the
migratory habit is still unclear.
How migrants find their destinations, that is, their navigational techniques,
(Schmidt-Koenig, 1975), also remain a mystery. Monarch butterflies are highly specific
in their annual migration to overwintering sites in Mexico (Calvert and Brower, 1980).
Walker (1985a) suggested that cloudless sulphurs from the southeast navigate toward the
Florida peninsula.

4
Migrating butterflies can maintain their compass direction and therefore must
have some orientation mechanism. It is possible that a time-compensated sun compass
is used (Walker, 1985a; Baker, 1978), but although Baker (1968a) has shown possible
azimuth orientation to the sun’s position, by Pieris rapae. this remains to be proven
(Able, 1980). That fall migrants can maintain their migratory direction on overcast days,
although not many fly then, weakens, but does not eliminate, this theory (Verheijen,
1978). Another possibility is that they have a magnetic compass. Evidence of
geomagnetic orientation has been found in the underwing moth, Noctua prónuba L.
(Baker and Mather, 1982), and is well known in honey bees (Lindauer and Martin,
1972). Magnetic particles have been detected in the monarch butterfly, although their
function has not been demonstrated (MacFadden and Jones, 1985). Jungreis (1987)
tested two of the Florida migrant butterflies, Phoebis sennae (L.) and Agraulis vanillae
(L.), and found no evidence of magnetic particles.
Migration requires a high degree of coordination among behavior, morphology
and metabolism to function as an adaptive life history strategy. Migrant insects must
undergo physiological changes that can sometimes be used to identify their migratory
tendency. One metabolic preparation for sustained flight that has been studied is energy
storage. A primary source of energy is fat, or lipids (Beenakkers, 1965), and increased
levels have been found in migrating locusts and monarch butterflies (Johnson, 1974).
It is known that monarchs build up these lipid stores prior to fall migration and deplete
them during long flights to Mexico (Beall, 1948; Brower, 1977). Whether they
replenish, or slow, the depletion of these stores by nectar feeding enroute has still not

5
been determined (Meier and Fivizzani, 1980). Beall (1948) reported that monarchs
captured in Ontario, before flight, were heavier than those caught after flight in
Louisiana. Fat stores are conserved carefully by thermoregulation through the over¬
wintering period and must remain sufficient to at least fuel the return journey to new
feeding or breeding grounds (Masters, 1988).
There has been very little experimental investigation of migration in the
laboratory. Most studies in this area have been done with insects in which flight activity
and/or orientation can be easily measured, and usually in tethered flight (Dingle, 1985).
So far, researchers have been unable to measure a butterfly’s level of flight activity and
preferred direction in the laboratory. Flight traps (Walker, 1985b; Walker and
Lenczewski, 1989), that exploit the behavioral characteristic of migrants to go up and
over obstacles, rather than around, can segregate these individuals in the field, but the
problems in the laboratory remain.
Florida Butterfly Migration
In a long term study using permanent flight traps in Gainesville, Walker (1991)
has identified at least eight species of butterflies that migrate southward each fall through
the Florida peninsula. The four principal migrants represent four families of
Lepidoptera: Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae (L.)
(Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long-tailed
skipper; and Precis coenia (Hübner) (Nymphalidae), the buckeye. These species
comprised 85% of Walker’s (1991) total trap catch and they also exhibited bidirectional

6
seasonal movement -- i.e. north in the spring and south in the fall. The total size of the
fall migration of these species has been estimated using preliminary data from this study
by Walker (1991) at about 80 million individuals. This is comparable to the estimates for
monarchs at Mexican overwintering sites (Calvert ei al, 1979). The most significant and
noticeable movements through the Florida peninsula are exhibited by the cloudless
sulphur and the gulf fritillary, which have been estimated by Walker (1991) to comprise
nearly 80% of the fall migration, or 64 million individuals annually.
A number of other species also show less directed or irregular migratory
movements (Walker, 1991). These include four pierids, Pieris rapae (L.), the cabbage
white; Eurema daira (Godart), the dainty sulphur; Eurema lisa (Boisduval & LeConte),
the little sulphur; Eurema nicippe (Cramer), the sleepy orange, as well as a nymphalid,
Vanessa virginiensis (Drury), the painted lady. Two hesperiid skippers are known to
migrate southward in the fall through north Florida, although there is very little known
of their movements since they are too small to be retained by Walker’s (1991) traps.
They are Lerema accius (Smith), the clouded skipper and Panoquina ocola (Edwards),
the ocola skipper. Walker (1991) also reported some individuals of the monarch
butterfly, Danaus plexippus (L.), as passing through Gainesville. The majority of the
monarch migration, however, is southwest toward Texas and Mexico (Brower, 1977).
The recurring spectacle of huge butterfly migrations along the Gulf coast
(Urquhart and Urquhart, 1976) and through north central Florida (Walker, 1978, 1980,
1985ab, 1991) is locally well known, but is inadequately represented in the literature.
The Florida Gulf coast is a renowned stopover along the major North American

7
migratory route for birds (National Geographic, 1979). It seems to function this way,
in some respects, for migrant butterflies also. Large numbers of monarchs, gulf
fritillaries and long tailed skippers are often seen feeding at Baccharis halimifolia along
the Gulf coast, particularly in the St. Marks or Appalachicola area (pers. obs.) during
the fall. Cloudless sulphurs are also present, but not in such great numbers as the other
species and there are some indications (Walker, pers. comm.) that they avoid coastlines.
Despite the intriguing presence of these large aggregations of migrant butterflies which
are obvious to the most casual observer, the subsequent movements of these individuals,
except for the monarch, remain largely unknown. These species often exhibit erratic
flight near the coastline, thus making it very difficult to find a preferred pattern. There
are reports of movements west, following the coastline (Urquhart and Urquhart, 1976),
but I have seen migration in both directions along the coast and also south across the
water, as well as north back to land. The migrants seen massed along the Gulf are
spectacular, and understandably, have attracted the most attention. Away from the Gulf
coast, most north Florida residents realize that the numbers of these species increase in
the fall, but are generally unaware of their persistent southward flights. Once the
butterflies pass into the Florida peninsula, we know virtually nothing of the migration
beyond Gainesville (Walker, 1991).
Most migrants pass through north peninsular Florida during September through
November and it has been presumed they join resident populations in subtropical south
Florida. Although the migratory flights are so conspicuous through north Florida during
fall, this does not seem to be the case in the southern parts of the state. Although the

8
migrant species seem more common in south Florida during this time (Lenczewski,
1980), there has only been one report of Agraulis vanillae and Urbanus proteus migration
in Melbourne (Williams, 1958). In addition, there are no data reporting any large
aggregations of these species in south Florida, which is lepidopterologically very well
explored, especially during the fall and winter months (Lenczewski, 1980). Perhaps, as
suggested by Walker (1985a), the densities become insignificant when the migrants settle
into the large areas that are available.
The only spring migrant trapped at Gainesville in large numbers has been Precis
coenia. the buckeye. Although there must be some method of northern recolonization,
there is only minimal direct evidence of spring return migrations by the long tailed
skipper, Urbanus proteus. Spring migrations of Phoebis sennae and Agraulis vanillae are
also small and little noticed except for significant flight trap catches (Walker, 1991). The
total northern movement during spring of the four species has been estimated using
preliminary data from this study as eight million individuals (Walker, 1991).
As stated previously, of the insect migrants, selection for precision in destination
is perhaps most evident in the monarch butterfly (Brower, 1977), where large scale
movements to several precisely located overwintering sites are coordinated for millions
of individuals. The monarch completes the southward journey in reproductive diapause
and the same individuals return, at least part of the way, during spring to repopulate the
northern range (Brower, 1962). For the migrant butterflies observed moving through
north Florida, it is clearly evident that the butterflies are mating, ovipositing and stopping
to feed at nectar sources enroute. This reproductive activity must result in subsequent

9
generations that will also presumably join the migration south and perhaps colonize
northern areas in which they have had no experience, the following spring. As Baker
(1978) has suggested, the behavior and ecology of these species may demonstrate a
sequence in evolutionary development along the spectrum of the migratory habit and can
provide new insights into these behaviors.
Florida provides an ideal geographical location for the study of butterfly
migration. The peculiar configuration of the state results in a number of migratory
species from the southeastern United States being funneled through the peninsula
presumably to the southern latitudes where they join resident populations and overwinter.
Although there are reports of occasional butterflies some distance at sea (Baust sí ai,
1981), the difficulty of long oceanic trips (Brower, 1962) most likely precludes the
majority of butterflies from flying across the Gulf of Mexico, and certainly, the Atlantic.
The peninsula of Florida, then, acts as a trap with a concentration of individuals
contained by extensive water masses on three sides. The butterfly migration through
Florida perhaps ranks among the world’s phenomenal insect movements in that it
involves a large number of unrelated species from a wide area within a short period of
time. Larsen (1988ab) has reported a similar mixed butterfly migration during spring
and fall in southern India. The movement involves more than twenty species, numbering
approximately four million spring migrants and as many as 100 million fall migrants.
The width of the migratory front reported by Larsen (1988ab) was, at most, only 12 km
in the spring. It may also be true that the majority of the Florida spring return migration
does not pass through Gainesville.

10
At the very least, understanding why, how and where these butterflies are going
presents a fascinating puzzle. As stated previously, outside of Gainesville, Florida,
virtually nothing is known of the phenology or extent of the migratory movements. The
goal of this study was therefore to contribute to general knowledge of the migration,
primarily of Phoebis sennae and Agraulis vanillae. elsewhere in peninsular Florida.
Research Plan
The investigations in this study focused on the migratory patterns in peninsular
Florida of the two principal migrant butterflies, Agraulis vanillae and Phoebis sennae.
Some attention was also given Urbanus proteus and Precis coenia. The research was
designed as a four part study, carried out during the falls of 1986 through 1988 and
addressed the following questions:
1. Flight direction - Is the mean flight direction of migrants throughout Florida the same
as in Gainesville?
2. Numbers of migrants - Is the migration density across Florida along the Gainesville
latitude constant? How many migrants pass into central Florida annually?
3. Phenology of movement - Does the peak migration at southern sites occur at
approximately the same time as it does in Gainesville? Where does the
migration stop?
Estimates of migration density for the cloudless sulphur and gulf fritillary were
made along the Gainesville latitude (29.65°). From these data, a "migration profile" or
cross section, through the stream of migration for each species was constructed. This

11
information was used, together with Walker’s (1991) data from Gainesville, to estimate
the size of the annual fall migration of these two species moving through north central
Florida. The phenology of the migration was studied by trapping the four principal
migrant species using directional flight traps at five stations from south Georgia to south
central Florida. Counts of north and south flying Phoebis sennae and Agraulis vanillae
also were made on other west-east transects along the Florida peninsula.
The individual chapters have been written as drafts of separate publications.
This has unavoidably resulted in some repetition. The numerical assignment of weeks
for calendar dates during which sampling was done with all methods is given in Table
1-1. The location of all sites used for pole counts and flight trap stations are shown in
Fig. 1-1.
Biology of the Butterflies
Phoebis sennae
Commonly known as the cloudless sulphur, this is a fast flying butterfly of
temporary, disturbed habitats. It is found in the eastern and southern United States,
rarely straying to Canada, through the West Indies and south to Argentina (Lenczewski,
1980). In both the temperate and tropical zones, the subspecies of sennae exhibit strong
migratory habits (Howe, 1975) and C.B. Williams (1938) nicknamed this "The
Travelling Butterfly." The subspecies eubule occurs commonly in the southeastern
United States, straying as far north as Canada (Klots, 1951) during the summer and fall

GEORGIA
FLORIDA
A - Arcadia LE
B - Bartow LA
C - Cross City |_B
CB - Crescent Beach LP
CL - Clermont N
G - Gainesville P
H - Hawthorne S
HA - Hastings j
I - Interlachen v
IM - Immokalee
Leesburg
- Lake Alfred
- LaBelle
- Lake Placid
- Newberry
- Palatka
â–  Steinhatchee
- Trenton
- Valdosta
Fig. 1.1-Location of pole count and flight trap sampling locations used during the fall of 1986, 1987 and 1988.

Table 1-1. Calendar dates of sampling and the corresponding assigned week number for
all sampling methods used during 1986-1988.
1986
1987
1988
WEEKS
1
01-07 SEP
AUG 30-05 SEP
AUG 28-03 SEP
2
08-14
06-12
04-10
3
15-21
13-19
11-17
4
22-28
20-26
18-24
5
29-05 OCT
27-03 OCT
25-01 OCT
6
06-12
04-10
02-08
7
13-19
11-17
09-15
8
20-26
18-24
16-22
9
27-02
25-31
23-29
10
03-09 NOV
01-07 NOV
30-05 NOV
11
10-16
08-14
06-12
12
17-23
15-21
13-19
13
24-30
22-28
20-26
14
29-05 DEC
27-02 DEC
15
06-12
03-10
months. Generally, the species occurs in the Mississippi Valley to Central Illinois and
along the Atlantic coastal plain to New Jersey (Calhoun sí jd, 1990). It is known to
reproduce as far north as west central Illinois (Sedman and Hess, 1985) and Virginia
(Clark and Clark, 1951). In Florida, this species has been recorded for every month of
the year throughout the state, but individuals are most common during the late summer
and fall in north central Florida (Lenczewski, 1980; D. Baggett, pers. comm.). This
subspecies is also resident in the states adjacent to Mexico (Howe, 1975). During some
years, individuals are frequent in eastern Kansas and Missouri where they may be seen
flying in a southeast direction in late summer (Howe, 1975). Similar flights have been

14
recorded in Mississippi and Alabama (Mather and Mather, 1958; Lambremont, 1968),
apparently heading southeast toward the Florida peninsula. Walker’s (1985a) observation
of systematic changes in P. sennae’s flight direction at many locations in the southeast
suggests that these migrants are navigating toward peninsular Florida. The cloudless
sulphur does not exhibit the coastal build up of other species mentioned and, as Walker
has suggested (pers. comm.), may detect the "proper" time to change its flight direction
before hitting the Gulf or Atlantic coastlines from up to 60 miles away! Gaddy and
Laurie (1983) note some consistent discrepancies in flight direction along the South
Carolina coastline for Phoebis sennae. In August, September and early October of 1978-
1980, they noted northeastern migrations along the immediate coast and, what seemed
to be random movements, inland. As previously discussed, flights adjacent to coastlines
can be erratic. Migrating butterflies often seem "confused" and fly in what appear to be
inappropriate directions, more often than not following the coastline in either or both
directions.
Adult males are yellow above and unmarked. Females are a deeper, more
orange yellow, fringed with dark brown marginal spots. The adults are particularly
attracted to red flowers, many of which flower in the fall and are an important source
of nectar for migrating butterflies during that time (pers. obs.). Larvae are a pale
yellowish green with a yellow lateral stripe along each side. They are most often found
at the young shoots of their leguminous hostplants and sometimes web leaves together
for shelter (Howe, 1975). The larvae of this butterfly feed on a wide range of
leguminous hostplants. There are at least 50 species of hostplants recorded and adults

15
will opportunistically oviposit on plants according to availability. The preferred larval
hostplants in Florida for this species include various species of Cassia, particularly
obtusifolia (L.) (pers. obs.), and as Howe (1975) reported in Louisiana, the partridge
pea, Chamaecrista cinérea (L.). Two generations reportedly occur in the northern part
of the range and although Howe (1975) reports that breeding is continuous in the Gulf
region, including Florida, there are no records of such activity during the winter months,
at least in the latter area. During 1987, Calhoun ei al (1990) reported cloudless sulphur
individuals sighted in Wisconsin and New York. Numbers of individuals seen that
summer were unusually large and sightings extended the northern range significantly.
The species was common in Mississippi, Illinois, Kentucky, Indiana, Ohio and West
Virginia. He described temporary breeding colonies of Phoebis sennae in Ohio and West
Virginia, where apparently, due to the unusually warm winter of 1986-1987, individuals
overwintered and were available in early spring to utilize newly emerged hostplants. The
larval hostplants of this species usually senesce in late fall and are absent through the
winter months. Interestingly, these breeding populations in Ohio and West Virginia were
located along river banks, which Calhoun (1990) considered to be primary corridors of
dispersal.
Agraulis vanillae
This butterfly, also known as the gulf fritillary, is commonly found in
temporary, disturbed habitats such as roadsides and fields, usually near its larval
foodplant, during the late summer and fall months in Florida. The species is widespread,

16
ranging throughout tropical America, from Argentina north to the southern United States
(Howe, 1975) and throughout the West Indies (Riley, 1975). The eastern subspecies,
nigrior. occurs from Florida, west to Louisiana, north to North Carolina and is also
found on Bermuda. Although this species has been reported as far north as New York
(Howe, 1975), it is not commonly found above southern Virginia (Clark and Clark,
1951) on the eastern coast. Unlike the cloudless sulphur, this species is not usually seen
during winter months (December-February) in northern Florida. However, there was an
exceptionally warm winter in Gainesville during 1986-87 and adults were seen flying in
January (pers. obs.).
Like the other heliconiids, the larvae of Agraulis vanillae feed only on species
of Passiflora. In Georgia and north central Florida, the caterpillars are commonly found
on the passion flower, Passi flora incamata L. (pers. obs.). The larva has a black body,
with three pairs of dorso-lateral, orange-brown stripes. There are six rows of branching
tubercles on the body and a pair of longer tubercles curving back on the head. The adult
is distinctively colored with bright orange upper wings marked with black and the
hindwing underside covered with metallic silver spots and they are particularly attracted
to the flowers of Spanish needle, Bidens pilosa and lantana, Verbena spp. According to
Klots (1951), there are three or more broods in the south. In Gainesville, Florida, the
butterflies arrive and begin to oviposit as soon as plants are available in early spring,
continuing to breed until the winter months, usually late November (pers. obs.).

17
Urbanus Proteus
Urbanus proteus. the long tailed skipper, is a minor pest of cultivated beans,
known as the bean leaf roller in agricultural literature. This species is distributed from
Argentina through the United States and West Indies. The mainland populations
comprise the nominate subspecies, with all of the insular populations falling into a second
subspecies (Howe, 1975). In the United States, U. proteus proteus is found from
Connecticut, south to Florida and west to Texas, Arkansas, Arizona and California. The
larvae are green and have a dark mid-dorsal line with yellow lateral lines and green
stripes below these. The area between the stripes is dotted with black and yellow spots.
The head is large, reddish brown, with two yellow spots between the ocelli and the
mouthparts. The caterpillars come out to feed from leaf shelters that they construct on
their leguminous hostplants, moving to a larger one at each instar. The major foodplants
include such legumes as Bauhinia. Clitoria mariana. Desmodium. Phaseolus. Soja,
Vigna. Wisteria and Pueraria lobata (Lenczewski, 1980). In the adult, the thorax, wing
bases and hind wings are a conspicuous metallic green. There are four separate square
spots on the central band of the forewing and a larger square spot under the end of the
cell. The male has a costal fold, the antennal club is yellowish brown below and the
range of wingspread is 38-50 mm (Howe, 1975). Adults are found flying in open areas,
particularly fields and along roadsides in disturbed sites. Three generations are reported
in Florida but only representatives of the late-summer brood arrive in Virginia where this
species is considered a casual visitor (Clark and Clark, 1951). Although Howe (1975)
claims it occurs throughout the year, there are no records during the summer months of

18
June and July in south Florida (Dade or Monroe Cos.). Instead, the recently introduced,
nonmigratory species, II. dorantes, with very similar habits is common during this period
(Lenczewski, 1980).
Precis coenia
Commonly known as the buckeye, this butterfly is the only Florida migrant
which demonstrates relatively large scale spring northward migration (Walker, 1991).
The genus Precis is largely confined to the tropical regions of the world, only two
species occur in North America. There has been considerable confusion concerning the
systematics of this species, which can be quite variable in size and color. P. coenia is
found from southern Canada, west to California, Arizona and south through tropical
America and Cuba. Clark and Clark (1951) reported two seasonal forms in Virginia.
The spring/autumn form had an average forewing length of 27 mm in females and 24
mm in males. The summer form averaged a slightly greater wing length and darker
color, characterized by an irregular reddish band on the underside of the hind wings
which the spring/autumn form lacks. Clark and Clark (1951) noted that the wings in
nearly all of the spring/autumn specimens they collected in Virginia were damaged.
These worn wings may indicate long distance travel and this corresponds well to what
has been observed of migration periods for this species. Mather (1967) found light and
dark forms to be correlated with seasonal changes. He found the major seasonal shift
from dark to light occurs between March and April and from light to dark between
August and September. The shifts were found to coincide with a change in the mean

19
temperature from below 16°C to above 16°C and from above 27°C to below 27°C and
not with rainfall, as previously suspected. Scott (1975) investigated the issues of male
territoriality and migratory tendency which seem to be at theoretical odds. To the casual
observer, buckeye males seem highly territorial, dashing from their resting places, to
chase intruders or meet females. Scott (1975) concluded that it is a "psuedo-territorial"
tendency and is not maintained for any significant length of time at a particular site.
The adults are often seen in open country, especially sitting along dirt roads near
disturbed, weedy fields. Three broods occur in the southeast (in Virginia from late May
to late fall). Although the butterflies of the last brood may be seen on warm days
through the first half of winter, they are not found in the spring in Virginia (Clark and
Clark, 1951). This species is considered a summer breeding resident in Virginia and
there are no records of overwintering individuals. However, Clark and Clark (1951)
report having seen ragged, overwintering individuals in March, April and early May in
Washington, D.C. Howe (1975) describes the adults as hibernating and sometimes
migrating. In southern Florida at least, the subspecies, P. coenia coenia. is reported for
every month of the year (Lenczewski, 1980).
The larva has a dark gray body, striped or spotted with orange-yellow. There
are a number of short, branching spines on the body and one pair on top of the head.
The larval foodplants are Ruellia (Acanthaceae); Plantago (Plantaginaceae); Antirrhinum.
Buchnera. Gerardia harperi. Linaria. Mimulus. Scrophularia lanceolata and others
(Scrophulariaceae); Ludwidgia (Onagraceae); Sedum (Crassulaceae); Verbena prostrata

20
and Acuba (Verbenaceae). The Scrophulariaceae appear to be the most important larval
hostplants in Florida.
Other Migrants
Pieris rapae (L.l. An introduction from England into Quebec in 1860, this well-
known migrant has spread rapidly throughout most of North America. Also known as
the cabbage white, it can be a serious agricultural pest on cruciferous crops. Although
very common in agricultural fields in north central Florida, this species is now rather
rare in south Florida (Lenczewski, 1980). Walker (1991) reported northward movement
in the spring in Florida, but it seems this species is also capable of overwintering as far
north as Canada (Scott, 1986).
Vanessa virginienis (Drurv). Although the American painted lady is a common
species in north Florida, this butterfly has been reported by Lenczewski (1980) as
infrequent in south Florida (Dade and Broward Cos). The range of this species extends
from coast to coast in the U.S., from southern Canada to Colombia, the Canary Islands
and Hawaii. It is rare in the Antilles and an occasional vagrant in Europe. This species
was not previously known to be migratory, at least to the extent of V. cardui. As found
for P. rapae in Gainesville, Florida, V. virginiensis showed northward movement in the
spring but individuals were seldom caught in the fall (Walker, 1991). Possibly their
movements are not related to temperature since Scott (1986) reports this species as also
overwintering in Canada.

21
Eurema lisa (Boisduval and Le Conte). Scott (1986) reports this species as
overwintering in the southeastern U.S. and extending its range north to Canada in
summer. At Gainesville, Walker (1991) trapped £. lisa only in the fall when it was
moving south. This species is a well-known migrant and has been reported on occasions
to fly over the Caribbean, the Atlantic and the Gulf of Mexico in huge swarms (Howe,
1975; Klots, 1951; Lenczewski, 1980).
Eurema daira (Godart). This species overwinters in the southeastern Coastal
Plain and spreads north a few hundred kilometers in summer (Scott, 1986). Walker
(1991) found no significant movement south in the fall through Gainesville except in the
fall of 1985.
Eurema nicippe (Cramer). This species sometimes moves in a seemingly
inappropriate direction, northward during fall flights (Walker, 1991) and reportedly
overwinters throughout the southeastern U.S. (Howe, 1975). However, it has not been
reported in December, January or March in south Florida Dade or Monroe Cos.,
although it is very common there in the fall (Lenczewski, 1980).
Danaus plexippus (L.). The monarch is undoubtedly the best studied migrant
butterfly. Large numbers of monarchs aggregate along the Florida panhandle coast each
fall. Most of this eastern population flies west along the Florida coast and down through
Texas to overwintering sites in Mexico (Brower, 1977). The migrating monarchs
generally fly at altitudes (Gibo, 1986) where flight traps are of little use, although
Walker (1991) did trap 15 in ten years and reported a significant southward bias in the
fall. Those individuals that enter Florida become physiologically "trapped" when warm

22
temperatures end their reproductive diapause, and subsequently, migratory tendency
(Brower, 1961; T. Van Hook, pers. comm.). There are reports of several small
overwintering colonies along the west coast of Florida (Brower, 1961; and pers. com.)
but the breeding status of these is unknown. About 20% of female monarchs along the
Florida Gulf coast during fall migrations are mated and will oviposit on available
milkweeds in Gainesville also during that time (T. Van Hook, pers. comm.). This
species is also known to breed in south Florida during the winter months, but the
subsequent migratory tendencies of the offspring are not known (Lenczewski, 1980).
Panoquina ocola (Edwards!. This small skipper is often found concentrated in
large numbers along the Gulf coast with other migrants in the fall (pers. obs.) and has
been reported by Walker (1978) as migrating through Gainesville. A mass movement
of these skippers in Louisiana was described by Penn (1955). Found through most of
the southeast, from Virginia to Florida, west to Texas, Kentucky and Arkansas, it has
also been reported as far north as New Jersey (Howe, 1975). There are no records of
this species in south Florida (Dade and Monroe Cos.) in January (Lenczewski, 1980).
Lerema accius ('Smith'), The range of this skipper is from New England to
Florida, west to Illinois, Arkansas, Texas and south to northern South America. Howe
(1975) reports it as scarce northward but common in the southern states, with records
from February to November. Walker (1978) reported some fall southward movement
for this species and it is present during every month of the year in south Florida
(Lenczewski, 1980).

CHAPTER 2
FLIGHT DIRECTION
Introduction
Of all insect migratory movements, those of butterflies are the most evident and
most accessible to study. Unlike their nocturnal counterparts, the moths, butterflies are
highly visible, daytime fliers which move through the boundary layer (Williams, 1930;
Baker, 1978). This is the layer of air near the ground where wind velocity is less than
the insect’s air speed. The thickness of this boundary layer is variable, determined by
wind velocity as well as the air speed maintained by the individual (Taylor, 1958;
Pedgley, 1982). Most other insect migrants, such as leafhoppers (Taylor and Reling,
1986), noctuid moths, locusts and aphids (Johnson, 1969), travel at much higher
altitudes. Radar studies have shown that they are generally flying with the wind
(Pedgley, 1982), although they may take off from the ground only when wind direction
is favorable (Williams, 1958). Flight within a few meters of the ground, however,
affords the individual more control over direction and also the opportunity to cease flying
by clinging to vegetation should conditions become unfavorable. This type of flight
allows an observer to easily identify species as well as to observe flight direction and
behavior under variable environmental conditions.
23

24
There are at least eight species of migrating butterflies that pass throughnorth
central Florida (Walker, 1978, 1980, 1985ab, 1991) each fall. The four most obvious
and abundant species are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis
vanillae (L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the
long-tailed skipper; and Precis coenia (Hübner) (Nymphalidae), the buckeye. Outside
of Walker’s studies (1978, 1980, 1985ab, 1991; Walker and Riordan, 1981), mostly
based in Gainesville, Florida, there is very little known about the flight directions of
these species. The conspicuously yellow cloudless sulphur, especially, has been reported
travelling in southeasterly fall flights throughout the southeastern states (Lambremont,
1968; Shannon, 1916; Williams, 1930, 1958; Clark and Clark, 1951; Howe, 1975;
Urquhart and Urquhart, 1976; Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981).
Although there are reports of inappropriate northern flights (Gaddy and Laurie, 1983;
Muller, 1977), or variable movements, especially along coastlines (Urquhart and
Urquhart, 1976), the majority of individuals appear to be navigating toward the Florida
peninsula (Walker, 1985a). Walker’s trapping and observational data (1978, 1980,
1985ab, 1991) clearly demonstrate that the migrants pass through Gainesville, in north
central Florida, each fall and maintain an average flight path approximating 141°.
However, for most of the rest of the state, flight paths are unknown. My study was
undertaken to investigate previously unknown migratory flight directions of the four
principal migrant species at various sites throughout the Florida peninsula.

25
Materials and Methods
Flight Azimuths
During the fall migratory season (September-November) of 1985-1988, flight
azimuths for the four migrant species were recorded at a number of locations throughout
the Florida peninsula. Sites selected for observations were large, open areas, such as
high school football fields or large lawns. Hours for observation were peak flight
periods established by Walker (1985a) from 0730 to 1430 HRS LMT. Observation
periods were also limited to weather conditions that are favorable to migration and were
terminated if more than 50% of the sky was obscured by clouds, temperatures dropped
below 21 °C or wind speed exceeded 3 m/sec. For each observation, these conditions
were noted as described in detail by Walker (1985a): civil time, species, appearance of
sun, percentage of blue sky visible, wind speed, wind direction and air temperature. The
butterfly passing closest to the observer was selected and a sighting of its flight path was
then made from the point where it had passed, to where it disappeared on the horizon,
noting the magnetic bearing with a Suunto KB 14 compass. Such observations continued
with the next individual spotted and so forth, also as described by Walker (1985a).
Magnetic bearings were adjusted for magnetic declination, resulting in true bearings and
civil time was converted to local mean time. Walker (1985a) found that mean migratory
directions at Gainesville remained approximately 141 ° regardless of species, season, time
of day or wind. For the purposes of this study, migration data for all dates at a locality
were grouped for each species. These data were analyzed using BUTTAZ.BAS, a

26
computer program designed by Walker (1985a) to calculate mean direction (M.D.),
length of mean vector (r) and the 95% confidence interval of M.D. (Batschelet, 1981,
Zar, 1984).
Mark-Release-Recapture
A number of individuals of Phoebis sennae were also marked and released
during the fall migration season of 1987. This mark-release-recapture took place at two
sites in Gainesville, Florida, from September through December. Both sites had gardens
with a number of flowering plants that the adults used as nectar sources. The most
northern site was Kanapaha Botanical Gardens and the second site was a Gainesville
residence located eight miles south of Kanapaha Gardens on Williston Rd. Between my
visits, observations of marked individuals were recorded by the resident, Dr. Lincoln
Brower, resulting in more complete information at this site. A total of 800 cloudless
sulphurs were netted over a period of three days from 10-12 September and given a
number on the upper and lower surfaces of the right wing with a permanent felt tip
marker, color coded for each site. Sites were visited three times per week until 23
November 1987 and all marked individuals were noted. Observations at the residential
site were continued by Dr. Brower throughout December.
Results and Discussion
The mean flight directions during fall migration 1985-1988 through various sites
in Florida for Phoebis sennae are shown in Fig. 2-1. The length of the line represents

A=Arcadia n=13,r=.58,161 ,p< 01
AP=Alligalor Point n=18,r=.26,68
B=Bartow n=27,r=.88,146,p<.001
BG=Belle Glade n=0
C=Cross City n=51,r=.89,171,p<.001
CB=Ctescent Beach n=0
CL=Clermont n=8,r=.45,191
CW=Clewiston n=4,r=.77,123
E=Everglades City n=2j=.51,32
EP=Eastpoiri n=3,r=.9,46
F=Frostproof n=3,r=.72,92
FC=Fishcreek n=4,r=53320
G=Gainesville 141 (Walker, 1985a)
GS=Glen St. Mary n=5j=.59,126
H=Hawthome n=61,r=.81,152,p<.001
HA=Hastings n=9,r=.33,175
l=Interlachen n=34,r=.94,168,p< 001
IM=Inimokalee n=12j=.62,170 ,p<.01
J=Jasper n=4,r=.!2,162
KB=Keeton Beach n=6,r=.29,330
L=LakeCity n=4,r= 80,146
LE= Leesburg n=8,r=.96,133,p<001
LA=Lake Alfred n=7,r=.74,165,p<.02
LB=LaBeUe n=13,r=.66,153,p<.005
LP=Lake Placid n=4,r=.79,71
N=Newberry n=41j=.80,146,p<.001
OOIga n=13/=.84,178 ,p<.001
OK=Okeechobee n=12j=.65,77,p<.005
P=Palalka n=17,r=.72,143 ,p<.001
PE=Perry n=18^=.93,144 ,p< 001
S=Steinhatchee n=46,r=.42,86,px. 001
SG=St. George n=2,r=l,54
T=Trenton n=25,r=.72,151 ,p<001
TA=Tallahassee n=9,r=.32,94
W=White Springs n=2,r=.88,167
YJ=YeehawJunction n=l,r=l,159
Fig. 2-1.-Summary of mean flight directions at various sites throughout Florida for Phoebis sennae during fall migrations,
1985-1988 (direction of arrow=mean flight direction, value given in key for each site; length of line=mean vector r; bold
letters=significant r value).
to

28
the mean vector (r) with sites labeled in bold letters indicating statistically significant
vector values for those sites with n >5 using Rayleigh’s z test (Zar, 1984). Gainesville
data is, as previously reported by Walker (1985a), 141°. The observations made in this
study along the latitude of Gainesville also show a similar pattern of movement. The
inland sites south of Gainesville correspond to the expected flight direction if migrants
were continuing along the same track observed at Gainesville. Although not statistically
significant, due to small sample sizes at many sites, the movement along coastlines
seemed confused and quite variable. Along the Gulf coast, butterflies usually followed
the coastline even in inappropriate directions. Previous reports of Phoebis sennae flight
directions are southeastward inland and coastwise when near the coast, as reported by
a number of different observers over many years (Shannon, 1916; Williams, 1930, 1958;
Clark and Clark, 1951; Lambremont, 1968; Howe, 1975; Urquhart and Urquhart, 1976;
Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981; Walker, 1985a).
Phoebis sennae is particularly sparse on the east coast considering the southeast
bearings of inland migrants. It has been observed by Walker (1985a) that even 90 km
from the coast, butterflies in south Georgia fly more southerly than further inland, as
though they can detect the coast. This trend is also seen here and no individuals of any
migrant species were observed at Crescent Beach, the easternmost site of the transect.
The reason for an absence of migrants along the east coast portion of the transect is not
clear. Migrating butterflies do tend to follow coastlines possibly using a different
orientation mechanism used from inland flight. If north or south coastline flight is not
distinguished, then by avoiding the east coast altogether, northward flight during the fall,

29
possibly resulting in freezing, could be prevented. Dave Baggett has reported
(pers.comm.) large numbers of long-tailed skippers flying south along the St. John’s
River near Jacksonville during some years. This geographical feature may serve as a
path when found and, if it can be detected at a distance, may create a funnel drawing
migrants away from the east coast areas. Calhoun el al (1990) reported some
colonization by cloudless sulphurs along rivers that appeared to serve as "dispersal
corridors." It has also been observed (see Chapter 3) that this "shadow" or absence of
migration along the east coast seemed to widen inland with distance travelled south.
Other possible reasons for the absence of these migrating butterflies along the Florida
east coast may be the lack of an appropriate habitat or some other adverse environmental
conditions found there.
Throughout the state, significant migratory flight directions for Phoebis sennae
are usually south or southeast. It is noteworthy that in this species, there appears to be
a shift in mean direction immediately north of Lake Okeechobee, as though the migrants
"detect" the lake. At the town of Okeechobee, the mean migratory track shifted to the
east and most individuals (79%) avoided flying toward the large expanse of water. No
cloudless sulphurs were observed along the southern boundary of the lake at Belle Glade,
just a short time after seeing many individuals flying through Okeechobee. Again, this
seemed to be consistent with observations that the butterflies at the northern border of
the lake were going around, rather than flying directly over water. A distinct southward
migratory flight was detected as far as Immokalee, just southwest of Lake Okeechobee.
This is also the northern limit for much of south Florida’s more tropical fauna and flora

30
and the southern boundary for many northern species. Historically, this is about the
point at which damage from a hard freeze is unlikely. It is important to consider the
number of visits to a site and the pattern of individual flight directions shown in Fig. 2-2,
rather than the mean vector alone. For example, although the mean vector at
Steinhatchee (Fig. 2-1) is approximately due east, from Fig. 2-2, it is evident that the
individuals had no consensus in migratory direction. The same is true at Alligator Point
during the only visit there. The eastward vector in Fig. 2-1 really is a result of south
and north flying individuals as shown in Fig. 2-2.
The mean flight direction patterns for the gulf fritillary throughout the state are
similar to those described for the cloudless sulphur and are shown in Fig. 2-3.
Significantly directed movement for the gulf fritillary continued south to Immokalee, but
fewer individuals were observed. Whereas Phoebis sennae is not common along the gulf
coast, Agraulis vanillae masses along the panhandle coast in large numbers (pers. obs.).
Also, unlike Phoebis sennae. the gulf fritillary showed no evidence of avoiding Lake
Okeechobee. Three individuals were observed flying southeast from Okeechobee and
others were apparent at the opposite shore in Belle Glade, heading in the same direction.
However, this species too avoided Crescent Beach, the most eastern site along the
Atlantic Coast.
Whereas some cloudless sulphurs do overwinter in north Florida and buckeyes
have also been reported to do so (Walker, 1978; Scott, 1986), this is not usual for the
gulf fritillary. The more highly directed flight of the gulf fritillary may be necessary to
ensure survival before freezing temperatures occur. The individual flight directions

ARCADIA ALLIGATOR POINT BARTOW REI .IP GLADE CROSS CITY
O
CLERMONT CRESCENT BEACH CLEWISTON EVERGLADES CITY EASTPOINT
© O ©
© o
FROSTPROOF FISHCREEK GLEN ST. MARY HAWTHORNE HASTINGS
© © ©
INTERLACHEN IMMOKALLEE JASPER LAKE CITY
^ J 0 ® 0
LEESBURG LAKE ALFRED LABELLE LAKE PLACID NEWBERRY
0 Q
ADCLLL fLAV.
© 0
%
OLGA OKEECHOBEE PALATKA PERRY STEINHATCHEE
31
ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
0
Or 0 0~
Fig. 2-2.--Azimuths of Phoebis sennae flight directions taken at various locations in
Florida during fall migration 1985-1988. The diameter of the circle represents the
number of visits (5 mm = l visit); the length and width of the line represents the number
of individuals (2 mm = l, in bold 2 mm=4); the orientation of the line represents the
flight direction.

A=Arcadia n=6,r=.98,170 .jx.OOl
AP=Alligator Point n=13,r=.17,164
B=Bartown=10,r=.84,174 ,p<.001
BG=BeIle Glade n=0
C=Cross City n=21,r=.89,156 ,p<.001
CB=Crescent Beach n=o
CL= Clermont n=8,i=.6,142
CW=Gewiston n=5,r=.92,146
E=Everglades City n=6,r=.62,109
EP=Eastpoint n=2j=.99,277
F=Frostproof n=l,r=l,163
FC=Fishcreek n=0
G=Gainesville (Walker, 1985a)
GS=Glen St. Mary n=0
H=Hawthome n=27 j=.84,160 ,p<.001
HA=Hastings n=4j=.84,132
I=lnterlachen n=14j=.96,179 ,p<001
IM=Immokalee n=16j=.72,164 ,p<.001
J=Jasper n=4j=.92
KB=Keeton Beach n=4,r=.7,290
L=Lake Cityn=4,r= 66,154
LE=Leesburg n=5,r= 54,163
LA=Lake Alfred n=8,r=.96,155 ,p< 001
LB=LaBelle n=2,r=.97,164
LP=Lake Placid n=3,r=.99,175
N=NewberTyn=16,r=.97,148 ,p<.001
0=01ga n=2j=.17,253
OK=Okeechobee n=4j=.96,143
P=Palatka n=12j=.84,172 ,p<.001
PE= Perry n=3j=.94,151
S=Steinhatchee n=13j=.80,137 ,p<.001
SG=St. George n=3,r=.56,141
T=Trenton n=6,r=.34,147
TA=Tallahassee n=9,r=.32,146
W=White Springs n=5,r=.58,174
YJ=Yeehaw Junction n=l »r=l ,159
P
Fig. 2.3.--Summary of mean flight directions at various sites throughout Florida of Agraulis vanillae during fall migrations, 1985-
1988 (direction of arrow = mean flight direction, value given in key for each site; length of line=mean vector r; bold
letters=significant rvalue).
u>
N>

33
observed throughout Florida during fall 1985-1988 are shown in Fig. 2-4. Like P.
sennae. Agraulis vanillae showed a great deal of variation in flight direction along the
Florida panhandle coast at sites such as Alligator Point. However, at
Steinhatchee, which is further south along the west coast, there was no evidence of the
confused flight that 2. sennae had shown and movement was significantly biased
southeast.
The summary of mean flight directions of Urbanus proteus throughout Florida
are shown in Fig. 2-5. Its migratory flight patterns are similar to those of the gulf
fritillary. Also common along the Gulf coast, this species exhibited a great deal of
variation in flight at coastal sites. This is evident in the individual flight directions from
various sites in Florida shown in Fig. 2-6. For example, at Keeton Beach, movement
was in every direction. Oddly, a similar variation in flight was noted at Yeehaw
Junction, an inland location just north of Lake Okeechobee. However, a few miles
directly south, at Okeechobee, the flight of individuals was very consistently due south.
It appeared they were heading out across the water, but just south of the lake, at
Clewiston and Belle Glade, there were no sightings of incoming individuals. It is
possible that the long tailed skipper makes shorter, more localized flights than the other
two species discussed at inland sites. This may also account for a great deal of
variability in flight directions.
Very few individuals of Precis coenia were recorded throughout the state during
fall migrations. This is consistent with Walker’s (1991) findings from flight traps in
Gainesville that most of the movement is northward in the spring. In this study, of the

34
FROSTPROOF FISHCREEK
O o
DMTERLACHEN
OLGA OKEECHOBEE
© o
ST. GEORGE TRENTON
GLEN ST. MARY HAWTHORNE
HASTINGS
LAKE CITY
©
LABELLE LAKE PLACID NEWBERRY
TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
©
Fig. 2-4.—Azimuths of Agraulis vanillae flight directions taken at various locations in
Florida during fall migration 1985-1988. The circle diameter represents the number of
visits (5 mm = l); the length and width of the line represents the number of individuals
observed (2 mm = l); the orientation of the line represents the flight direction.

N
A=Arcadia n=9,r=81,168
AP=Alligator Point n=0
B=Bartow n=8,r=.66,156
BG=Belle Glade n=0
C=Cross City n=12,r=.82,147
CB=Crescent Beach n=0
CL=Clermont n=3,r=.82,147
CW=Clewiston n=0
E=Everglades City n=6j=.48.121
EP=East point n=0
F=Frostproof n=2,r=.94,143
FC=Fishcreek n=2,r=.64,318
G=Gainesville (Walker, 1985a)
GS=Glen St. Mary n=0
H=Hawthome n=6,r=.89,157
HA=Hastings n=3j=.31,338
I=Interlachen n=3,r=.67,135
IM=lmmokalee n=ll j=.68,177
J=Jasper n=l,r=l,152
KB=Keeton Beach n=22,r=.28,168
L=Lake City n=0
LE=Leesburg n=12j=.89,145
LA=Lake Alfred n=2tr=l,347
LB=LaBelle n=7j=.53,215
LP=Lake Placid n=l,r= 1,277
N=Newberry n=l,r= 1,280
0=01ga n=5,r=.87,168
OK=Okeechobee n=6,r=.97,186
P=Palatka n=o
PE=Perry n=2,r=l,162
S=Steinhatchee n=23,r=.91,120
SG=SL George n=0
T=Trenton n=3,r=.98,152
TA=Tallahassee n=0
W=White Springs n=0
YJ=Yeehaw Junction n=13j=.55,144
Fig. 2-5.--Summary of mean flight directions at various sites throughout Florida of Urbanus proteus during fall migrations of 1985-
1988. (The direction of the arrow=mean flight direction, value given in key for each site; the length of line=mean vector r; bold
letters = significant r value). w

3e
ARCADIA
ALLIGATOR POINT
BARTOW
BELLE GLADE
CROSS CITY
©
O
©
O
©
CLERMONT
Q
CRESCENT BEACH
O
CLEWISTON
o
EVERGLADES CITY
®
EASTPOINT
O
FROSTPROOF
FISHCREEK GLEN ST. MARY
HAWTHORNE
HASTINGS
GOO
IMMOKALLEE INTERLACHEN JASPER LAKE CITY
o ©
O
LEESBURG LAKE ALFRED TABELLE LAKE PLACID NEWBERRY
©
© o o
OLGA OKEECHOBEE PALATKA PERRY STEINHATCHEE
Q CD
Q G©
ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
O Q O O
Fig. 2-6.--Azimuths of Urbanus proteus flight directions during fall migrations, 1985-
1988. The circle diameter represents the number of visits (5 mm = l); the length and
width of the line represents the number of individuals (2 mm=l); the orientation of the
lline represents the flight direction.

37
two sites where there were more than two individuals recorded, the movement was
southward. The only location where buckeyes were sighted south of Gainesville was in
Bartow, central Florida and those few individuals were headed north.
Mark-Release-Recapture
In two instances, cloudless sulphurs were recaptured at some distance from the
mark-release sites in Gainesville. These are the first instances of an individual Phoebis
sennae being tracked over a relatively long distance and time. The first individual had
been marked at Kanapaha Gardens and was seen again feeding at flowers at the Williston
Rd. residence, eight miles south of there, ten days later. The second individual had been
marked at the Williston Rd. residence and was recaptured in Bronson, Florida, 51 km
southwest of Gainesville, 14 days later. This cloudless sulphur had evidently taken
refuge in some potted plants during a cold evening and had been brought into the house.
In both instances, the butterflies travelled a much shorter distance than would be expected
from the elapsed time. It has been calculated (Arbogast, 1966; Balciunas and Knopf,
1977) that migrating cloudless sulphurs or gulf fritillaries go 15-20 km/hr under ideal
conditions. Thus, such a journey could easily be completed in one day.

CHAPTER 3
NUMBERS OF MIGRANTS
Introduction
Each fall, a migratory mass of butterflies comprised of at least eight species
from four families, passes through north central Florida. The most conspicuous of these
migrants is a pierid, Phoebis sennae (L.), commonly known as the cloudless sulphur.
This butterfly has been reported throughout the southeast during fall migratory flights
which are usually toward the Florida peninsula (Walker, 1985a). The three other
principal migrant species are Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary;
Urbanus proteus (L.) (Hesperiidae), the long-tailed skipper; and Precis coenia (Hiibner)
(Nymphalidae), the buckeye. Large numbers of these species have been noted along the
Florida Gulf coast (Urquhart and Urquhart, 1976) as well as passing through north
Florida regularly (Walker 1985ab, 1978, 1980, 1991).
It has been difficult to estimate the size of this migratory flight. Walker (1991)
has operated traps for more than ten years in Gainesville, Florida, but the magnitude of
migration had not been monitored at any other location. In order to accurately estimate
the total numbers of butterflies moving south into central Florida each fall, the density
of migratory flights must be monitored at other locations along the latitude of Gainesville
(29.65°). Walker’s (1991) long term trapping results from Gainesville could then be
38

39
projected to estimate the absolute numbers moving south each fall. The goal of this
study was to construct a "migration profile" of density estimates for the two principle
migrants, Phoebis sennae and Agraulis vanillae. along a transect across north central
Florida at the latitude of Gainesville. This cross section through the migratory stream
will then be used to estimate the total fall migration of Phoebis sennae and Agraulis
vanillae into peninsular Florida.
Materials Methods
Latitudinal Pole Counts
In the fall of 1986, a counting system previously used by Walker (1985a) was
used to estimate the numbers of migrants moving through various sites along the
Gainesville, Florida, latitude (29.65°). Three 3-meter lengths of PVC pipe were used
with the second pipe set in the ground 15 meters from the first, and the third, 30 meters
beyond the second. All three pipes were along a straight line at 51-231°, perpendicular
to the average flight track of 141° established in Gainesville by Walker (1985a). All
observations were made in large, clear areas (i.e. athletic fields or lawns) to avoid any
vegetative or structural influence on the paths taken by migrants. The air temperature
(°C), percentage of clear sky and sun exposure (b=bright, h=hazy, o=obscured by
clouds, p=disk obscured, but position easily determined ±5°) were noted. Wind speed
was measured with a hand held pith ball anemometer and direction was noted. The
numbers of individuals of the four principal migrant species were counted crossing
between two poles for three five-minute periods with a one minute break between each

40
observation period. Cloudless sulphurs were easily observed at 45 meters and therefore
individuals flying between the first and third poles, were counted. Long-tailed skippers,
gulf fritillaries and buckeyes were less visible and were counted if within 15 meters
(between the first and second poles). Counts were made at least once every three weeks
from 1 September to 2 November 1986 at most of the ten sites (see Table 1.1 for
assigned weeks and Fig. 1-1 for location of sites). This period has been shown by
Walker (1991) to include >95% of the total migration through Gainesville.
Observations were adjusted for time of day differences by using a "time-of-day"
factor, T, from Walker’s (1985a) observations of each species’ flight variation throughout
the day. The average for each half hour was calculated from three days of observation
(3 October 1982 and 4, 11 October 1983) by Walker of percent individuals flying
throughout the day. This average percent of individuals flying was then multiplied by
the T factor which would adjust up to 100% for that time period. A 15 minute
adjustment to T was made by averaging two adjacent half hourly observations. For
example, if 6.4% of the day’s total was the average seen flying at 1000 HRS (EDT)
(counts made between 0945-1014) then T would be 15.6. An average of 5.9% flying at
1030 HRS (counts between 1015-1044) results in a T of 16.9. To adjust for 15 minute
periods, the average percent flying for both half hourly observations was averaged with
a result of 6.15, the resulting T factor for observations at 1015 HRS would then be 16.3.
Observations were grouped according to their closest time period in 15 minute intervals
and multiplied by the appropriate T factor. The west-east transect sites were
approximately 23 kms apart and extended the width of the state from Steinhatchee to

41
Crescent Beach, a distance of about 200 km. All results of pole counts were summarized
by averaging the net numbers of individuals/m/minute from samples made at each site
throughout the season.
Migration Profiles
During the fall migrations of 1986, another technique was devised to estimate
the net migration across west/east transects. Counts of north and south flying butterflies
were made while driving along selected roads running west to east. This method covered
entire transects rather than the 15 or 45 m samples made during the pole counts.
Transects with low vegetation and therefore good visibility were selected. The major
transect was at the latitude of Gainesville, through the same sites as the pole counts.
Ideally, the entire transect was travelled roundtrip once per week, with half the counts
completed on a particular day. Only cloudless sulphurs and gulf fritillaries were included
in the drive counts because they were easily identifiable at a distance. To be counted,
a butterfly had to at least have reached the centerline of the road before the car passed.
A speed of approximately 97 kph was maintained and counts were recorded for every 16
km of transect from which the net number of indiv/min was calculated. Standard time
was noted at the beginning and end of each segment of transect, as was weather.
Weather conditions recorded were the percentage of blue sky visible, the appearance of
the sun (b=bright, h=hazy, o=obscured, p=disk obscured, but position easily
determined ±5°), air temperature (°C), wind direction and speed (m/sec with a hand
held anemometer).

42
During this study, the analysis of the data presented several problems incurred
by the distance and time required for sampling and daily and seasonal changes in the
density of migration. Walker and Riordan (1981) have shown that by sampling only on
days that meet certain weather criteria, day-to-day fluctuations in numbers can be
minimized. Driving periods were limited to weather conditions favorable to migration
and were terminated if more than 50% of the sky was obscured by clouds, the
temperature dropped below 21 °C or wind speed exceeded 3 m/sec. There are however,
hourly differences in migration throughout the day with peak flight times usually from
1000 to 1400 HRS EDT (Walker, 1985). Since it was not possible to sample
simultaneously along the transect sections, it was necessary to accommodate the daily
changes in peak flight periods. Adjustments for time of day differences were made to
the actual net numbers observed during drive counts (D) by multiplying with the "time
of day" factor, T (see explanation for pole counts above).
As a result of time limitations, unpredictable weather changes and other
occurrences during sampling periods, not all segments along a transect were always
visited on a particular sampling trip. To equalize for the differences in sample number
from each transect segment, the time adjusted counts (DnT) were then transformed
relative to a time corrected "base" segment (DjT) which was always sampled during each
trip. For the western portion of the transect (Gainesville-Steinhatchee), this base was the
Gainesville-Newberry segment. For the eastern portion of the transect (Gainesville-
Crescent Beach), the base was the Gainesville-Hawthome segment. The net number of

43
individuals observed along each transect segment was converted to a proportion of the
net number observed along the base segment = DnT/D,.
Another important influence on the amount of migratory movement observed is
a seasonal factor (Walker, 1991). Since the drive counts were made over a period of
several months, numbers had to be adjusted for season. In addition, partial profiles from
different drive counts had to be combined to estimate the entire trans-state profile.
Walker (1991) has collected phenological information on the migration pattern in
Gainesville, Florida for more than ten years. His two permanent traps sample
continuously throughout the season and have been shown to reflect migration patterns
reliably. To eliminate the effects of season and to unify the east and west transects,
Walker’s trap catches were used to quantify the "strength" of migration for each day that
drive counts were made. The net numbers of gulf fritillaries and cloudless sulphurs
captured in Walker’s two traps were summed for the period during which drive counts
were made that year and a mean net number was calculated. The daily net number
caught in his traps was then divided by the mean net number to produce a factor that
would adjust to the mean net number. The reciprocal of this factor was called the
migration index, I, for that day. If the migration was particularly strong, the migration
index was correspondingly low. The adjusted data from drive counts was then multiplied
by the migration index appropriate to the day that sampling occurred. In this way, the
high numbers observed during the peak migratory season were reduced and the more
sparse, late or early migration was augmented, in effect eliminating the seasonal

44
difference. The results of these calculations yielded the estimated migration profiles for
1986, 1987 and 1988.
D„ = NET # INDIVIDUALS/MINUTE FROM DRIVE COUNTS
T = TIME OF DAY ADJUSTMENT
D, = NET # INDIVIDUALS/MINUTE AT BASE SITE (TIME CORRECTED)
I = MIGRATION INDEX
ESTIMATED MIGRATION PROFILE = I (D^T/Dj).
The summary migration profiles were then derived from the median segment-by¬
segment values of the yearly estimated profiles for each species. These cross sections
of migration density through north central Florida were used to estimate the annual
number of P. sennae and A. vanillae butterflies flying into central Florida by applying
the migration profiles to Walker’s (1991) Gainesville permanent trap data.
Results and Discussion
Latitudinal Pole Counts
Cloudless sulphur
Pole counts for the fall 1986 cloudless sulphur migration are summarized in Fig.
3-1. The two coastal sites showed no net movement. At Steinhatchee, this was due to
equal numbers of individuals flying north and south, and, at Crescent Beach it was due
to a lack of migrants. All net movement southward was statistically significant, except
for Palatka, also near the east coast, where not many individuals were seen. Peak

45
STNCRC TRN NWB GNV HAW ITL PLK HST CRB
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-1.--Summary of pole counts showing Phoebis sennae migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65°). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p <.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB = Newberry, GNV = Gainesville, HAW = Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.

46
numbers of individuals were counted through the central sites of Newberry, Gainesville
and Hawthorne, averaging about .45 indiv/m/min. The highest count was through
Gainesville at .57 indiv/m/min and numbers declined toward both the Gulf and Atlantic
coasts.
Gulf fritillarv
Pole counts taken of the gulf fritillary migration during fall 1986 are shown in
Fig. 3-2. All net movement was southward and the peak migration was concentrated
slightly more to the west, passing through Trenton, Newberry and Gainesville. Peak
numbers were recorded in Trenton at .58 indiv/m/min followed by Newberry at .48
indiv/m/min. Whereas the cloudless sulphur had little or no migration on either coast,
the west coast migration for the fritillary was relatively strong (.3 indiv/m/min). Few
individuals were counted at Hastings, which was also the only site lacking statistical
significance to the southward net movement. At Crescent Beach however, directly along
the coast, movement was recorded at .13 indiv/m/min. The greater presence along the
coastlines of this species corresponds to the gulf fritillary buildup along the panhandle
coast.
Long tailed skipper
The summary for fall 1986 pole counts of long tailed skipper net movement
across the west-east transect at the latitude of Gainesville is shown in Fig. 3-3. The
greatest number of individuals move south through the central sites of Gainesville-
Hawthome (.21 indiv/m/min), but migration of Urbanus proteus is also substantial along
both the Gulf and Atlantic coastlines (. 14 and . 13 indiv/m/min, respectively). Centrally

47
„„ STN CRC TRN NWBGNV HAWITL PLK HST CRB
O.D r — —
» i i i T i i i ~— r"
O 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-2.--Summary of pole counts showing Agraulis vanillae migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65°). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p C.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB = Newberry, GNV=Gainesville, HAW=Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.

48
STN CRC TRN NWB GNV HAW ITL PLK HST CRB
0.6
0.5
z
5
2 0.4
>
Q
5 0.3
h
LU
z 0.2
-a
O
<
0.1
0
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-3.--Summary of pole counts showing Urbanus proteus migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65°). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p <.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN = Trenton, NWB = Newberry, GNV = Gainesville, HAW = Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.

49
located Gainesville, as well as the two coastal areas were the only locations with
statistically significant movement (chi-square, p < .05). When the sampling period was
subdivided to early, middle and late migration, a single, statistically southward biased
peak shifted eastward from Steinhatchee to Gainesville to Crescent Beach, respectively.
The total fall migration resolved into three peaks, which may represent three pathways
into central Florida. Large numbers of long tailed skippers were counted near the west
coast, but the highest counts were in the central region, passing through Gainesville.
A third peak in migration began at the St. John’s River in Palatka and extended to the
east coast. Dave Baggett (pers. comm.) has reported large numbers of II. proteus
migrating along the St. John’s River during the fall of some years. This pattern may
reflect some geographical obstacles, visual attractions, habitat preferences or source
areas, such as bean fields.
Migration Profiles
Cloudless sulphur
Profile estimates of Phoebis sennae migration in 1986 across the west-east
transect at the Gainesville latitude are shown in Fig. 3-4 a,b,c,d,e,f,g,h. The peak net
southward movement along the western half of the transect (=STN-GNV portion)
occurred through the TRN-NWB or NWB-GNV segments on 14 October, during week
7 of sampling. Throughout the first five weeks, the movement along the western transect
remained fairly constant and relatively low, <_ 2 indiv/min. During week 6, this rate
doubled and in week 7, counts in the western segments (Steinhatchee-Newberry)
increased dramatically. Along the peak transect segment (CRC-TRN, Fig. 3-4g, west)

Fig. 3-4 a,b,c,d,e,f,g,h.—Profile estimates of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65°) from 13 September to 28 October
1986. Counts were taken on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC = Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK=Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.

•)
WEST
WEEK 2
EAST
|o
i*
* 4
h g.
Ill D
z
3 20
<30
40
13 SEP
14 SEP
â„¢ â– 
â–  X X
BTNCftC TUN NWB Of
â– 
X X
VtTN CUC TRN NWB ONV
• « it in i» M4 iao m im
KILOMETERS
C)
WEST
WEEK 4
EAST
5 2
5 4
K 6
Z 8
3io
*12
28 SEP
26 SEP
“ 1 "
X
5TN ene TRN NWB
â–  â–  â– 
x X
QNV HAW m. PLKH8TCRB
0 16 32 48 64 BO 96 112128144160176192
KILOMETERS
b)
I»
S 2
z
5 4
6
! 10
12
EAST
WEEK 3
EAST
20 SEP
rwpi
20 SEP trip]
Tn -
ONVNAWrTL PLK MBT CUB G
W HAW ITL PLK HST CRB
96 112126144160176192 96 112126144160176192
KILOMETERS
WEST
WEEK 5 : OCT 2
EAST
KILOMETERS
EAST
WEEK 7
EAST
lo
Í2
• 4
ÍL 6
z
3 8
< 10
12
‘ , 140CT
TWP1
14 OCT
TRIP 2
â–  - "
X X
ONV HAW m. PLK HST CRB ON
8 â–  1
X X
r HAW ITL PLKHST CRB
KILOMETERS
9611212814416017619296112128144160176192
KILOMETERS
WEST WEEK 7 WEEK 8 EAST
KILOMETERS
h)
EAST WEEK 9 EAST
TRIP 1 28 OCT
28 OCT TRIP 2
I'1
X X
3NVHAW ITL PLKHSTCRI
i fl
mv haw m
X X
PLKHST CRB
96 11212814416017619296 112128144160176192
KILOMETERS

52
at this time, the net number of individuals flying south/minute was nearly 40, 8 times
greater than seen previously. Unfortunately, that was the only drive count through CRB-
TRN during the entire fall. During week 9, the GNV-HAW and ITL-PLK segments
along the eastern transect also increased, but only to about a third of what was seen on
the Gulf coast. On 16 October, the migration through the two Gulf coast segments
(STN-CRC, CRC-TRN) was southward and twenty times higher than that of any other
segment, at any other time, throughout the season. It is not known if large numbers of
migrants regularly pass through these segments. The one additional sample through
STN-CRC (Fig. 3-4c, west) and four samples at TRN-NWB (Fig.3-4 a,c,d,e, west) seem
to indicate that at least during the early season it is not common. It is possible that such
large scale movements result from synchronous emergence of a locally produced
generation. Large fields observed in the Trenton area had extensive stands of Cassia
obtusifolia. a primary hostplant of cloudless sulphurs in this area.
Similar profile estimates of the cloudless sulphur migration during fall 1987 are
shown in Fig. 3-5 a,b. All net movement was southward and greatest along the western
transect through GNV-HAW during week 6. This segment had more than 300 indiv/min
passing south, but the net catch for Gainesville that day was only one cloudless sulphur,
flying south! This resulted in a high migration index that perhaps inflated what was an
average day in the HAW-ITL segment. Along the eastern transect, the numbers were
highest through the TRN-NWB segment during week 7, with 4.5 indiv/min. This
transect was sampled on 17 October, almost one year to the day when extremely large

53
WEST
WEEK 4
WEEK 6
EAST
KILOMETERS
bt WEST wee|( 7 WEEK 8 EAST
KILOMETERS
Fig. 3-5 a,b.--Profile estimates of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65°) from 26 September to 18 October
1987. Counts were taken on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW = Hawthorne, ITL = Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.

54
numbers had been noted there in 1986. The peak numbers along this western transect
in 1987 were only one tenth of what they were the previous year.
The profile estimates showing Phoebis sennae migration along this transect in
fall of 1988 are given in Fig. 3-6 a,b,c,d. Each transect consists of a roundtrip made
during the same day and close in time (trip 1 and 2). Each trip consistently showed the
peak net movement occurred southward through the central, NWB-INT, or western STN-
TRN segment and was lowest in the east.
All profile estimates made of Phoebis sennae net movement during the fall of
1986, 1987 and 1988 are summarized as yearly migration profiles in Fig. 3-7 a,b,c.
Most transect segments showed a statistically significant (chi-square test, p < .05) net
southward movement throughout the sampling period. The exceptions were PLK-HST,
where in 1986 and 1988, net movement was not significantly biased and HST-CRB, in
1987, where no individuals were sighted. During 1986 (Fig. 3-7a), the peak southward
movement was through the western segment of CRC-TRN, 37 indiv/min, followed by
STN-CRC with 14 indiv/min. As was pointed out in the individual 1986 profile
estimates (Fig. 3-4g), this was a result of one sample taken on 16 October. If this trip
were not included, the central transect segment from Trenton to Palatka would contain
most of the southward movement. Peak movement south for the fall of 1987 (Fig. 3-7b)
was through the HAW-INT segment at nearly 200 indiv/min, the highest rate recorded
through three years of sampling. The secondary peak was through the adjacent segment
of GNV-HAW with about 11 indiv/min. Both of these segments showed an average

a) WEST b) east
KILOMETERS KILOMETERS
WEST EAST
Fig. 3-6 a,b,c,d.--Profile estimates of Phoebis sennae migration (net number of individuals/minute, adjusted for time of day,
sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65°) from 18 September to 12
October 1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB=Newberry, GNV =Gainesville, HAW = Hawthome, ITL=Interlachen, PLK=Palatka, HST=Hastings,
CRB=Crescent Beach. x=unsampled sites.

Fig. 3-7 a,b,c.--Migration profiles summarizing estimates of Phoebis sennae migration
across a west-east transect in north central Florida along the latitude of Gainesville
(29.65°). Yearly migration profiles were derived from the median estimate (net number
of individuals/minute, adjusted for time of day, sample size and season). Counts were
made from 13 September-28 October 1986, 26 September-18 October 1987 and 18
September-12 October 1988 on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW = Hawthorne, ITL = Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically significant (chi-
square, p < .05) southward movement and open bars were sites with no statistical bias
(chi-square, p < .05) in net movement. Total number of visits made to each site are
shown above the bars.

ADJ NET * INDIV/MIN SOUTH ADJ NET * INOIV/MIN SOUTH
57
KILOMETERS
KILOMETERS
c)
KILOMETERS

58
migration rate, but traps in Gainesville caught very little. As a result, the high migration
index for that day has increased numbers for these segments. If these days were not
included, the GNV-HAW segment would still have the peak southward movement at 2.3
indiv/min, followed by the TRN-NWB segment at 1.4 indiv/min. There were no
cloudless sulphurs observed during the one trip made through the HST-CRB segment
along the Atlantic coast in 1987. The migration profile for fall of 1988 (Fig. 3-7c)
showed a peak through HAW-INT of 2.7 indiv/min, nearly twice that of the next highest
peak through NWB-GNV (1.4 indiv/min). Movement along either coast was low, with
both east and west segments below .5 indiv/min. For all three years, the Atlantic coastal
segment of transect, Palatka to Crescent Beach, consistently had the least number of
migrants, averaging about .18 indiv/min.
Gulf fritillarv
The profile estimates of Agraulis vanillae migration during fall 1986 are shown
in Fig. 3-8 a,b,c,d,e,f,g,h. On the western transect, movement mostly occurred through
the TRN-NWB segment and peaked during week 6 (Fig. 3-8e) with 20 indiv/min. The
Gulf coast segment of STN-CRC, was not sampled during this time, but the following
week (Fig. 3-8g), it showed peak movement south at 8 indiv/min. Numbers were low
along the eastern transect during all the weeks sampled, except for large numbers (23
indiv/min) during week 9 (trip 2, Fig. 3-8h) through INT-PLK. All peak movement
through the season occurred within this segment, except for week 9 (trip 1, Fig. 3-8h)
where GNV-HAW had nearly 15 indiv/min.

Fig. 3-8 a,b,c,d,e,f,g,h.--Profile estimates of Apraulis vanillae migration (net number
of individuals/minute, adjusted for time of day, sample size and season) across a west-
east transect at the latitude of Gainesville, Florida (29.65°) from 13 September to 28
October 1986. Counts were made on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.

60
b)
♦ I
•)
KILOMETERS
WEST WEEK 6 :12 OCTOBER WEST
Â¥ 0
5 *
z
«10 .
020
<
25
TRIP 1
TRIP 2
X X
ITN CMC TUN
â– 
NWB ONV
X X
5TN CPC TUN
â– 
NWB QNV
0 16 32 48 64 60 96 8 24 40 56 72 88
KILOMETERS
I’
EAST WEEK 3 : 20 SEPTEMBER EAST
: *
3
*10
TRIP 1
TRIP 2
1 "
mv maw m. nx mr cm
■ ■ — ■
X
MV HAW m. PU MT CM
96 11212B144 160 176 192 96 112129 144 160 176192
KILOMETERS
«»)
$ I
2
0
I *
s
Z 4
*
G *
z
-> 0
o
* 10
KILOMETERS
EAST WEEK 7 :14 OCTOBER
EAST
TRIP 1
TRIP 2
T"'1
X X
GMV HAW m. FIX M8T CM
â–  " "
X X
3MV HAW m. PLK H8T CPB
96 11212814416017619296 112126144160176192
KILOMETERS
9)
Z o
5
> 2
m -
G 6
O 8
<
10
KILOMETERS
KILOMETERS

61
Profile estimates for the gulf fritillary migration during fall of 1987 are shown
in Fig. 3-9 a,b. All net movement was southward and the migratory peak was recorded
during week 7 (Fig. 3-9b, west) along the TRN-NWB segment at nearly 15 indiv/min.
This increase in numbers was seen one year and a day after the extraordinarily large
numbers of Phoebis sennae were noted along the Gulf coast in 1986 (Fig. 3-4g).
Numbers of individuals seen along the eastern transect remained low and the only
segments with significant movement southward were the central areas of GNV-HAW and
HAW-INT.
Profile estimates of gulf fritillary migration during fall of 1988 are shown in
Fig. 3-10 a,b,c,d. Each weekly sample is a round trip during the same day and most of
the patterns were similar between trips 1 and 2, except during week 3 of sampling (Fig.
3-10a). Although trip 1 of week 3 showed fairly uniform movement across the western
transect (all < 2 indiv/min), trip 2 peaked through TRN-NWB with > 10 indiv/min.
Other than the TRN-NWB segment on this day, STN-CRC had the next largest number
of individuals, during week 3 and week 5 (Fig. 3-10c, trips 1 and 2). The Atlantic coast
had relatively low numbers of migrants (< 2 indiv/min) during week 4 (Fig. 3-10b), but
in week 6 showed an increase with peaks at HAW-INT (trip 1) and INT-PLK (trip 2).
Yearly migration profiles of Agraulis vanillae movement during the fall of 1986,
1987 and 1988 are shown in Fig. 3-11 a,b,c. The profile for 1986 (Fig. 3-1 la) peaked
through the INT-PLK segment (7 indiv/min) with the next peak at TRN-NWB (4.8
indiv/min). During 1987 (Fig. 3-1 lb), the TRN-NWB transect segment also contained
the migration peak at 7.5 indiv/min. This rate was nearly three times the number

62
a)
>
O
*
H-
LU
Z
“9
Q
<
b)
>
Q
Z
4»
h
lil
Z
”3
o
<
WEST
WEEK 4
WEEK 6
EAST
o
1 -
2 -
3 -
4
26 SEP
11
STN CRC TRN
NWB
GK V
4 OCT
XXX
HAW ITL PLK HST CRB
“i 1 1 1 1 1 1 1 1 1 1 1 r
0 16 32 48 64 80 96 112 128 144 160 176 192
JN
ts
WEST
KILOMETERS
WEEK 7
WEEK 8
EAST
17 OCT
6
10
15
18 OCT
STN CRC TRN â„¢ NWB GNV HAW ITL PLK HST CRB
i ! i i i r
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-9 a,b.-Profile estimates of Agraulis vanillae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65°) from 26 September to 18 October
1987. Counts were made on nine transect segments defined by 10 sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.

WEST
N
â– f
S
KILOMETERS
WEST
WEEK 5 : 20 SEP
0
2
4
5
s
10 k
TRIP 1
I
I I
TRIP 2
I â– 
laptepc , TRM NYyBWH'pTW CBC TRW KWB ONVI
0 16 32 48 64 80 96 0 16 32 48 64 80 06
KILOMETERS
b)
I
a
z
*
tü
â–¡
<
<»)
EAST
WEEK 4
: 25 SEP
0
TRIP 1
TRIP 2
2
â– I"
1
4
6
8
10
INVHAWm PLK HST CRB
JNV HAW m. PLK HST CRG
112120136152168184200112128144160176192
KILOMETERS
EAST
WEEK 6:12 OCT
+
KILOMETERS
Fig. 3-10 a,b,c,d. Profile estimates of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day,
sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65°) from 18 September to 12 October
1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthome, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.
x=unsampled sites.

Fig. 3-11 a,b,c.--Yearly migration profiles of Agraulis vanillae movement across a west-
east transect at the latitude of Gainesville (29.65°). The yearly migration profiles are
from the median estimates (net number of individuals/minute, adjusted for time of day,
sample size and season) during 13 September-28 October 1986,26 September-18 October
1987 and 18 September-12 October 1988. Counts were made on nine transect segments
defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen,
PLK=Palatka, HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically
significant (chi-square, p < .05)southward movement and open bars were sites with no
statistical bias (chi-square, p < .05) in net movement. Number of visits to each site is
above bars.

65
b)
KILOMETERS
1987
0 16 32 48 64 80 96 112128144160176192
KILOMETERS
1988
STN CRC TRÑ NWB GNV HAW ITL PLK HST CRB
i i "T 1 T r“i “T—~—"T 1 — i " i
0 16 32 48 64 80 96 112128 144 1 60 1 76 1 92
KILOMETERS

66
recorded at the next highest peak, through CRC-TRN (2.7 indiv/min). All southward
movement was statistically significant (chi-square, p < .05) except at PLK-HST and
HST-CRB during 1987, where numbers were small. During 1988, the migration peak
occurred along the Gulf coast through STN-CRC, but at lower numbers than previous
years (approx. 4 indiv/min). During all three years, the lowest numbers of individuals
moved southward along the east coast.
Numbers of Migrants
A migration profile summarizing the results of 1986, 1987 and 1988 profiles of
Phoebis sennae migration is shown in Fig. 3-12. All net movement was southward
throughout the fall sampling period (13 Sept-28 Oct). The transect segment with the
greatest numbers of cloudless sulphurs moving south was HAW-INT at 2.7 indiv/min,
more than double any other segment. The migration rate was fairly even across the other
transect segments, averaging about 1 indiv/min, except for PLK-HST (.34 indiv/min) and
HST-CRB (.18 indiv/min), where numbers declined approaching the Gulf coast.
The migration profile summarizing Agraulis vanillae movement for 1986, 1987
and 1988 is given in Fig. 3-13. Peak migratory movement for the gulf fritillary was
through the TRN-NWB segment at 4.8 indiv/min. The other peak was along the west
coast segment, STN-CRC at 3.9 indiv/min. In the case of the gulf fritillary, all three
segments with peak movement were along the west coast. Migration east of Newberry
was low in general, but the transect segments with the least movement were PLK-HST
and HST-CRB, at .81 and.84 indiv/min respectively.

67
1986-1988
Fig. 3-12.--A migration profile summarizing Phoebis sennae migration across a west-east
transect at the latitude of Gainesville, Florida (29.65°). The profile is derived from a
median estimate (net number of individuals/minute, adjusted for time of day, sample size
and season) at each site during 13 September-28 October 1986, 26 September-18 October
1987 and 18 September-12 October 1988. Estimates were made along nine segments of
a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthorne, ITL=Interlachen,
PLK=Palatka, HST=Hastings, CRB=Crescent Beach.

68
O On i i i i" i i' i i i i i 1 r
< O 16 32 48 64 80 96 112 128 144 160 176 192 (KM)
1986-1988
Fig. 3-13.—A migration profile summarizing Agraulis vanillae migration across a west-
east transect at the latitude of Gainesville, Florida (29.65°). The profile is derived from
a median estimate (net number of individuals/minute, adjusted for time of day, sample
size and season) at each site during 13 September-28 October 1986, 26 September-18
October 1987 and 18 September-12 October 1988. Estimates were made along nine
segments of a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City,
TRN = Trenton, NWB = Newberry, GNV = Gainesville, HAW=Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.

69
Walker (1991) estimated the size of the seasonally recurring migration through
peninsular Florida by extrapolating his flight trap catches from Gainesville using a
preliminary estimate of the migration profile from this study. His estimate of the net
movement of migrating butterflies each fall was an average of 22.2 million individual
cloudless sulphurs and 41.3 million gulf fritillaries. The combined figure of 63.5 million
butterflies approaches the estimated number of monarchs reported to overwinter at
Mexican sites (Brower, 1985; Calvert gt ai, 1979). The average rate of net movement
for Phoebis sennae and Agraulis vanillae (1986-1988) from the migration profile of each
species was used to recalculate Walker’s (1991) data (Table 3.1). Walker’s (1991)
Table 3.1 Estimates of the Phoebis sennae and Agraulis vanillae fall migration into the
Florida peninsula (Lenczewski, 1992 and Walker, 1991). Average absolute net numbers
of individuals flying across a north central Florida transect at the latitude of Gainesville
(29.65°).
Mean profile height (Lenczewski, 1992)
££
1.03
£E
2.10
NWB-GNV profile height
1.03
1.43
Profile correction factor
1.00
1.47
Kilometers across state (Lenczewski, 1992)
190.00
190.00
Migrants trapped/6m (Walker, 1991)
800.00
868.00
Est. migrants into cen. Fla. (millions)
25.30
40.40
Est. trapping efficiency (Walker, 1985b)
0.60
0.35
Adj. migrants trapped/6m
1333.00
2480.00
Est. migrants into C. Fla. (millions)
42.20
115.00

70
calculations were based on an assumption that the average rate of migration across the
transect was no less than 60% of the rate at the longitude of his trapping site in
Gainesville. In actuality, for Phoebis sennae. his site fell within the NWB-GNV segment
of the transect which had a rate of 1.03 indiv/min, identical to the transect average.
Walker also used a conservative transect length of 170 km, whereas the actual transect
length used in this study is 190 km. Walker’s (1991) Gainesville trap yields of 800
individuals/6 m for Phoebis sennae were applied to these data resulting in an estimated
influx into central Florida of 25.3 million individuals. With further adjustment for
Walker’s (1985b) 60% trap efficiency, the size of the cloudless sulphur fall migration
into central Florida was estimated as 42.2 million individuals, nearly double the figure
given by Walker (1991).
These calculations were repeated using the migration profile information for
Agraulis vanillae. also shown in Table 3.1. The average migration along the latitude of
Gainesville was 2.10 indiv/min, as compared to the NWB-GNV segment, which had a
rate of 1.43 indiv/min, resulting in a profile correction factor of 1.47. After applying
Walker’s trap catch of 868 individuals/6m and an efficiency of 35%, the final results
were 115 million gulf fritillaries, more than double Walker’s (1991) estimate of 41.3
million. The total number of these two species of butterflies migrating south through
north central Florida in the fall was estimated at 157 million individuals annually. This
number does not include the buckeyes or long tailed skippers, which were estimated by
Walker as numbering 3.7 and 14.6 million, respectively. Since profiles were not
established for these species from drive counts, Walker’s estimates cannot be evaluated.

71
Using his numbers for these other two species, the total migration of these four principle
migrants is given as 175 million individuals annually.
Fall 1987 drive counts of Phoebis sennae net migration across west-east transects
south of Gainesville, Florida, during fall 1987 are given in Fig. 3-14 a,b,c,d,e,f. The
net numbers observed in these drive counts were adjusted only for time of day. Visits
to the same transects were approximately 4 weeks apart and week 9 was the only time
that the full north-south range of transects was travelled. Statistically significant
southward movement is evident south to the Crystal River-Leesburg transect (3) during
week 4 (Fig. 3-14a) and continued through week 9 (Fig. 3-14c). It is not known whether
individuals are continually moving southward throughout this time or if this is a
secondary wave of migrants from subsequent generations. From the pattern observed at
Gainesville, there was probably a continual, and gradually diminishing, net southward
movement throughout this period. By week 15 (Fig. 3-14f), this northern area no longer
showed any significant migration. The only time statistically significant migration was
noted south of Crystal River-Leesburg was during week 7 at the Zolfo Springs-Sebring
transect. As is seen with results of trap catches (discussed in Chapter 4), there was no
evidence of large scale migratory movement through south central Florida. Unless the
migratory track deviates from the mean flight direction observed at Gainesville, the
migration appears to diminish south of Leesburg. Small numbers of individuals may be
moving farther south, perhaps with a second generation augmenting populations around
the Zolfo Springs-Sebring area. Drive counts made during fall 1987 depicting the net
migration of Agraulis vanillae net migration across various west-east transects south of

Fig. 3-14 a,b,c,d,e.f.—Drive counts of Phoebis sennae migration (net number of
individuals/minute, adjusted for time of day) across various west-east transects south of
Gainesville, Florida during fall 1987. Transects with a southward or northward net
migration of statistical significance (chi-square test, p < .05), are shown with half filled
circles weighted in the appropriate direction. Net movement, either north or south, that
is not statistically significant (p < .05), is depicted with open circles. Lines with no
circles represent transects where there were no individuals seen. Number of individuals
and percentage flying south are given in parentheses after each transect.

73
4. WILD WOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
E. BROOKSVILLE-HILL N DALE
7. BAYONET POINT-SAN ANTOMO
8. WESLEY CHAPEL-ZEPHRHILLS
#. LAKE LAND-HA INES CITY
10. BARTOW-LAKE WALES
11. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIDGE-LAKE PLACID
o NORTH BIAS
O o NO BIAS
GNV LAT (29.65)
NONE SEEN
7 MM =20 KM

74
Gainesville, Florida are shown in Fig. 3-15 a,b,c,d,e,f. A pattern similar to P. sennae
is seen for this species which also demonstrated a statistically significant southward
movement from Inglis-Belleview through Crystal River-Leesburg during week 4.
Subsequently, the farthest significant southward movement seen occurred during week
7 through Bartow-Lake Wales. Again, since the more northern transects were not
sampled at this time, it is not known if southward migration continued there during this
period, but presumably it did. Significant migration was still seen through the north
central transects during week 9 and, from the patterns seen along the Gainesville transect
(Fig. 3-11), it is likely that this continued. No individuals were observed on any trips
south of Gainesville (Fig. 3-15 d,e,f) after this time and Bee Ridge-Lake Placid also
showed no significant migration throughout the season for this species.
The area known as the Green Swamp is located between transects 5 and 7. It
is a large area maintained as wilderness, with some grazing areas, for cattle. The wet
areas along the Withlahochee River support a variety of flowering plants, especially in
the disturbed habitats where cattle are allowed to graze. Large numbers of migrants have
been sighted there during some years, particularly in late November (H. Nigg, pers.
com.). This largely unexplored area may contain many resources and provide an ideal
environment for adult winter survival. There is evidence from Gainesville (Chapter 4),
that some cloudless sulphur individuals will remain in an area with nectar sources for
several weeks.

Fig. 3-15 a,b,c,d,e.f.—Drive counts of Agraulis vanillae migration (net number of
individuals/minute, adjusted for time of day) across various west-east transects south of
Gainesville, Florida during fall 1987. Transects with a southward or northward net
migration of statistical significance (chi-squared test, p < .05), are shown with half filled
circles weighted in the appropriate direction. Net movement, either north or south, that
is not statistically significant (p < .05), is depicted with open circles. Lines with no
circles represent transects where there were no individuals seen. Number of individuals
and percentage flying south are given in parentheses after each transect.

76
4. WILDWOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
6. BROOKSVILLE-HILL N DALE
7. BAYONET POINT-SAN ANTON»
B. WESLEY CHAPEL-ZEPHRHILLS
>. LAKELAND-HAINES CITY
10. BARTOW-LAKE WALES
11. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIOGE-LAKE PLACID
d * SOUTH BIAS
© © NORTH BIAS
O O NO «AS
GNVLAT (29.65)
NONE SEEN
7 MM =20 KM

CHAPTER 4
PHENOLOGY OF MOVEMENT
Introduction
Many migratory insects, such as aphids and locusts, are greatly dependent on
wind and weather conditions to direct their flight once they are airborne (Pedgeley, 1982;
Taylor, 1958). These insects, as well as nocturnal moths, fly long distances at high
altitudes requiring the use of sophisticated, expensive technology, such as radar, for
study (Rainey, 1951; Schaefer, 1976) or extensive mark-release-recapture programs (Li
et al, 1964; Showers eí al, 1989). Butterflies are particularly suitable as subjects for the
study of insect migration because they are daytime fliers and move within a variable
boundary layer, usually within 4-6 meters of the ground (Edwards and Richman, 1977).
Wind speeds within this layer are low enough to allow individual control of flight
direction (Johnson, 1969). As the butterflies move near ground level, an observer can
identify species as well as note behavior and preferred flight tracks. Migrating butterflies
typically continue their flight on a linear track and will fly up and over obstacles rather
than around (Williams, 1930). This behavioral characteristic can be exploited to
selectively capture migrants for research purposes and to quantify movement by using
flight traps. Butterflies also have large wing surfaces which make it relatively easy to
conspicuously mark individuals for recapture (Walker and Wineriter, 1981).
77

78
Despite all these advantages, and, probably because they are not agriculturally
significant pests, the often spectacular movements of these insects have been largely
ignored. A few workers (Arbogast, 1965, 1966; Balciunas and Knopf, 1977) have made
observations over short periods of time and space, but it has been difficult to
simultaneously monitor migrations over distance. Through a massive tagging campaign,
Urquhart and Urquhart (1976) succeeded in determining the destination of Danaus
plexippus (L.), the monarch butterfly, in the Mexican mountain ranges. The migrations
of the great southern white, Ascia monuste L., along the Florida east coast were detailed
by Nielsen and Nielsen (1952; Nielsen, 1961) over a period of at least 20 years. In
Germany, Roer (1959, 1961a, 1961b, 1962, 1968, 1969, 1970) has described the
movements of Agíais urticae L., Inachis io, Nvmphalis antiopa and other European
butterflies, while Baker (1968ab, 1969) has studied Pieris rapae in England. For more
than ten years, Walker (1978, 1980, 1985ab, 1991) has maintained directional flight traps
in Gainesville, Florida that continuously monitor the annual migrations of eight species
of butterflies through this area. This type of data collection eliminates observer bias,
time of day or seasonal influences and permits correlation with environmental factors
(Walker and Riordan, 1981). A continuous, long term documentation of migration in
this manner is unique and has provided reliable information about the species
composition, phenology and size of the migration, at least through Gainesville.
There are three species that comprise most of the migration through north
central Florida in the fall (Walker, 1991): Phoebis sennae (L.) (Pieridae), the cloudless
sulphur; Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; and Urbanus proteus

79
(L.) (Hesperiidae), the long-tailed skipper. Another important migrant in Florida is
Precis coenia Hübner (Nymphalidae), the buckeye, but most of its migration occurs
northward in the spring with the southward fall movement being relatively small. This
pattern is the reverse of what is seen for the other species listed. Together, these four
species make up more than 85% of Walker’s (1991) total trap catch. Walker (1991)
captured at least 100 individuals of five other species in his ten years of trapping. Pieris
rapae (L.) (Pieridae) and Vanessa virginiensis (Drury) (Nymphalidae) showed significant
northward movement in the spring, but were seldom caught in the fall. Eurema lisa
(Boisduval & LeConte) (Pieridae) had significant southward movement through
Gainesville in the fall, but was never caught in the spring. Another pierid, Eurema daira
(Godart) moved south in the fall of one year, but showed no statistical bias in either
direction over all years combined. Finally, the pierid, Eurema nicippe (Cramer) was
found in some years to move northward in the fall, a seemingly inappropriate direction.
The monarch passes primarily along the gulf coast to Texas in its migrations south
(Brower et ai, 1985) although, a few individuals do move through north central Florida
(Walker, 1991). Since this species flies at a higher altitude than others (Gibo, 1986),
they are seldom captured in flight traps.
Although the movements of these species have been well documented in
Gainesville by Walker, migratory activities in the rest of the state are virtually unknown.
In particular, the timing, or phenology, of the migration south of Gainesville is not
known, nor are the characteristics of the migratory front, as far as density and species
composition. Arbogast (1966) and Balciunas and Knopf (1977) determined the flight

80
speed of migrating gulf fritillaries and long tailed skippers to be approximately 12-22 kph
under ideal conditions. Weather permitting, peak migration times are usually 0900-1600
HRS LMT (Walker, 1985a) daily, suggesting that individuals may travel as much as 100
km/day. If the same individuals that pass through Gainesville during the peak of
migration move directly to lower latitudes, a similar pattern of peaks would be expected
at sites to the south. However, if peaks lag to the south by about 3 weeks, the
approximate generation time for gulf fritillaries (Arbogast, 1965, 1966), movement south
could be in a stepwise, generational manner.
Materials and Methods
Longitudinal Pole Counts
During 1986, pole counts (see description of methods in Chapter 3) were
initially made at Gainesville and five sites south, Clermont, Bartow, Arcadia, LaBelle
and Immokalee (see Chapter 1, Fig 1-1 for location of all sample sites). The primary
purpose of these counts was to get some information on when migration began and how
far it extended south of Gainesville before planning more extensive data collection with
traps. Counts were made from 13 September to 2 November 1986, at approximately two
week intervals (see Chapter 1, Table 1-1 for dates of all sampling). This period was
figured by Walker (1991) to include 95% of the total seasonal migration through
Gainesville. The six sites along the north-south transect were spaced approximately 62
km apart, a total distance of 372 km. All net numbers were adjusted as explained
previously (Chapter 3) for time of day differences (T) using Walker’s (1985a)

81
observations of daily flight periodicity. These adjusted net numbers were averaged to
yield a seasonal summary for each site along the north-south transect.
Portable Flight Traps
For the purposes of this study, a portable flight trap was designed (Walker and
Lenczewski, 1989) to quantify the phenology of butterfly migration along the Florida
peninsula. Walker (1985a) has determined that the migratory track through Gainesville,
Florida remained relatively constant at 141° annually. A migratory route was projected
north and south (321° and 141°) from Gainesville and pairs of these flight traps were
erected on it at five sites (an average of 105 km apart) from southern Georgia to south
central Florida: Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida. Migration was monitored with these traps through the fall of 1987 and 1988.
The traps were placed in pairs along an ENE-WSW line (67.5°-247.5°), one trap facing
north and one facing south. These traps, erected approximately perpendicular to the axis
of the Florida peninsula and to the mean direction of migration, were left in place from
13 September through 12 December 1987. Butterflies captured by the south-facing trap
were flying 321 ± 90°; those captured by the north-facing trap were heading 157.5 ±
90°. All traps were situated in large, open fields to eliminate the influence of vegetation
or other obstacles on the preferred flight tracks of migrants. As the migrant butterflies
tried to fly over a main wall of polyester screening, they passed through a narrow slot
into a duct that led to two holding cages made of 1/4 inch hardware cloth. Captured
butterflies were removed from the holding cages semiweekly throughout the trapping

82
period. Walker and Lenczewski (1989) estimated that semi-weekly service recovered at
least 90% of all individuals entering the traps’ collecting cages. After testing other
construction materials in the spring of 1988, the traps were modified, primarily to
increase durability. The backs of eight polyester traps were replaced with monofilament
shrimp netting. Two of the polyester netting traps were left intact. An additional four
traps were built of the same design but with all monofilament shrimp netting. All these
traps were found to be similar in efficiency (Walker and Lenczewski, 1989), but the
monofilament shrimp netting was more durable and easier to work with. In 1988, traps
were in place from 28 August through 26 November.
The sites were 157, 100, 70 and 92 km apart, respectively, a total of 419 km
from northernmost to southernmost site. The types of traps used each year, at each site,
are as follows:
1. Valdosta State College, Valdosta, Georgia (30.85° lat, 83.29° long).
1987 - two all polyester traps, 1988 - four polyester traps with shrimp
net backing.
2. Green Acres Farm, University of Florida, Gainesville, Florida (29.65°
lat, 82.44° long). 1987 - two all polyester traps. 1988 - no traps at this
location, data used from Walker (1991).
3. IFAS/AREC, University of Florida, Leesburg, Florida (28.90° lat,
81.54° long). 1987 - two all polyester traps, 1988 - four polyester traps
with shrimp net backing.
4. IFAS/CREC, University of Florida, Lake Alfred, Florida (28.00° lat,

83
81.44° long). 1987 - two all polyester traps, 1988 - four all shrimp net
traps.
5. Archbold Biological Station, Lake Placid, Florida (27.45° lat, 81.23°
long). 1987 and 1988 - two all polyester traps.
During the fall of 1987, two all polyester traps were in operation at all five
sites, one facing north and one facing south. In fall of 1988, four polyester traps with
shrimp net backing were used in Valdosta and Leesburg. Four all shrimp net traps were
placed at Lake Alfred, and the Lake Placid station had two all polyester traps similar to
those used in 1987. During 1988, all stations had two traps facing north and two south,
in an alternating pattern, except for Lake Placid, where one trap was positioned north
and one south. There were no traps placed at the Gainesville site in 1988 since similar
data were available from Walker’s (1985b) permanent flight traps. The dates for each
week sampled are as given in Chapter 1, Table 1-1.
Results and Discussion
Cloudless Sulphur
Longitudinal pole counts
Pole counts of Phoebis sennae net movement along a north-south transect from
Gainesville to Immokalee during 1986 are shown in Fig. 4-1 a,b,c,d,e,f. All statistically
significant (p < .05) net movement at the six sites was southward throughout the period
sampled. All three weeks sampled at Gainesville (a) showed significant southward
movement with the peak occurring during week 2. Of all the sites sampled, Gainesville

84
a)
GAINESVILLE
a)
LABELLE
b)
d)
Í J
g)
CLERMONT
WEEKS
ARCADIA
XX XX
10
WEEKS
WEEKS
Fig. 4-1 a,b,c,d,e,f.--Pole counts of Phoebis sennae movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September through 7 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p <.05). x=no counts made.

85
had the greatest numbers moving south (1 indiv/m/min) during peak migration, well over
three times that of the next largest movement at Clermont (.25 indiv/m/min). At
Clermont (b), which is located approximately 136 kms south of Gainesville, the cloudless
sulphur did not achieve a statistically significant southward movement until five weeks
later, during week 7. The rest of the sites had no significantly directed movement,
except for LaBelle, where cloudless sulphurs also migrated southward (.2 indiv/m/min)
during week 7.
The summary of Phoebis sennae pole counts along the transect from Gainesville
south is shown in Fig. 4-2. Gainesville had the highest counts with an average of about
.6 indiv/m/min passing southward. Clermont and LaBelle, the only other sites with
significant southward movement, had a rate of about .1 indiv/m/min.
Flight trap catches
Weekly percentages of total net trap catch of Phoebis sennae during the fall of
1987 at five sites along a north-south transect through Gainesville are shown in Fig. 4-3
a,b,c,d,e. All statistically significant net movement was southward for this species,
falling between weeks 4 and 12. The early weeks were not sampled completely at any
of the sites, and the week in which half of the total net catch was achieved is probably
not a reliable estimate of mid-migration point. The Gainesville data (b) were corrected
for this by using information taken from Walker’s (1991) permanent traps near the same
location, shown in Fig. 4-4c. In the case of the cloudless sulphur, the percent total net
catches for weeks 5-13 were reduced by 33% (Walker’s trap catch during the first 4
weeks). The appropriate adjustment was applied in a similar manner for the other three

86
Fig. 4-2.-Summary of pole counts of Phoebis sennae (average net number of
individuals/meter/minute, adjusted for time of day) through Gainesville, Clermont,
Arcadia, LaBelle, and Immokalee, Florida during 11 September through 8 November
1986. Numbers below the bars represent total number of counts at that site.

87
a)
VALDOSTA
b)
GAINESVILLE PTP#&6 WALKER! 1991)
•)
i a a 4 •• 7 •• io ii u 1* 14 if
WEEKS
LAKE ALFRED
GAINESVILLE
1 2 3 4 6 • 7 6 ■ 10 11 12 13 14 16
WEEKS
d)
f)
If 14
WEEKS
LEESBURG
10 11 12 It 14 If
WEEKS
LAKE PLACID
i 2 3 4 a a 7 a a io 1112131416
WEEKS
Fig. 4-3 a,b,c,d,e,f.--Weekly percentage of seasonal total net trap catch of Phoebis
sennae at Valdosta, Georgia, Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida in flight traps during 13 September through 12 December 1987. Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05),
shaded bars show nonsignificant (p > .05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right corner. The midpoint of
migration is designated by an asterisk, x=trap not operating. Total percent migration
in b) was reduced by 33% on the basis of data in c).

88
species. The estimate for mid-migration was either week 5 or 6 at all sites. The close
coordination of mid-migration seems to indicate that the same wave of individuals passed
through all five areas. The most southern site, Lake Placid, had no net movement, and
even though several individuals were seen and two were trapped, there was no indication
of migration. In fact, the numbers of cloudless sulphurs were surprisingly low at all
other sites, as compared to Gainesville. The reason for low numbers at Valdosta at least
could be that since the first three weeks were not sampled, most of the migrants may
have passed southward earlier.
The percentage of total net catch of Phoebis sennae captured at the five sites in
1988 are shown in Fig. 4-4 a,b,c,d,e,f,g,h,i. Gainesville data is taken from Walker
(1991). All statistically significant movement was southward biased. At Valdosta (Fig.
4-4a), for the cloudless sulphur, catches in traps 1 and 2 were not well coordinated.
Trap 2 went out of commission after week 6 and the percentage of total catch was
adjusted in relation to the first six weeks of trap 1 (i.e. decreased by 40%). During
1988, traps were in place by week 1, at which time there was no evidence of migration.
The mid-migration point for trap 1 was during week 6, and the only other significant
southward movement was seen during week 7. Southward migration occurred during the
second week for trap 2, but sample size was small and the trap may already have been
damaged. Walker’s (1991) data from the two permanent traps located in Gainesville
(Fig. 4-4 c,d) both show mid-migration occurred during week 7. The two traps at
Leesburg and the average of the two traps at Lake Alfred all suggest a midpoint during
week 8 at those sites. The only significant southward movement seen from catches at

Fig. 4-4 a,b,c,d,e,f,g,h,i.--Weekly percentage of seasonal total of net trap catch of
Phoebis sennae at Valdosta, Georgia, Leesburg, Lake Alfred and Lake Placid, Florida
during 28 August-26 November 1988. Data for Gainesville, Florida are from Walker
(1991). Solid bars represent statistically significant southward biased movement (chi-
square test, p < .05), shaded bars show nonsignificant (p > .05) southward movement.
Clear bars depict nonsignificant (p > .05) northward movement. The total net catch
(southbound minus northbound) for the period is shown in the upper right comer. The
midpoint of migration is designated by an asterisk, x=trap not operating. Total percent
migration in b) was reduced by 40% on the basis of data in a).

CO
CD
o
0
% TOTAL NET CATCH % TOTAL NET CATCH
% TOTAL CATCH
% TOTAL NET CATCH
% TOTAL NET CATCH
% TOTAL NET CATCH ~ % TOTAL NET CATCH
A TRAP 1 D) VALDOSTA TRAP 2

91
Lake Alfred trap 2 was during week 11. At these four sites, from Valdosta to Lake
Alfred, the mid-point of migration for traps 1 and 2 fell between weeks 6-9 and it is
found almost uniformly (except for Lake Alfred, trap 2) at a one week difference
between sites. Lake Placid had no statistically significant bias in movement at any time
during the trapping period, although five individuals were trapped and others were seen.
The percentage of total net trap catches of cloudless sulphurs are summarized
for both the 1987 and 1988 fall migration seasons in Fig. 4-5 a,b,c,d,e. The 1988 trap
1 and 2 data were tested for homogeneity and combined before calculating the mean
percent total net catch for both years. The first four weeks of Gainesville 1987 and all
of 1988 data used in this summary are from Walker’s (1991) permanent trap catches as
shown in Figs. 4-3c and 4-4 c,d. The migration midpoint occurred between weeks 5
through 7 at all of the four main sites. Lake Placid had no evidence of migration with
a total net of only one individual heading south in two years. The net numbers captured
at Valdosta (Fig. 4-5a) were surprisingly low, totaling only 70 individuals in two years,
and similar to catches at Lake Alfred (Fig. 4-5d). During 1987 alone, Gainesville traps
(Fig. 4-3b) captured nearly three times that number. The total net number at Leesburg
for the two years was 275. This was unexpected since it was assumed that most
individuals begin travelling south through Valdosta and then move through Gainesville.
During 1987, the earlier part of the migration may have been missed, but in 1988 the
pattern was similar with comparable numbers at Gainesville-Leesburg and Valdosta-Lake
Alfred. Subsequent breeding could augment the migratory stream between Leesburg and
Gainesville, accounting for the larger numbers caught at these sites. The sites south of

12346678 8 101112131416
WEEKS
1 2 3 4 6 8 7 8 8 10 1112131416
WEEKS
C)
LEESBURG
WEEKS
d)
LAKE ALFRED
WEEKS
e)
LAKE PLACID
WEEKS
Fig. 4-5 a,b,c,d,e.--Summary of percent total net trap catches of Phoebis sennae at five
sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred
and Lake Placid, Florida) during 30 August-12 December 1987 and 28 August-26
November 1988. Solid bars represent net southward movement and clear bars show net
northward movement. Total net catch (southbound minus northbound) is shown in the
upper right comer and x=unsampled weeks. Data for the first 4 weeks of 1987 and all
of 1988 Gainesville are from Walker (1991).

93
Valdosta have milder winter temperatures and the population build up during summer
months may be much larger than at more northern sites, resulting in more winter
survival. Cloudless sulphurs are commonly seen in Gainesville on warm days throughout
the winter, but do seem affected by freezing temperatures which would be more likely
in Valdosta. Following a particularly severe freeze in Gainesville on 23 December 1989,
this species was not seen again until 18 February the following year. This was despite
unusually high temperatures after the hard freeze.
Gulf Fritillarv
Longtiydinai pQig counts
Pole counts of Agraulis vanillae net movement during 1986 along six north-south
sites from Gainesville to Immokalee, Florida are shown in Fig. 4-6 a,b,c,d,e,f. All net
movement at the sites was southward during the period sampled. Of the weeks sampled,
the peak statistically significant migration (chi-square, p < .05) for this species occurred
during week 6 in Gainesville at approximately .55 indiv/m/min. The only other site
with significant movement was the next site, Clermont, about 180 kms south of
Gainesville, with .42 indiv/m/min. There were no individuals of Agraulis vanillae
counted at either LaBelle or Immokalee during the sampling period.
The pole count summaries of gulf fritillary net movement at six sites from
Gainesville to Immokalee is shown in Fig. 4-7. Like the cloudless sulphur, all net
movement was southward for the gulf fritillary. The greatest numbers of individuals
travelled south through Gainesville at an average rate of nearly .4 indiv/m/min.
Numbers of individuals moving through the more southern sites of Clermont and Bartow

4 6 6
WEEKS
BARTOW
ARCADIA
IMMOKALEE
Fig. 4-6 a,b,c,d,e,f.--Pole counts of Agraulis vanillae movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p <.05). x=no counts made.

95
z
.6
1
.4
(/>
.2
<
D
O
0
>
Q
.2
Z
.4
h*
LU
Z
.6
GNV CLM BRT ARC LBL IMK
LOCATION
Fig. 4-7.-A summary of pole counts of Agraulis vanillae (average net number of
individuals/meter/minute, adjusted for time of day) through Gainesville, Clermont,
Bartow, Arcadia, Labelle and Immokalee, Florida during 11 September-8 November
1986. Numbers below the bars represent total number of visits to that site.

96
decreased greatly and there was no significant migration recorded south of these areas.
Flight trap catches
The weekly percentages of seasonal total net trap catch of Agraulis vanillae
along a north-south transect through five sites from Valdosta, Georgia to Lake Placid,
Florida during fall 1987 are shown in Fig. 4-8 a,b,c,d,e,f. All statistically significant
(chi-square, p < .05) net movement is southward. The midpoint of migration occurred
in Valdosta (Fig. 4-8a) no later than week 5. The lack of data for the early weeks has
probably shifted the midpoint forward and there has been no adjustment made for this.
The pattern seen from the two Gaineville sites suggest a buildup before the mode; thus
the Valdosta mode during week 4 may also represent the median. Significant southward
movement occurred at Valdosta as late as week 11. With the same trapping effort,
significant southward migration occurred during weeks 5, 6 and 7 at Gainesville (Fig.
4-8b) with mid-migration seen from Walker’s (1991) traps (Fig. 4-8c) as occurring
during week 5. Net numbers of gulf fritillary individuals were comparable in Valdosta
and Gainesville, numbering 72 and 69, respectively, unlike the large discrepancy
described for the cloudless sulphur at these sites. The Valdosta mid-migration point of
week 4 or 5 and Gainesville mid-migration point of week 5 suggest that the same
individuals may travel through south Georgia to north central Florida in a timely manner.
It is also possible that throughout this general area, the initiation of migratory flight is
coordinated. Although numbers captured in Valdosta and Gainesville were similar, the
net movement at Leesburg (Fig. 4-8c) seemed low. In actuality, the total numbers
captured there, 102, were higher than any other site, but there was almost equal

97
a)
VALDOSTA
b)
C)
WEEKS
GAINESVILLE PTP #3 (WALKER, 1991)
50 i
e)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 13
WEEKS
LAKE ALFRED
GAINESVILLE
d)
I 140
-
t 120
O 100
I 80
-J 60
% TOT/
ro £
o o o
UL
WEEKS
LEESBURG
12l4 WEEKS
LAKE PLACID
11 12 11 1* 11
2 3 4 S « 7 • 8 10 11 12 13 14 IS
WEEKS
WEEKS
Fig. 4-8 a,b,c,d,e,f.--Weekly percentage of total net trap catch of Agraulis vanillae at
Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida during
13 September-12 December 1987. Solid bars represent statistically significant southward
biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p > .05)
southward movement. Clear bars show nonsignificant (p > .05) northward movement.
The total net catch (southbound minus northbound) for the period is shown in the upper
right comer. The midpoint of migration is designated by an asterisk, x=traps not
operating.

98
bidirectional movement throughout the season. The midpoint of migration at Leesburg
was also during week 5 and this was the only week of statistically significant net
movement (biased southward). During visits to this site throughout the collecting period,
Passiflora incamata. a larval hostplant, was noted as abundant in the surrounding
cultivated crop fields. Gulf fritillaries, especially females, were observed visiting and
ovipositing on these plants. The presence of these hostplants may have affected flight
trap catches by diverting individuals to those sites and the butterflies may have remained
longer, taking advantage of breeding sites. It was seen from captures at Lake Alfred
(Fig. 4-8d), some 70 kilometers south of Leesburg, that the southward migration
continued well into week 12 of the season although the median was in week 8. The
favorable breeding conditions at Leesburg may have caused the 2-3 week delay seen at
Lake Alfred. There also may have been a high population of residents (non-migratory)
at Leesburg which could have been travelling short distances and were also trapped.
This may have resulted in the trapping pattern seen at Lake Alfred. From these data,
it appears that migration through Valdosta, Gainesville and Leesburg (a distance of 265
km) occurs during weeks 4 or 5, but the movement does not reach Lake Alfred, only 90
km south of Leesburg until 2-3 weeks later. Using an extremely conservative estimate
of travel time, individuals that fly south through Gainesville would be expected to reach
Lake Alfred within two weeks. According to estimates of adult longevity for the gulf
fritillary by Arbogast (1965), unmated adults had a mean life span of 18.4 ± 3.0 days
in captivity. It seems reasonable to assume that in nature, subjected to various stresses,
the normal life span may be shorter. Arbogast’s (1965) estimates of the developmental

99
rate from egg to adult at 28.5-29.5 °C for this species is approximately 23 days. The
observed situation at Leesburg may indicate that adult butterflies, presumably having
arrived from Gainesville north, have ceased migrating and are utilizing available
hostplants for oviposition. From known estimates of developmental rates and flight
speeds, it is possible that a second generation travels southward to Lake Alfred.
Although a few individuals were seen at Lake Placid, there was no significant migration
recorded during the fall and only one individual was captured heading south during week
14.
The percent of the total net trap catch of gulf fritillaries during fall 1988 is
shown in Fig. 4-9 a,b,c,d,e,f,g,h,i. As seen in 1987, all statistically significant (chi-
square, p < .05) net movement throughout the season was southward at each site. Also
as seen the previous year, mid-migration at Valdosta trap 1 (Fig. 4-9a) was during week
5 and this was also the peak for trap catches during the 6 weeks sampled with trap 2
(Fig.4-9b). Walker’s (1991) data during 1988 are given for Gainesville (Fig. 4-9 c,d)
and the midpoint is shown as week 6. Leesburg traps 1 and 2 (Fig. 4-9 e,f) had
corresponding midpoints also during week 8. Whereas in 1987, the only southward
biased movement was during week 5, 1988 catches from traps 1 and 2, respectively,
revealed a full 8 and 4 weeks of significant net southward movement. Movement during
1988 through Leesburg was less bidirectional than the previous year and the migration
midpoints were much closer to the values for Lake Alfred (Fig. 4-9 g,h). There was still
a three week difference in the midpoint of migration between Valdosta and the mid¬
central site of Lake Alfred, and this year, also Leesburg. Again, this spacing of

Fig. 4-9 a,b,c,d,e,f,g,h,i.-Weekly percentage of total net trap catch of Agraulis vanillae
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August
through 26 November 1988. Data for Gainesville, Florida are from Walker (1991).
Solid bars represent statistically significant southward biased movement (chi-square test,
p < .05). Shaded bars show nonsignificant (p > .05) southward movement. Clear bars
show nonsignificant (p > .05) northward movement. The total net catch for the period
is shown in the upper right comer. The midpoint of migration is designated by an
asterisk, x=traps not operating.

% TOTAL NET CATCH % TOTAL NET CATCH % TOTAL CATCH % TOTAL NET CATCH
101
a)
VALDOSTA TRAP 1
VALDOSTA TRAP 2
WEEKS
GAINESVILLE PTP3 (WALKER,1991)
WEEKS
LEESBURG TRAP 1
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
LAKE ALFRED TRAP 1
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
d)
WEEKS
GAINESVILLE PTP5 (WALKER,1991)
50 i
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
LEESBURG TRAP 2
WEEKS
LAKE ALFRED TRAP 2
h)
LAKE PLACID
1 2 3 4 5 6 7 8 9 10111213
WEEKS
Ox50
o
t 40
o
t30
z
H
*
-
*
f\
N=6
X
*
1 2 3 4 5 6 7 8 9 1011 12 13
WEEKS

102
migration waves could represent generational movement. There was no evidence of
significant movement through Lake Placid (Fig. 4-9i).
The summary of percent total net trap catches of Agraulis vanillae individuals
along the north-south transect during fall 1987 and 1988 is shown in Fig. 4-10 a,b,c,d,e.
All statistically significant (chi-square test, p < .05) net movement was southward.
There was a time span of three weeks in the migration midpoints from Valdosta, in south
Georgia, to the south central Florida site of Lake Alfred. This lag fits more closely the
hypothesis of synchronous initiation of movement, or travel by the same individuals,
rather than successive generations. Migration midpoints occurred in Valdosta and
Gainesville during week 5, Leesburg during week 6 and Lake Alfred during week 8 and
the migration lasted at least 9 weeks at all sites. Significant net movement ended at
Valdosta by week 12 and at Leesburg by week 13. Movement south at Lake Alfred and
Gainesville continued at least through weeks 12 and 13, respectively. There was no
evidence of significant net movement of gulf fritillaries through Lake Placid. Although
only seven individuals were caught throughout the two years of trapping, all were headed
south. Over the two year period, this would constitute a significant southward
movement.
Long Tailed Skipper
Longitudinal pole counts
The pole counts of net movement Urban us proteus net movement along a north-
south transect from Gainesville to Immokalee, Florida during 1986 are shown in Fig. 4-
11 a,b,c,d,e,f. The southward movement for this species at Gainesville was similar to

103
a)
VALDOSTA
C)
WEEKS
LEESBURG
WEEKS
b)
d)
GAINESVILLE
,x x
123466789 101112131416
WEEKS
LAKE ALFRED
12346878 9 10 1112131415
WEEKS
e)
12346678 9 10 1112131416
WEEKS
Fig. 4-10 a,b,c,d,e. Summary of percent total net trap catches of Agraulis vanillae at
five sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake
Alfred and Lake Placid, Florida) during 13 September-12 December 1987 and 28
August-26 November 1988. Solid bars represent net southward movement and clear bars
show net northward movement. Total net catch is shown in the upper right comer and
x=trap not operational. Data for the first 4 weeks and all of 1988 Gainesville are is
from Walker (1991).

104
c)
e)
Z-2
to
b
i -6
D o
z 8
* 1
tjl.2
BARTOW
l
1
WEEKS
*
Fig. 4-11 a,b,c,d,e,f. Pole counts of Urbanus proteus movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p < .05), x=no counts made.

105
the gulf fritillary and cloudless sulphur. However, the only statistically significant
migration of the season was through Gainesville during weeks 6 and 8. The peak rate
there was during week 6, at about 1.1 indiv/m/min, which was nearly twice the peak rate
of the gulf fritillary and slightly more than the cloudless sulphur. Although there was
net movement at all other sites, mostly southward, none of it was significant.
The summary of net movement of H. proteus observed during pole counts along
the north-south transect in 1986 is shown in Fig. 4-12. All net movement is southward,
but the only statistically significant (chi-square, p < .05) bias was through Gainesville
and Clermont, the two most northern sites. The numbers moving through Gainesville
averaged about .6 indiv/m/min and those at Clermont about, .2 indiv/m/min. The lack
of significantly biased directional movement south of Clermont is probably a result of
low sample numbers, since subsequent catches in flight traps revealed southward
movement through Lake Alfred.
Flight trap catches
The percent total net trap catch of Urbanus proteus along a north-south transect
from Valdosta, Georgia, to Lake Placid, Florida, during 1987 is shown in Fig. 4-13
a,b,c,d,e,f. All statistically significant (chi-square, p C.05) net movement for this
species was southward and was completed by weeks 5 or 6 at all sites. Valdosta (a),
Gainesville (b) and Leesburg (d) had surprisingly similar numbers (net = 37, 37 and 45)
of individuals captured. Midpoints at all sites, except Lake Placid, were during week
4 or 5 (Fig. 4-13 a,c,d,e). Numbers of individuals trapped at Lake Alfred (Fig. 4-13 e)
were very low (n=9), with no significant movement shown for any week throughout the

106
Z .6
2
GNV CLM BRT ARC LBL ¡MK
LOCATION
Fig. 4-12. A summary of pole counts showing net movement of Urbanus proteus
(average net number of individuals/meter/minute, adjusted for time of day) through
Gainesville, Clermont, Bartow, Arcadia, Labelle and Immokalee, Florida during 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p < .05), x=no counts made. Numbers below the
bars represent total number of visits to that site.

107
a)
b)
VALDOSTA
GAINESVILLE
1 2 3 4 B 6 7 8 9 101112131416
WEEKS
C)
d)
GAINESVILLE PTP#3&5 (WALKER,1991)
2 3 4 5 • 7 8 8 10 11 12 13 14 16
WEEKS
LEESBURG
12 3 4
• 7 8 9 10 11 12 13 14 16
WEEKS
a)
LAKE ALFRED
1 2 3 4 6 6 7 8 9 101112131416
WEEKS
LAKE PLACID
WEEKS
1 2 3 4 6 8 7 8 9 10 11 12 13 14 16
WEEKS
Fig. 4-13 a,b,c,d,e,f.-Weekly percentage of seasonal total net trap catch of Urbanus
proteus at Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid,
Florida during 13 September-12 December 1987. Solid bars represent statistically
significant southward biased movement (chi-square test, p < .05), shaded bars show
nonsignificant (p > .05) southward movement. Clear bars show nonsignificant (p > .05)
northward movement. The total net catch for the period is shown in the upper right
comer. The midpoint of migration is designated by an asterisk. x=trap not operating.

108
period, but the net movement for the season was significantly southward. There were
no long tailed skippers captured in flight traps at Lake Placid (Fig. 4-13 e), although
several adults were observed feeding at flowers during each visit.
The percent total net trap catch of Urbanus proteus at five north-south sites
during 1988 is shown in Fig. 4-14 a,b,c,d,e,f,g,h,i. All statistically significant (chi-
square, p < .05) net movement was southward. Numbers of individuals captured at
Valdosta were small, with a total net catch of 16 for traps 1 and 2 (Fig. 4-14 a,b).
There was no statistically biased movement revealed by trap 1, but trap 2 had significant
(p <.05) net southward movement during week 5. Walker’s (1991) Gainesville data
indicates the mid-migration point at that site was during week 9 (Fig. 4-14 c,d). This
is also true for Leesburg, both trap 1 and 2 (Fig. 4-14 e,f). The average midpoint of the
two Lake Alfred traps (Fig. 4-14 g,h) was during week 10. South of Valdosta, where
significant net movement south ended after week 5, all sites sampled (except Lake Placid)
showed net southward movement through at least week 12. No long tailed skippers were
captured at Lake Placid (Fig. 4-14i).
The summary of percent total net trap catches of Urbanus proteus for 1987 and
1988 is shown in Fig. 4-15 a,b,c,d,e. All statistically significant (chi-square, p <.05)
net movement is southward. The average midpoint of migration for these two years in
Valdosta was during week 5. Gainesville migration midpoint occurred during week 7.
All movement stopped at Valdosta by week 9, but continued at low levels in Gainesville
through at least week 13. Leesburg migration midpoint was also during week 7 and
movement there ended by week 13. Lake Alfred had a bimodal pattern showing strong

Fig. 4-14 a,b,c,d,e,f,g,h,i.--Weeklypercentage of total net trap catch of Urbanus proteus
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August-
26 November 1988. Data for Gainesville, Florida are from Walker (1991). Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05).
Shaded bars show nonsignificant (p >.05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right comer. The midpoint of
migration is designated by an asterisk, x =trap not operating.

110
VALDOSTA TRAP 1
3 4 5 6 7 8 9 1011 1213
WEEKS
VALDOSTA TRAP 2
2 3 4 5 6 7 8 9 1011 1213
WEEKS
C) GAINESVILLE PTP3 (WALKER, 1991)
WEEKS
WEEKS
9) LAKE ALFRED TRAP 1
WEEKS
d) GAINESVILLE PTP5 (WALKER,1991)
WEEKS
WEEKS
") LAKE ALFRED TRAP 2
50
0 * * spa is^a
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
WEEKS

Ill
b)
GAINESVILLE
z 60
O
H 50
t 40
z 30
£ 20
° 10
123466789 101112131416
WEEKS
1 2 3 4 6 • 7 S 9 10 11 12 13 14 16
WEEKS
LAKE ALFRED
12346678 9 10 1112131416
WEEKS
e)
O
<
o
<
â–º-
o
I-
#
WEEKS
50
40
30
20
10
0
123466789 101112131416
LAKE PLACID
N-0
X
WEEKS
Fig. 4-15 a,b,c,d,e.-Summary of percent total net trap catches of Urbanus proteus at
five sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake
Alfred and Lake Placid, Florida) during 13 September-12 December 1987 and 28
August-26 November 1988. Solid bars represent net southward movement and clear bars
show net northward movement. Total net catch (southbound minus northbound) is shown
in the upper right comer, x=trap not operating. Data from the first 4 weeks of 1987,
and all of 1988, collected in Gainesville are from Walker (1991).

112
movement in the early weeks (3-5) and again, during later weeks (10-12). In fact, all
sites seemed to have some semblance of this bimodal pattern, not quite so clearly
separated as at Lake Alfred. This pattern could represent a migration surge, perhaps by
a second generation, later in the season. There were no long tailed skippers captured in
Lake Placid in either year of sampling, although individuals were seen feeding in the area
at each visit. These individuals may have been part of a resident, non-migratory
population or numbers in the area may have been too low for effective trapping.
Trapping results indicate that this species does not pass through Lake Placid in any
significant numbers.
Buckeye
Flight trap catches
The percent total seasonal net trap catch of Precis coenia along the north-south
transect during 1987 is shown in Fig. 4-16 a,b,c,d,e,f. This species showed no
significantly (chi-square, p < .05) biased movement for this species at any of the sites.
Net numbers were small, the most caught were seven individuals at Gainesville (Fig. 4-
16b). This is not surprising, as the sizable migration northward for this species is largely
in the spring. Of the sites where 4 or more individuals were caught, Valdosta,
Gainesville and Lake Alfred (Fig. 4-16 a,b,c,e), the migration midpoint was week 5 or
6. Leesburg (Fig. 4-16d) had a net northward movement of only one individual and
there were no buckeyes caught at Lake Placid (Fig. 4-16e).
The percent total net trap catches of Precis coenia at north-south sites during
1988 are shown in Fig. 4-17 a,b,c,d,e,f,g,h,i. Neither trap 1 or 2 (Fig. 4-17 a,b) at

113
1 2 3 4 5 6 7 S 8 10 11 12 13 14 15
C) WEEKS
GAINESVILLE PTP 3&5 (WALKER,1991)
b)
GAINESVILLE
LEESBURG
WEEKS
e)
LAKE ALFRED
f)
WEEKS
LAKE PLACID
WEEKS
Fig. 4-16 a,b,c,d,e,f."Weekly percentages of total net trap catch of Precis coenia at
Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida during
13 September-12 December 1987. Solid bars represent statistically significant southward
biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p > .05)
southward movement. Hatched bars show significant northward movement (chi-square
test, p < .05) and open bars show nonsignificant (p > .05) northward movement. The
total net catch (southbound minus northbound) for the period is shown in the upper right
comer. The midpoint of migration is designated by an asterisk. x=trap not operating.

Fig. 4-17 a,b,c,d,e,f,g,h,i.--Weekly percentage of total net trap catch of Precis coenia
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August-
26 November 1988. Data for Gainesville, Florida are from Walker (1991). Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05).
Shaded bars show nonsignificant (p >.05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right comer. The midpoint of
migration is designated by an asterisk, x=traps not operating.

115
a)
c)
WEEKS
LEESBURG TRAP 1
WEEKS
LEESBURG TRAP 2
g)
1 2 3 4 5 6 7 6 9 1011 1213
WEEKS
LAKE ALFRED TRAP 1
WEEKS
LAKE ALFRED TRAP 2
1 23456789 1011 1213
WEEKS

116
Valdosta showed any significant net movement during the season. Walker’s (1991) data
from Gainesville revealed a median at weeks 6 and 7 for traps 3 and 5 (Fig. 4-17 c,d),
respectively. Leesburg trap 1 (Fig. 4-17e) netted the most individuals of any site (n = 17)
sampled and was the only site with significant net southward movement (week 4). There
were no individuals caught in trap 2 (Fig. 4-17f) at Leesburg. Both traps at Lake Alfred
recorded slight net movement northward, and Lake Placid traps caught only one
individual during the entire season.
A summary of percent total net trap catches of Precis coenia for 1987 and 1988
is shown in Fig. 4-18 a,b,c,d,e. Overall, the only statistically significant (chi-square,
p C.05) net movement for this species during 1987-1988 sampling periods was
southward. Mid-migration at all sites occurred between weeks 3-5. All significant
movement ceased after week 7 in Valdosta, but continued at low levels through at least
week 13 in Gainesville. Movement south continued at Leesburg and Lake Alfred through
week 11. Only one individual was trapped in Lake Placid. As was shown for the other
three migrant species, the lack of substantial trap catches or large numbers of individuals
suggests there is no significant migration through Lake Placid. These indications are
further substantiated by Dr. Mark Deyrup, insect ecologist, Archbold Biological Station,
Lake Placid, Florida. Gulf fritillaries were seen ovipositing on Passiflora during the fall
of 1988 and occasionally large flights of Eurema daira were observed, but no other
conspicuous migratory flights were noted (M. Deyrup, pers. comm.).

117
a)
VALDOSTA
1 2 3 4 6 6 7 8 6 10 111213 1416
WEEKS
e)
b)
GAINESVILLE
WEEKS
LAKE ALFRED
â– 
N = 7
•
N
•
N
««rífiri..
u
XXX
1 2 3 4 6 6 7 8 9 10 11 12131416
WEEKS
LAKE PLACID
WEEKS
Fig. 4-18 a,b,c,d,e.—Summary of percent total net trap catches of Precis coenia at five
sites along a north-south transect (Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred
and Lake Placid, Florida) during 13 September-12 December 1987 and 28 August-26
November 1988. Solid bars represent net southward movement and open bars show net
northward movement. Total net catch (southbound minus northbound) is shown in the
upper right comer, x=traps not operating. Data for the first 4 weeks of 1987 and all of
1988 Gainesville data are from Walker (1991).

CHAPTER 5
SUMMARY AND DISCUSSION
This study has confirmed that the four principal migrants, Phoebis sennae.
Agraulis vanillae. Urbanus proteus and Precis coenia migrate through the Florida
peninsula each fall in a southeasterly manner similar to that described by Walker (1991)
in Gainesville. Observations of flight azimuths for three of these species (excluding the
buckeye) were made at various locations throughout the state during the fall of 1986.
Most significant mean flight directions were consistent with the 141° previously
described for Gainesville (Walker, 1991). Flight was generally erratic and variable along
the coastlines, as had also been reported by previous observers. The mean flight
directions noted in fall of 1986 indicated that significant southward movement continued
to the southern border of Lake Okeechobee. This is approximately the latitude at which
damage from freezing temperatures is unlikely in the state, the northern limit for many
tropical south Florida species, as well as the southern limit for some temperate ones.
Drive counts along an east-west transect were made across north central Florida
at the latitude of Gainesville (29.65°) during the fall of 1986-1988. These counts yielded
a migration profile, or cross section through the migratory stream, which was used to
estimate the density of individuals moving south into central Florida annually. The
highest density of the cloudless sulphur was through the central parts of the state, with
118

119
peak numbers recorded just east of Gainesville. The greatest density for the gulf
fritillary migration was observed along the western sections of the state, with peak
numbers noted just west of Gainesville. Flight azimuth observations, drive counts and
pole counts made across north central Florida have all demonstrated that the numbers of
cloudless sulphurs and gulf fritillaries are very low along the Atlantic coast. This is
consistent with Walker’s (pers. comm.) reports that these species gradually changed their
migratory tracks, from southeast to south, along an east-west transect from the central
to Atlantic coastal regions in south Georgia. Long tailed skippers were not included in
the migration profiles, but pole counts at sites along the same transect suggest that
migration is more common for this species along both the Atlantic and Gulf coasts, as
well as through Gainesville. The four species are alike in having a summer breeding
range that includes much of the eastern United States, and although the overwintering
ranges are not adequately described, they appear to be restricted to peninsular Florida
or to peninsular Florida and the adjacent coastal plain. The buckeye is the only one of
these species that has a large northward spring migration. Since the fall migration is
comparatively small (Walker, 1991), very few individuals of this species were seen
during these fall studies.
Using the density estimates from migration profiles and previous data from
permanent traps in Gainesville (Walker, 1991), the total number of cloudless sulphurs
and gulf fritillaries moving south each fall into central Florida was estimated at 42 and
115 million individuals, respectively. When Walker’s (1991) previous estimates for the

120
long tailed skipper and buckeye were included, the total number of butterflies migrating
south through the Florida peninsula was estimated to be 175 million individuals annually.
The phenology of the migration was studied by trapping the four major migrant
species using directional flight traps at five stations, from south Georgia to south central
Florida, during the fall of 1987 and 1988. All statistically significant net movement was
southward, and extended to the south central site of Lake Alfred, about 192 km south of
Gainesville. There was no statistically significant movement recorded for the buckeye,
but numbers captured were low. There was also no evidence from flight trap catches or
drive counts at the Lake Placid latitude (27.45°) that large numbers of migrants regularly
move into extreme south Florida to overwinter.
The midpoints of migration revealed through fall 1987 and 1988 flight trap
catches suggest several possibilities for the pattern of movement along the Florida
peninsula. The movement along the southward route was closely coordinated in time for
all species and usually occurred within the first three weeks of October. This time span
does not indicate a step-wise generational flight that occured over short distances. One
possibility is that individuals fly long distances along the migratory route. Another
possibility is that populations increase throughout the state during late summer and fall
migration is more or less synchronous over a large area. In this scenario, individuals fly
relatively short distances, followed by a second generation. For two cloudless sulphurs
marked in Gainesville, the duration of travel was longer than expected from the
knowledge of ground speed. These individuals traveled only a distance of 13 and 51 km
in 10 and 14 days, respectively. As an alternative to the two possibilities mentioned,

121
some individuals may have a facultative migration during which they follow a wave of
short-lived larval and adult resources as they are available. The bimodal pattern noticed
in the seasonal number of individuals caught in flight traps, two peaks separated by about
three weeks, suggests that a second wave of migrants, probably offspring, move along
later in the season.
While migrating, these species remain reproductively active. Most females are
mated and will oviposit abundantly on available hostplants. In north Florida, the larval
hostplants senesce shortly after the migration peaks and these, as well as adult nectar
sources, may be limited resources for these species. In fact, some cloudless sulphur
individuals marked in Gainesville remained near the same areas with abundant nectar
sources, at least through December. As mentioned previously, the cloudless sulphur and
buckeye are capable of regularly overwintering in north Florida and the adjacent coastal
plain. During unusually warm winters, a few gulf fritillaries may survive in north
Florida, but this is not common.
An overall picture of the butterfly migration through the Florida peninsula is still
elusive and requires more detailed studies in many areas throughout the state. The
question of just how far individuals migrate and at what rate and consistency is still
unclear. Although the trapping results of this study suggest that movement south occurs
in a closely coordinated manner, it does not rule out the possibility that migration is
synchronous over a large area. The availability of larval hostplants and adult nectar
sources seem to play an important role, particularly in the migration of the cloudless
sulphur. Hostplant phenology needs to be closely monitored throughout the year on a

122
latitudinal gradient. This, in conjunction with a mark-release-recapture and trapping
program for adult populations would explore the relationship between resource
availability and migratory movements. Breeding patterns need to be described along a
north-south gradient in order to identify generational movements. A large scale tagging
program throughout the southeast is the most definitive way to identify the long distance
movements of individuals. The recent development of an inexpensive, and very portable
flight trap (Walker, pers. comm.), makes this type of program more feasible, and one
has been undertaken by J. Whitesell (pers. comm.) in south Georgia with the
participation of local high schools over the past several years.
How northern regions are recolonized is also intriguing and particularly
important with regard to planning future conservation practices. Even with so many
millions of individuals moving through the peninsula, no specific "overwintering" area
has been identified. From the patterns seen during the three years this study was
conducted, it appears that the western half of south central Florida is the limit of
significant movement south. There are large areas of wilderness, much of this within
the Southwest Florida Water Management District, along the Withlahoochee River which
should be investigated as potentially vital winter survival sites. It is possible that a very
small percentage of individuals survive scattered along the southern part of the migratory
route, and these slowly repopulate the northern range the following year. This also
appears to fit the gradual population buildup which seems to occur throughout the
summer in north Florida. If this is true, then a very large genetic pool is drastically
reduced each year and any large scale catastrophic incident may significantly effect the

123
survival of this reduced population. Another interesting aspect to this problem is the
presence of resident populations in south Florida, particularly the gulf fritillary, which
seems to be common there throughout the year. Massive migrations of the four species,
so obvious in the northern regions of the state, have never been reported in extreme
south Florida, and the possible genetic differences between the apparently resident and
migrant populations, especially the gulf fritillary, would be worth exploring.
It is clear from this, and previous studies, that the Florida panhandle, Gulf coast
and central portion of the state constitute a major pathway for butterfly migration. The
large numbers of individuals involved in this annual migration through the Florida
peninsula is impressive. The number of different species involved, from a large area of
the southeastern region of the United States, makes it one of the world’s most spectacular
insect migrations.

REFERENCES
Abel, K.P. 1980. Mechanisms of orientation, navigation and homing. Pp. 283-373 in
Animal Migration, Orientation, and Navigation, S.A. Gauthreaux, Jr. (ed.),
Academic Press, New York. 387 pp.
Arbogast, R.T. 1965. Biology and migratory behavior of Agraulis vanillae (L.)
(Lepidoptera, Nymphalidae). Ph.D. Dissertation, University of Florida,
Gainesville. University Microfilms International, Ann Arbor, Michigan. 96 pp.
Arbogast, R.T. 1966. Migration of Agraulis vanillae (Lepidoptera, Nymphalidae) in
Florida. Fla Entomol. 49:141-145.
Baker, R.R. 1968a. Sun orientation during migration in some British butterflies. Proc.
R. Entomol. Soc. Lond. (A) 43:89-95.
Baker, R.R. 1968b. A possible method of evolution of the migratory habit in butterflies.
Phil. Trans. (B) 253:309-341.
Baker, R.R. 1969. The evolution of the migratory habit in butterflies. J. Anim. Ecol.
38:703-746.
Baker, R.R. 1978. The Evolutionary Ecology of Animal Migration. Homes and Meier,
New York. 1012 pp.
Baker, R.R. and J.G. Mather, 1982. Magnetic compass sense in the large yellow
underwing moth, Noctua prónuba L. Anim. Behav. 30:543-548.
Balciunas, J. and K. Knopf, 1977. Orientation, flight speeds, and tracks of three species
of migrating butterflies. Fia. Entomol. 60:37-39.
Batschelet, E. 1981. Circular Statistics in Biology. Academic Press, New York. 371
pp.
Baust, J.G., A.H. Benton and G.D. Aumann, 1981. The influence of off-shore
platforms on insect dispersal and migration. Bull. Ecol. Soc. Am. 27:23-25.
124

125
Beall, G. 1946. Seasonal variation in sex proportion and wing length in the monarch
butterfly, Danaus plexippus L. Trans. R. Entomol. Soc. (Lond.) 97:337-353.
Beall, G. 1948. The fat content of a butterfly, Danaus plexippus L., as affected by
migration. Ecology 29:80-94.
Beall, G. and C.B. Williams, 1945. Geographical variation in the wing length of Danaus
plexippus. Proc. R. Entomol. Soc. (Lond. A) 20:65-76.
Beenakkers, A.M.T. 1965. Transport of fatty acids in Locusta migratoria during
sustained flight. I. Insect Physiol. 11:879-888.
Brower, L.P. 1961. Studies on the migration of the monarch butterfly. I. Breeding
populations of Danaus plexippus and D. gilippus berenice in south central
Florida. Ecology 42:76-83.
Brower, L.P. 1962. Biology of the monarch butterfly. Ecology 43:181-182.
Brower, L.P. 1977. Monarch migration. Nat. Hist. 86:40-53.
Brower, L.P. and Calvert, W.H. 1985. Foraging dynamics of bird predators on
overwintering monarch butterflies in Mexico. Evolution 39:852-868.
Calhoun, J.V., T.J. Allen and D.C. Iftner, 1990. Temporary breeding populations of
Phoebis sennae eubule L. (Lepidoptera: Pieridae) in Ohio and West Virginia.
J. Res. Lepid. 28:123-125.
Calvert, W.H. and L.P. Brower, 1986. The location of monarch butterfly (Danaus
plexippus! overwintering colonies in Mexico. J. Lep. Soc. 40:164-187.
Calvert, W.H., L.E. Hedrick and L.P. Brower, 1979. Mortality of the monarch
butterfly (Danaus plexippus L.): avian predation at five overwintering sites in
Mexico. Science 204:847-851.
Clark, A.H. and L.F. Clark, 1951. The Butterflies of Virginia. Smithsonian Institution,
Washington. 239 pp.
Danthanarayana, W. 1986. Introductory chapter. Pp. 1-10 in Insect Flight: Dispersal
and Migration, W. Danthanarayana (ed.), Springer-Verlag, New York. 289 pp.
Dingle, H. 1968. Life history and population consequences of density, photo-period, and
temperature in a migrant insect, the milkweed bug, Oncopeltus. Am. Nat.
102:149-163.

126
Dingle, H. 1972. Migration strategies of insects. Science 175:1327-1335.
Dingle, H. 1985. Chapter 8. Migration. Pp. 375-415 in Comprehensive Insect
Physiology, Biochemistry and Pharmacology, G.A. Kerkut and L.I. Gilbert
(eds.), Vol. 9. Pergamon Press, Oxford.
Dingle, H., C.K. Brown and J.P. Hegmann, 1977. The nature of genetic variance
influencing photoperiodic diapause in a migrant insect, On copel tus fasciatus.
Am. Nat. 111:1047-1059.
Dumont, H.J. and B.O.N. Hinnekint, 1973. Mass migration of dragonflies, especially
in Libellula quadrimaculata: a review, a new ecological approach and a new
hypothesis. Odontológica 2:1-20.
Edwards, G.B. and D.B. Richman, 1977. Flight heights of migrating butterflies. Fla.
Entomol. 60:30.
Gaddy, L.L. and P. Laurie, 1983. Notes on the autumnal northward migration of the
cloudless sulphur, Phoebis sennae (Pieridae), along the South Carolina coast.
J. Lep. Soc. 37:166-167.
Gibo, D.L. 1986. Flight strategies of migrating monarch butterflies (Danaus plexippus
L.) in southern Ontario. Pp. 172-184 in Insect Flight, Dispersal and Migration,
W. Danthanarayana (ed.), Springer-Verlag, Berlin. 289 pp.
Howe, W.H. 1975. The Butterflies of North America. Doubleday, New York. 633 pp.
Istock, C.A. 1978. Fitness variation in a natural population. Pp. 171-190 in Evolution
of Insect Migration and Diapause, H. Dingle (ed.), Springer-Verlag, New York.
284 pp.
Johnson, C.G. 1969. Migration and Dispersal of Insects by Flight. Methuen, London.
763 pp.
Johnson, C.G. 1974. Insect migration: aspects of its physiology. Pp. 279-334 in The
Physiology of Insecta, M. Rockstein (ed.), Vol.3, Academic Press, New York.
Jungreis, S.A. 1987. Biomagnetism: an orientation mechanism in migrating insects?
Fla. Entomol. 70:277-283.
Kennedy, J.S. 1961. A turning point in the study of insect migration. Nature (Lond.)
198:785-791.

127
Klots, A.B. 1951. A Field Guide to the Butterflies of North America, East of the Great
Plains. Houghton Mifflin Co., Houston, Texas. 349 pp.
Lambremont, E.N. 1968. Mass one-directional flight of cloudless sulfurs (Pieridae) in
Alabama and Mississippi. I. Lep. Soc. 22:182.
Lamb, R.J. and P.A. MacKay, 1983. Micro-evolution of the migratory tendency,
photoperiodic response and developmental threshold of the pea aphid,
Acyrthosiphon pisum. Pp. 209-218 in Diapause and Life Cycle Strategies in
Insects, V.K. Brown and I. Hodek (eds.), Dr. W. Junk, Boston. 283 pp.
Larsen, T.B. 1988a. The anatomy of a major butterfly migration in southern India.
Atalanta 18:267-281.
Larsen, T.B. 1988b. Butterfly migration activity in southwestern India during September
and October of 1986. Atalanta 18:283-289.
Lenczewski, B. 1980. Butterflies of Everglades National Park. South Florida Research
Center Report T-588. 110 pp.
Li, K.P., H.H. Wong and W.S. Woo, 1964. Route of the seasonal migration of the
Oriental armyworm moth in the eastern part of China as indicated by a three-
year result of releasing and recapturing of marked moths. Acta Phytophvlac.
Sin. 3:101-110.
Lindauer, M. and H. Martin, 1972. Magnetic effect on dancing bees. Pp. 559-567 in
Animal Orientation and Navigation, S.R. Galler, K. Schmidt-Koenig, G.J.
Jacobs, R.E. Belleville (eds.). NASA SP-262, U.S. Gov. Printing Office,
Washington, D.C.
MacFadden, B.J. and D.S. Jones, 1985. Magnetic Butterflies: a case study of the
monarch (Lepidoptera, Danaiidae). Pp. 407-415 in Magnetite Biomineralization
and Magnetoreception in Organisms, J.L. Kirschvink, D.S. Jones and B.J.
MacFadden (eds.), Plenum Press, New York. 682 pp.
Masters, A.R. 1988. Monarch butterfly (Danaus plexippusl thermoregulatory behavior
and adaptations for overwintering in Mexico. Ecology 69:458-467.
Mather, B. 1967. Variation in Junonia coenia in Mississippi (Nymphalidae). J. Lep.
Soc. 19:139-160.
Mather, B. and K. Mather, 1958. The butterflies of Mississippi. Tulane Stud. Zool.
6:63-109.

128
Meier, A.H. and A.J. Fivizzani, 1980. Physiology of migration. Pp. 225-282 in
Animal Migration, Orientation, and Navigation, S.A. Gauthreaux, Jr. (ed.),
Academic Press, New York. 387 pp.
Muller, J. 1977. Notes on migrating Phoebis sennae eubule Linn. (Pieridae). News
Lep. Soc. 4.
National Geographic. 1979. Bird Migration. 156:154A (Suppl.).
Nielsen, A. 1961. On the habits of the migratory butterfly Ascia monuste L. in Florida.
Biol. Meddel. 23:1-81.
Nielsen, A. and E.T. Nielsen, 1952. Migrations of the pierid butterfly Ascia monuste
L. Florida. Entomol. Mededeelingen 26:386-391.
Pedgley, D. 1982. Windbome Pests and Diseases, Meteorology of Airborne Organisms.
John Wiley and Sons, New York. 250 pp.
Penn, G.H. 1955. Mass flight of ocola skippers (Panoquina ocola Edwards). Lep. News
Soc. 9:79.
Pyle, R.M. 1981. The Audobon Field Guide to North American Butterflies. Alfred A.
Knopf, New York. 917 pp.
Rainey, R. C. 1951. Weather and the movements of locust swarms: a new hypothesis.
Nature (Lond.) 168:1057-1060.
Rankin, M.A. 1978. Hormonal control of insect migratory behavior. Pp. 5-32 in
Evolution of Insect Migration and Diapause, H. Dingle (ed.), Springer-Verlag,
Berlin and New York. 284 pp.
Riley, N.D. 1975. A Field Guide to the Butterflies of the West Indies. Collins,
London. 224 pp.
Roer, H. 1959. Uber Flug- und Wandergewohnheiten von Pieris brassicae L. Z.
Angew. Entomol. 44:272-309.
Roer, H. 1961a. Zur Kenntnis der Populationsdynamik und des
Migrationsverhaltens von Vanessa atalanta L. im palaarktischen Raum. Beitr.
Entomol. 11:594:613.
Roer, H. 1961b. Ergebnisse mehijahriger Markierungsversuche zur Erforschung der
Flug- und Wandergewohnheiten europaischer Schmetterlinge. Zool. Anz.
167:456-463.

129
Roer, H. 1962. Experimentelle Untersuchungen zum Migrationsverhalten des Kleinen
Fuchs (Agíais urticae L.). Beitr. Entomol. 12:528-554.
Roer, H. 1968. Weitere Untersuchungen uber die Auswirkungen der Widening auf
Richtung und Distanz der Fluge des Kleinen Fuchses (Agíais urticae L.) (Lep.
Nymphalidae) im Rheinland. Decheniana 120:313-334.
Roer, H. 1969. Zur Biologie des Tagfauenauges, Inachis io L. (Lep. Nymphalidae).
unter besonderer Berucksichtigung der Wanderungen im mitteleuropaischen
Raum. Zool. Anz. 183:177-194.
Roer, H. 1970. Untersuchungen zum Migrationsverhalten des Trauermantels Nvmphalis
antiopa L.) (Lep. Nymphalidae). Z. Angew. Entomol. 4:388-396.
Roff, D.A. 1975. Population stability and evolution of dispersal in a heterogeneous
environment. Oecolopia 19:217-237.
Schaefer, G.W. 1976. Radar observations of insect flight. R. Entomol. Soc. (Lond.)
7:157-159.
Schmidt-Koenig, K. 1975. Migration and Homing in Animals. Ser. Zoophysiology and
Ecology 6. Springer-Verlag, Berlin. 99 pp.
Scott, J.A. 1975. Movement of Precis coenia. a "pseudoterritorial" submigrant
(Lepidoptera: Nymphalidae). I. Anim. Ecol. 44:843-850.
Scott, J.A. 1986. The Butterflies of North America: A Natural History and Field Guide.
Stanford Univ. Press, Stanford, California. 583 pp.
Sedman, Y. and D.F. Hess, 1985. The Butterflies of West Central Illinois. W. Illinois
Univ., Ser. Biol. Sci. No. 11. 120 pp.
Shannon, H.J. 1916. Insect migrations as related to those of birds. Science Monthly
3:227-240.
Showers, R.B. Smelser, A.J. Keaster, F. Whitford, J.F. Robinson, J.D. Lopez and S.E.
Taylor, 1989. Recapture of marked black cutworm (Lepidoptera: Noctuidae)
males after long-range transport. Environ. Entomol. 18:447-458.
Solbreck, C. 1985. Insect migration strategies and population dynamics. Pp. 641-662
in Migration: Mechanisms and Adaptive Significance, M.A. Rankin (ed.),
Contrib. Mar. Sci. 27 (Suppl). 868 pp.

130
Southwood, T.R.E. 1962. Migration of terrestrial arthropods in relation to habitat.
Biol. Rev. 37:171-214.
Stinner, R.E., C.S. Barfield, J.L. Stimac and L. Dohse, 1983. Dispersal and migration
of insect pests. Annu. Rev. Entomol. 28:319-335.
Taylor, L.R. 1958. Aphid dispersal and diurnal periodicity. Proc. Linn. Soc. (Lond.)
169:67-73.
Taylor, R.A.J. and D. Reling, 1986. Preferred wind direction of long distance
leafhopper (Empoasca fabael migrants and its relevance to the return migration
of small insects. J. Anim. Ecol. 55:1103-1114.
Urquhart, F.A. 1960. The Monarch Butterfly. Univ. of Toronto Press, Toronto. 361
pp.
Urquhart, F.A. and N.R. Urquhart, 1976. Migration of butterflies along the gulf coast
of north Florida. I. Lep. Soc. 30:59-61.
Verheijen, F.J. 1978. Orientation based on directivity, a directional parameter of the
animal’s radiant environment. Pp. 447-458 in Animal Migration, Navigation,
and Homing, K. Schmidt-Koenig and W. T. Keeton (eds.) Springer-Verlag,
Berlin. 462 pp.
Walker, T.J. 1978. Migration and re-migration of butterflies through north peninsular
Florida: quantification with Malaise traps. J. Lep. Soc. 32:178-190.
Walker, T.J. 1980. Migrating Lepidoptera: are butterflies better than moths? Fla.
Entomol. 63:79-98.
Walker, T.J. 1985a. Butterfly migration in the boundary layer. Pp. 704-723 in
Migration: Mechanisms and Adaptive Significance, M. A. Rankin (ed.), Contrib.
Mar. Sci. 27 (Suppl.). 868 pp.
Walker, T.J. 1985b. Permanent traps for monitoring butterfly migration: tests in Florida
1979-84. I Lep. Soc. 39:313-320.
Walker, T.J. 1991. Butterfly migration from and to peninsular Florida. Ecol. Entomol.
16:241-252.
Walker, T.J. and B. Lenczewski, 1989. An inexpensive portable trap for monitoring
butterfly migration. J. Lep. Soc. 43:289-298.

131
Walker, T.J. and A.J. Riordan, 1981. Butterfly migration: are synoptic-scale wind
systems important? Ecol. Entomol. 6:433-440.
Walker, T.J. and S.A. Wineriter, 1981. Marking techniques for recognizing individual
insects. Fla. Entomol. 64:18-29
Williams, C.B. 1930. The Migration of Butterflies. Oliver and Boyd, Edinburgh. 473
pp.
Williams, C.B. 1937. Recent progress in the study of some North American migrant
butterflies. Ann. Entomol. Soc. Am. 31:211-230.
Williams, C.B. 1958. Insect Migration. Collins, London. 235 pp.
Zar, J.H. 1984. Biostatistical Analysis, 2nd ed., Prentice-Hall, New York. 718 pp.

BIOGRAPHICAL SKETCH
Barbara Lenczewski was bom in New Haven, Connecticut, in 1952. She
received a Bachelor of Science degree from the University of Connecticut, Storrs, in
1978. The following year, she collected mammals in the Paraguayan Chaco, as part of
a University of Connecticut research team. Upon returning to the U.S., Barbara was
employed as a fire ecology technician with the South Florida Research Center,
Everglades National Park. Her work documenting the butterflies of the region resulted
in the publication of Butterflies of Everglades National Park in 1980. That same year,
she began graduate studies in insect ecology and a teaching assistantship at Florida State
University. In 1985, she was awarded the Master of Science degree for her study of
Trachymvrmex septentrionalis. a northern leafcutting ant. That fall, she began doctoral
studies in entomology at the University of Florida, pursuing a long time interest in
butterfly migration. During 1985-1988, Barbara was employed as a research assistant
in the Department of Entomology and Nematology through the mole cricket biological
control project. From 1988-1992, her lepidopterological horizons were further
broadened to include moth mating behavior, and, she worked as a research associate at
the U.S.D.A., Insect Attractant’s, Behavior and Basic Biology Research Laboratory in
Gainesville, Florida until 1992. She received her Ph.D., in 1992, from the University
of Florida.
132

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Thomas J. W
Professor of
Nematology
er, Chair
tomology and
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
>
íames E. Lloyd
Professor of Entomology and
Nematology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Professor of Entomology and
Nematology
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Lincoln P. Brower
Professor of Zoology

This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.'»
December 1992
Dean, College of Agriculture
Dean, Graduate School

UNIVERSITY OF FLORIDA



112
movement in the early weeks (3-5) and again, during later weeks (10-12). In fact, all
sites seemed to have some semblance of this bimodal pattern, not quite so clearly
separated as at Lake Alfred. This pattern could represent a migration surge, perhaps by
a second generation, later in the season. There were no long tailed skippers captured in
Lake Placid in either year of sampling, although individuals were seen feeding in the area
at each visit. These individuals may have been part of a resident, non-migratory
population or numbers in the area may have been too low for effective trapping.
Trapping results indicate that this species does not pass through Lake Placid in any
significant numbers.
Buckeye
Flight trap catches
The percent total seasonal net trap catch of Precis coenia along the north-south
transect during 1987 is shown in Fig. 4-16 a,b,c,d,e,f. This species showed no
significantly (chi-square, p < .05) biased movement for this species at any of the sites.
Net numbers were small, the most caught were seven individuals at Gainesville (Fig. 4-
16b). This is not surprising, as the sizable migration northward for this species is largely
in the spring. Of the sites where 4 or more individuals were caught, Valdosta,
Gainesville and Lake Alfred (Fig. 4-16 a,b,c,e), the migration midpoint was week 5 or
6. Leesburg (Fig. 4-16d) had a net northward movement of only one individual and
there were no buckeyes caught at Lake Placid (Fig. 4-16e).
The percent total net trap catches of Precis coenia at north-south sites during
1988 are shown in Fig. 4-17 a,b,c,d,e,f,g,h,i. Neither trap 1 or 2 (Fig. 4-17 a,b) at


123
survival of this reduced population. Another interesting aspect to this problem is the
presence of resident populations in south Florida, particularly the gulf fritillary, which
seems to be common there throughout the year. Massive migrations of the four species,
so obvious in the northern regions of the state, have never been reported in extreme
south Florida, and the possible genetic differences between the apparently resident and
migrant populations, especially the gulf fritillary, would be worth exploring.
It is clear from this, and previous studies, that the Florida panhandle, Gulf coast
and central portion of the state constitute a major pathway for butterfly migration. The
large numbers of individuals involved in this annual migration through the Florida
peninsula is impressive. The number of different species involved, from a large area of
the southeastern region of the United States, makes it one of the worlds most spectacular
insect migrations.


UNIVERSITY OF FLORIDA


60
b)
I
)
KILOMETERS
WEST WEEK 6 :12 OCTOBER WEST
Â¥ 0
5 *
z
10 .
020
<
25
TRIP 1
TRIP 2
X X
ITN CMC TUN

NWB ONV
X X
5TN CPC TUN

NWB QNV
0 16 32 48 64 60 96 8 24 40 56 72 88
KILOMETERS
I
§
EAST WEEK 3 : 20 SEPTEMBER EAST
: *
3
*10
TRIP 1
TRIP 2
1 "
MV maw m. nx mr cm

X
inv maw m puc mr cm
96 11212B144 160 176 192 96 112126 144 160 176192
KILOMETERS
)
$ i
2
0
I *
s
Z 4
*
G *
z
-> 0
o
* 10
KILOMETERS
EAST WEEK 7 :14 OCTOBER
EAST
TRIP1
TRIP 2
T"'1
X X
ONV HAW m. FIX M8T CM
"
X X
INV HAW m. PUC H8T C*B
96 11212814416017619296 112126144160176192
KILOMETERS
9)
2 o
5
> 2
O 8
<
10
KILOMETERS
KILOMETERS


ARCADIA ALLIGATOR POINT BARTOW REI .IP GLADE CROSS CITY
O
CLERMONT CRESCENT BEACH CLEWISTON EVERGLADES CITY EASTPOINT
O
o
FROSTPROOF FISHCREEK GLEN ST. MARY HAWTHORNE HASTINGS

INTERLACHEN IMMOKALLEE JASPER LAKE CITY
^ J 0 0
LEESBURG LAKE ALFRED LABELLE LAKE PLACID NEWBERRY
0 Q
ADCLLL LJUsJl ri-AL
0
%
OLGA OKEECHOBEE PALATKA PERRY STEINHATCHEE
31
ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
0
Or 0 0~
Fig. 2-2.--Azimuths of Phoebis sennae flight directions taken at various locations in
Florida during fall migration 1985-1988. The diameter of the circle represents the
number of visits (5 mm = l visit); the length and width of the line represents the number
of individuals (2 mm = l, in bold 2 mm=4); the orientation of the line represents the
flight direction.


A=Arcadia n=6,r=.98,170 .jx.OOl
AP=Alligator Point n=13,r=.17,164
B=Bartown=10,r=.84,174 ,p<.001
BG=BeIle Glade n=0
C=Cross City n=21,r=.89,156 ,p<.001
CB=Crescent Beach n=o
CL=Clenmont n=8,i=.6,142
CW=Gewiston n=5,r=.92,146
E=Everglades City n=6,r=.62,109
EP=Eastpoint n=2j=.99,277
F=Frostproof n=l,r=l,163
FC=Fishcreek n=0
G=Gainesville (Walker, 1985a)
GS=Glen St. Mary n=0
H=Hawthome n=27 j=.84,160 ,p<.001
HA=Hastings n=4j=.84,132
I=lnterlachen n=14j=.96,179 ,p<001
IM=Immokalee n=16j=.72,164 ,p<.001
J=Jasper n=4j=.92
KB=Keeton Beach n=4,r=.7,290
L=Lake Cityn=4,r= 66,154
LE=Leesburg n=5,r= 54,163
LA=Lake Alfred n=8,r=.96,155 ,p< 001
LB=LaBelle n=2,r=.97,164
LP=Lake Placid n=3,r=.99,175
N=NewberTyn=16,r=.97,148 ,p<.001
0=01ga n=2j=.17,253
OK=Okeechobee n=4j=.96,143
P=Palatka n=12j=.84,172 ,p<.001
PE= Perry n=3j=.94,151
S=Steinhatchee n=13j=.80,137 ,p<.001
SG=St. George n=3,r=.56,141
T=Trenton n=6,r=.34,147
TA=Tallahassee n=9,r=.32,146
W=White Springs n=5,r=.58,174
YJ=Yeehaw Junction n=l r=l ,159
Fig. 2.3.--Summary of mean flight directions at various sites throughout Florida of Agraulis vanillae during fall migrations, 1985-
1988 (direction of arrow = mean flight direction, value given in key for each site; length of line=mean vector r; bold
letters=significant rvalue).
u>
N>


127
Klots, A.B. 1951. A Field Guide to the Butterflies of North America, East of the Great
Plains. Houghton Mifflin Co., Houston, Texas. 349 pp.
Lambremont, E.N. 1968. Mass one-directional flight of cloudless sulfurs (Pieridae) in
Alabama and Mississippi. I. Lep. Soc. 22:182.
Lamb, R.J. and P.A. MacKay, 1983. Micro-evolution of the migratory tendency,
photoperiodic response and developmental threshold of the pea aphid,
Acyrthosiphon pisum. Pp. 209-218 in Diapause and Life Cycle Strategies in
Insects, V.K. Brown and I. Hodek (eds.), Dr. W. Junk, Boston. 283 pp.
Larsen, T.B. 1988a. The anatomy of a major butterfly migration in southern India.
Atalanta 18:267-281.
Larsen, T.B. 1988b. Butterfly migration activity in southwestern India during September
and October of 1986. Atalanta 18:283-289.
Lenczewski, B. 1980. Butterflies of Everglades National Park. South Florida Research
Center Report T-588. 110 pp.
Li, K.P., H.H. Wong and W.S. Woo, 1964. Route of the seasonal migration of the
Oriental armyworm moth in the eastern part of China as indicated by a three-
year result of releasing and recapturing of marked moths. Acta Phvtophvlac.
Sin. 3:101-110.
Lindauer, M. and H. Martin, 1972. Magnetic effect on dancing bees. Pp. 559-567 in
Animal Orientation and Navigation, S.R. Galler, K. Schmidt-Koenig, G.J.
Jacobs, R.E. Belleville (eds.). NASA SP-262, U.S. Gov. Printing Office,
Washington, D.C.
MacFadden, B.J. and D.S. Jones, 1985. Magnetic Butterflies: a case study of the
monarch (Lepidoptera, Danaiidae). Pp. 407-415 in Magnetite Biomineralization
and Magnetoreception in Organisms, J.L. Kirschvink, D.S. Jones and B.J.
MacFadden (eds.), Plenum Press, New York. 682 pp.
Masters, A.R. 1988. Monarch butterfly (Danaus plexippusl thermoregulatory behavior
and adaptations for overwintering in Mexico. Ecology 69:458-467.
Mather, B. 1967. Variation in Junonia coenia in Mississippi (Nymphalidae). J. Lep.
Soc. 19:139-160.
Mather, B. and K. Mather, 1958. The butterflies of Mississippi. Tulane Stud. Zool.
6:63-109.


126
Dingle, H. 1972. Migration strategies of insects. Science 175:1327-1335.
Dingle, H. 1985. Chapter 8. Migration. Pp. 375-415 in Comprehensive Insect
Physiology, Biochemistry and Pharmacology, G.A. Kerkut and L.I. Gilbert
(eds.), Vol. 9. Pergamon Press, Oxford.
Dingle, H., C.K. Brown and J.P. Hegmann, 1977. The nature of genetic variance
influencing photoperiodic diapause in a migrant insect, On copel tus fasciatus.
Am. Nat. 111:1047-1059.
Dumont, H.J. and B.O.N. Hinnekint, 1973. Mass migration of dragonflies, especially
in Libellula quadrimaculata: a review, a new ecological approach and a new
hypothesis. Odontolgica 2:1-20.
Edwards, G.B. and D.B. Richman, 1977. Flight heights of migrating butterflies. Fla.
Entomol. 60:30.
Gaddy, L.L. and P. Laurie, 1983. Notes on the autumnal northward migration of the
cloudless sulphur, Phoebis sennae (Pieridae), along the South Carolina coast.
J. Lep. Soc. 37:166-167.
Gibo, D.L. 1986. Flight strategies of migrating monarch butterflies (Danaus plexippus
L.) in southern Ontario. Pp. 172-184 in Insect Flight, Dispersal and Migration,
W. Danthanarayana (ed.), Springer-Verlag, Berlin. 289 pp.
Howe, W.H. 1975. The Butterflies of North America. Doubleday, New York. 633 pp.
Istock, C.A. 1978. Fitness variation in a natural population. Pp. 171-190 in Evolution
of Insect Migration and Diapause, H. Dingle (ed.), Springer-Verlag, New York.
284 pp.
Johnson, C.G. 1969. Migration and Dispersal of Insects by Flight. Methuen, London.
763 pp.
Johnson, C.G. 1974. Insect migration: aspects of its physiology. Pp. 279-334 in The
Physiology of Insecta, M. Rockstein (ed.), Vol.3, Academic Press, New York.
Jungreis, S.A. 1987. Biomagnetism: an orientation mechanism in migrating insects?
Fla. Entomol. 70:277-283.
Kennedy, J.S. 1961. A turning point in the study of insect migration. Nature (Lond.)
198:785-791.


A=Arcadia n=13,r=.58,161 ,p< 01
AP=Alligalor Point n=18,r=.26,68
B=Baitow n=27,r=.88,146,p<.001
BG-Belle Glade n=0
C=Cross City n=51,r=,89,l7l,p<.OOI
CB=Ctescent Beach n=0
CL=Clermont n=8,r=.45,191
CW=Clewiston n=4,r=.77,123
E=EvergladesCity n=2j=.51,32
EP=Eastpoiri n=3,r=.9,46
F=Frostproof n=3,r=.72,92
FC=Fishcreek n=4,r=53320
G=Gainesville 141 (Walker, 1985a)
GS=Glen St. Mary n=5j=.59,126
H=Hawthome n=61,r=.81,152,p<.001
HA=Hastings n=9,r=.33,175
l=Interlachen n=34,r=.94,168,p< 001
IM=Inimokalee n=12j=.62,170 ,p<.01
J=Jasper n=4,r=.!2,162
KB=Keeton Beach n=6,r=.29,330
L=LakeCity n=4,r= 80,146
LE= Leesburg n=8,r=,96,133,p< 001
LA=Lake Alfred n=7,r=.74,165,p<.02
LB=LaBeUe n=13,r=.66,153,p<.005
LP=Lake Placid n=4,r=.79,71
N=Newberry n=41 j=.80,146,p<.001
OOIga n=13/=.84,178 ,p<.001
OK=Okeechobee n=12,r=.65,77,p<.005
P=Palatka n=17,r=.72,143 ,p<.001
PE=Perry n=18,n=.93.144 ,p< 001
S=Steinhatchee n=46,r=.42,86,px. 001
SG=St. George n=2,r=l,54
T=Trenton n=25,r=.72,151 ,p<.001
TA=Tallahassee n=9,r=.32,94
W=White Springs n=2,r=.88,167
YJ=YeehawJunction n=l,r=l,159
Fig. 2-1.-Summary of mean flight directions at various sites throughout Florida for Phoebis sennae during fall migrations,
1985-1988 (direction of arrow=mean flight direction, value given in key for each site; length of line=mean vector r; bold
letters=significant r value).
to


52
at this time, the net number of individuals flying south/minute was nearly 40, 8 times
greater than seen previously. Unfortunately, that was the only drive count through CRB-
TRN during the entire fall. During week 9, the GNV-HAW and ITL-PLK segments
along the eastern transect also increased, but only to about a third of what was seen on
the Gulf coast. On 16 October, the migration through the two Gulf coast segments
(STN-CRC, CRC-TRN) was southward and twenty times higher than that of any other
segment, at any other time, throughout the season. It is not known if large numbers of
migrants regularly pass through these segments. The one additional sample through
STN-CRC (Fig. 3-4c, west) and four samples at TRN-NWB (Fig.3-4 a,c,d,e, west) seem
to indicate that at least during the early season it is not common. It is possible that such
large scale movements result from synchronous emergence of a locally produced
generation. Large fields observed in the Trenton area had extensive stands of Cassia
obtusifolia. a primary hostplant of cloudless sulphurs in this area.
Similar profile estimates of the cloudless sulphur migration during fall 1987 are
shown in Fig. 3-5 a,b. All net movement was southward and greatest along the western
transect through GNV-HAW during week 6. This segment had more than 300 indiv/min
passing south, but the net catch for Gainesville that day was only one cloudless sulphur,
flying south! This resulted in a high migration index that perhaps inflated what was an
average day in the HAW-ITL segment. Along the eastern transect, the numbers were
highest through the TRN-NWB segment during week 7, with 4.5 indiv/min. This
transect was sampled on 17 October, almost one year to the day when extremely large


78
Despite all these advantages, and, probably because they are not agriculturally
significant pests, the often spectacular movements of these insects have been largely
ignored. A few workers (Arbogast, 1965, 1966; Balciunas and Knopf, 1977) have made
observations over short periods of time and space, but it has been difficult to
simultaneously monitor migrations over distance. Through a massive tagging campaign,
Urquhart and Urquhart (1976) succeeded in determining the destination of Danaus
plexippus (L.), the monarch butterfly, in the Mexican mountain ranges. The migrations
of the great southern white, Ascia monuste L., along the Florida east coast were detailed
by Nielsen and Nielsen (1952; Nielsen, 1961) over a period of at least 20 years. In
Germany, Roer (1959, 1961a, 1961b, 1962, 1968, 1969, 1970) has described the
movements of Agais urticae L., Inachis io, Nvmphalis antiopa and other European
butterflies, while Baker (1968ab, 1969) has studied Pieris rapae in England. For more
than ten years, Walker (1978, 1980, 1985ab, 1991) has maintained directional flight traps
in Gainesville, Florida that continuously monitor the annual migrations of eight species
of butterflies through this area. This type of data collection eliminates observer bias,
time of day or seasonal influences and permits correlation with environmental factors
(Walker and Riordan, 1981). A continuous, long term documentation of migration in
this manner is unique and has provided reliable information about the species
composition, phenology and size of the migration, at least through Gainesville.
There are three species that comprise most of the migration through north
central Florida in the fall (Walker, 1991): Phoebis sennae (L.) (Pieridae), the cloudless
sulphur; Agraulis vanillae (L.) (Heliconiidae), the gulf fritillary; and Urbanus proteus


Cloudless sulphur 44
Gulf fritillary 46
Long tailed skipper 46
Migration Profiles 49
Cloudless sulphur 49
Gulf fritillary 58
Numbers of Migrants 66
CHAPTER 4. PHENOLOGY OF MOVEMENT 77
Introduction 77
Materials and Methods 80
Longitudinal Pole Counts 80
Portable Flight Traps 81
Results and Discussion 83
Cloudless Sulphur 83
Longitudinal pole counts 83
Flight trap catches 85
Gulf Fritillary 93
Longtiudinal pole counts 93
Flight trap catches 96
Long Tailed Skipper 102
Longitudinal pole counts 102
Flight trap catches 105
Buckeye 112
Flight trap catches 112
CHAPTER 5. SUMMARY AND DISCUSSION 118
REFERENCES 124
BIOGRAPHICAL SKETCH 132
vi


REFERENCES
Abel, K.P. 1980. Mechanisms of orientation, navigation and homing. Pp. 283-373 in
Animal Migration, Orientation, and Navigation, S.A. Gauthreaux, Jr. (ed.),
Academic Press, New York. 387 pp.
Arbogast, R.T. 1965. Biology and migratory behavior of Agraulis vanillae (L.)
(Lepidoptera, Nymphalidae). Ph.D. Dissertation, University of Florida,
Gainesville. University Microfilms International, Ann Arbor, Michigan. 96 pp.
Arbogast, R.T. 1966. Migration of Agraulis vanillae (Lepidoptera, Nymphalidae) in
Florida. Fla Entomol. 49:141-145.
Baker, R.R. 1968a. Sun orientation during migration in some British butterflies. Proc.
R. Entomol. Soc. Lond. (A) 43:89-95.
Baker, R.R. 1968b. A possible method of evolution of the migratory habit in butterflies.
Phil. Trans. (B) 253:309-341.
Baker, R.R. 1969. The evolution of the migratory habit in butterflies. JL Anim. Ecol.
38:703-746.
Baker, R.R. 1978. The Evolutionary Ecology of Animal Migration. Homes and Meier,
New York. 1012 pp.
Baker, R.R. and J.G. Mather, 1982. Magnetic compass sense in the large yellow
underwing moth, Noctua prnuba L. Anim. Behav. 30:543-548.
Balciunas, J. and K. Knopf, 1977. Orientation, flight speeds, and tracks of three species
of migrating butterflies. Fia. Entomol. 60:37-39.
Batschelet, E. 1981. Circular Statistics in Biology. Academic Press, New York. 371
pp.
Baust, J.G., A.H. Benton and G.D. Aumann, 1981. The influence of off-shore
platforms on insect dispersal and migration. Bull. Ecol. Soc. Am. 27:23-25.
124


CHAPTER 2
FLIGHT DIRECTION
Introduction
Of all insect migratory movements, those of butterflies are the most evident and
most accessible to study. Unlike their nocturnal counterparts, the moths, butterflies are
highly visible, daytime fliers which move through the boundary layer (Williams, 1930;
Baker, 1978). This is the layer of air near the ground where wind velocity is less than
the insects air speed. The thickness of this boundary layer is variable, determined by
wind velocity as well as the air speed maintained by the individual (Taylor, 1958;
Pedgley, 1982). Most other insect migrants, such as leafhoppers (Taylor and Reling,
1986), noctuid moths, locusts and aphids (Johnson, 1969), travel at much higher
altitudes. Radar studies have shown that they are generally flying with the wind
(Pedgley, 1982), although they may take off from the ground only when wind direction
is favorable (Williams, 1958). Flight within a few meters of the ground, however,
affords the individual more control over direction and also the opportunity to cease flying
by clinging to vegetation should conditions become unfavorable. This type of flight
allows an observer to easily identify species as well as to observe flight direction and
behavior under variable environmental conditions.
23


91
Lake Alfred trap 2 was during week 11. At these four sites, from Valdosta to Lake
Alfred, the mid-point of migration for traps 1 and 2 fell between weeks 6-9 and it is
found almost uniformly (except for Lake Alfred, trap 2) at a one week difference
between sites. Lake Placid had no statistically significant bias in movement at any time
during the trapping period, although five individuals were trapped and others were seen.
The percentage of total net trap catches of cloudless sulphurs are summarized
for both the 1987 and 1988 fall migration seasons in Fig. 4-5 a,b,c,d,e. The 1988 trap
1 and 2 data were tested for homogeneity and combined before calculating the mean
percent total net catch for both years. The first four weeks of Gainesville 1987 and all
of 1988 data used in this summary are from Walkers (1991) permanent trap catches as
shown in Figs. 4-3c and 4-4 c,d. The migration midpoint occurred between weeks 5
through 7 at all of the four main sites. Lake Placid had no evidence of migration with
a total net of only one individual heading south in two years. The net numbers captured
at Valdosta (Fig. 4-5a) were surprisingly low, totaling only 70 individuals in two years,
and similar to catches at Lake Alfred (Fig. 4-5d). During 1987 alone, Gainesville traps
(Fig. 4-3b) captured nearly three times that number. The total net number at Leesburg
for the two years was 275. This was unexpected since it was assumed that most
individuals begin travelling south through Valdosta and then move through Gainesville.
During 1987, the earlier part of the migration may have been missed, but in 1988 the
pattern was similar with comparable numbers at Gainesville-Leesburg and Valdosta-Lake
Alfred. Subsequent breeding could augment the migratory stream between Leesburg and
Gainesville, accounting for the larger numbers caught at these sites. The sites south of


CHAPTER 1
INTRODUCTION
Literature Review
General Migration
Migration has been recognized in recent years as an important element in the
dynamics of insect populations (Solbreck, 1985). Long distance movements of butterflies
(Williams, 1930), locusts (Johnson, 1969), aphids (Kennedy, 1961), moths (Stinner si
al, 1983), dragonflies (Dumont and Hinnekint, 1973) and insects of many other groups
have been recorded by observers across practically every continent (Williams, 1958), at
high altitudes (Rainey, 1951), and many miles out to sea (Baust el al, 1981). Although
some insect flights seem to function as random dispersal by forces beyond the
individuals control (Johnson, 1969), closer behavioral studies reveal that active initiation
of flight occurs in many species and correlates with favorable wind conditions and
appropriate seasons (Baker, 1978; Taylor and Reling, 1986). In other words, the
migrants will "choose" the appropriate winds, during the appropriate times of year, by
which they may be carried.
It was perhaps due to his interest in butterflies that Williams (1930) stressed the
concept of individual control over flight direction in his early definition of migration.
This individual control over flight direction is particularly evident in butterfly migrations.
1


54
numbers had been noted there in 1986. The peak numbers along this western transect
in 1987 were only one tenth of what they were the previous year.
The profile estimates showing Phoebis sennae migration along this transect in
fall of 1988 are given in Fig. 3-6 a,b,c,d. Each transect consists of a roundtrip made
during the same day and close in time (trip 1 and 2). Each trip consistently showed the
peak net movement occurred southward through the central, NWB-INT, or western STN-
TRN segment and was lowest in the east.
All profile estimates made of Phoebis sennae net movement during the fall of
1986, 1987 and 1988 are summarized as yearly migration profiles in Fig. 3-7 a,b,c.
Most transect segments showed a statistically significant (chi-square test, p < .05) net
southward movement throughout the sampling period. The exceptions were PLK-HST,
where in 1986 and 1988, net movement was not significantly biased and HST-CRB, in
1987, where no individuals were sighted. During 1986 (Fig. 3-7a), the peak southward
movement was through the western segment of CRC-TRN, 37 indiv/min, followed by
STN-CRC with 14 indiv/min. As was pointed out in the individual 1986 profile
estimates (Fig. 3-4g), this was a result of one sample taken on 16 October. If this trip
were not included, the central transect segment from Trenton to Palatka would contain
most of the southward movement. Peak movement south for the fall of 1987 (Fig. 3-7b)
was through the HAW-INT segment at nearly 200 indiv/min, the highest rate recorded
through three years of sampling. The secondary peak was through the adjacent segment
of GNV-HAW with about 11 indiv/min. Both of these segments showed an average


105
the gulf fritillary and cloudless sulphur. However, the only statistically significant
migration of the season was through Gainesville during weeks 6 and 8. The peak rate
there was during week 6, at about 1.1 indiv/m/min, which was nearly twice the peak rate
of the gulf fritillary and slightly more than the cloudless sulphur. Although there was
net movement at all other sites, mostly southward, none of it was significant.
The summary of net movement of H. proteus observed during pole counts along
the north-south transect in 1986 is shown in Fig. 4-12. All net movement is southward,
but the only statistically significant (chi-square, p < .05) bias was through Gainesville
and Clermont, the two most northern sites. The numbers moving through Gainesville
averaged about .6 indiv/m/min and those at Clermont about, .2 indiv/m/min. The lack
of significantly biased directional movement south of Clermont is probably a result of
low sample numbers, since subsequent catches in flight traps revealed southward
movement through Lake Alfred.
Flight trap catches
The percent total net trap catch of Urbanus proteus along a north-south transect
from Valdosta, Georgia, to Lake Placid, Florida, during 1987 is shown in Fig. 4-13
a,b,c,d,e,f. All statistically significant (chi-square, p C.05) net movement for this
species was southward and was completed by weeks 5 or 6 at all sites. Valdosta (a),
Gainesville (b) and Leesburg (d) had surprisingly similar numbers (net = 37, 37 and 45)
of individuals captured. Midpoints at all sites, except Lake Placid, were during week
4 or 5 (Fig. 4-13 a,c,d,e). Numbers of individuals trapped at Lake Alfred (Fig. 4-13 e)
were very low (n=9), with no significant movement shown for any week throughout the


BARTOW
ARCADIA
IMMOKAUEE
Fig. 4-6 a,b,c,d,e,f.Pole counts of Agraulis vanillae movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p <.05). x=no counts made.


97
a)
VALDOSTA
b)
C)
WEEKS
GAINESVILLE PTP #3 (WALKER, 1991)
50 i
e)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 13
WEEKS
LAKE ALFRED
GAINESVILLE
d)
I 140
-
t 120
O 100
I 80
-J 60
% TOT/
ro £
o o o
L
WEEKS
LEESBURG
12l4 WEEKS
LAKE PLACID
11 12 11 1* 11
2 3 4 S 7 8 10 11 12 13 14 IS
WEEKS
WEEKS
Fig. 4-8 a,b,c,d,e,f.--Weekly percentage of total net trap catch of Agraulis vanillae at
Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida during
13 September-12 December 1987. Solid bars represent statistically significant southward
biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p > .05)
southward movement. Clear bars show nonsignificant (p > .05) northward movement.
The total net catch (southbound minus northbound) for the period is shown in the upper
right comer. The midpoint of migration is designated by an asterisk, x=traps not
operating.


CHAPTER 5
SUMMARY AND DISCUSSION
This study has confirmed that the four principal migrants, Phoebis sennae.
Agraulis vanillae. Urbanus proteus and Precis coenia migrate through the Florida
peninsula each fall in a southeasterly manner similar to that described by Walker (1991)
in Gainesville. Observations of flight azimuths for three of these species (excluding the
buckeye) were made at various locations throughout the state during the fall of 1986.
Most significant mean flight directions were consistent with the 141 previously
described for Gainesville (Walker, 1991). Flight was generally erratic and variable along
the coastlines, as had also been reported by previous observers. The mean flight
directions noted in fall of 1986 indicated that significant southward movement continued
to the southern border of Lake Okeechobee. This is approximately the latitude at which
damage from freezing temperatures is unlikely in the state, the northern limit for many
tropical south Florida species, as well as the southern limit for some temperate ones.
Drive counts along an east-west transect were made across north central Florida
at the latitude of Gainesville (29.65) during the fall of 1986-1988. These counts yielded
a migration profile, or cross section through the migratory stream, which was used to
estimate the density of individuals moving south into central Florida annually. The
highest density of the cloudless sulphur was through the central parts of the state, with
118


113
1 2 3 4 5 6 7 S 8 10 11 12 13 14 15
C) WEEKS
GAINESVILLE PTP 3&5 (WALKER,1991)
b)
GAINESVILLE
LEESBURG
WEEKS
e)
LAKE ALFRED
f)
WEEKS
LAKE PLACID
WEEKS
Fig. 4-16 a,b,c,d,e,f."Weekly percentages of total net trap catch of Precis coenia at
Valdosta, Georgia; Gainesville, Leesburg, Lake Alfred and Lake Placid, Florida during
13 September-12 December 1987. Solid bars represent statistically significant southward
biased movement (chi-square test, p < .05), shaded bars show nonsignificant (p > .05)
southward movement. Hatched bars show significant northward movement (chi-square
test, p < .05) and open bars show nonsignificant (p > .05) northward movement. The
total net catch (southbound minus northbound) for the period is shown in the upper right
comer. The midpoint of migration is designated by an asterisk. x=trap not operating.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BUTTERFLY MIGRATION THROUGH THE FLORIDA PENINSULA
By
Barbara Lenczewski
December 1992
Chairman: Dr. Thomas J. Walker
Major Department: Entomology and Nematology
At least eight species of butterflies migrate southward during fall of each year
through the Florida peninsula. The four major migrants, representing four families of
Lepidoptera, are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis vanillae
(L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the long tailed
skipper; and Precis coenia (Hbner) (Nymphalidae), the buckeye. Drive counts of the
net southward movement of the cloudless sulphur and gulf fritillary were made across
north central Florida at the latitude of Gainesville (29.65) during fall 1986-1988. This
migration profile, or cross section through the migratory stream, was used to estimate
the density of individuals moving south into central Florida annually. The highest
density of the cloudless sulphur was through the central parts of the state and that of the
gulf fritillary, along the western sections. Peak numbers were recorded just east and just
vii


58
migration rate, but traps in Gainesville caught very little. As a result, the high migration
index for that day has increased numbers for these segments. If these days were not
included, the GNV-HAW segment would still have the peak southward movement at 2.3
indiv/min, followed by the TRN-NWB segment at 1.4 indiv/min. There were no
cloudless sulphurs observed during the one trip made through the HST-CRB segment
along the Atlantic coast in 1987. The migration profile for fall of 1988 (Fig. 3-7c)
showed a peak through HAW-INT of 2.7 indiv/min, nearly twice that of the next highest
peak through NWB-GNV (1.4 indiv/min). Movement along either coast was low, with
both east and west segments below .5 indiv/min. For all three years, the Atlantic coastal
segment of transect, Palatka to Crescent Beach, consistently had the least number of
migrants, averaging about .18 indiv/min.
Gulf fritillarv
The profile estimates of Agraulis vanillae migration during fall 1986 are shown
in Fig. 3-8 a,b,c,d,e,f,g,h. On the western transect, movement mostly occurred through
the TRN-NWB segment and peaked during week 6 (Fig. 3-8e) with 20 indiv/min. The
Gulf coast segment of STN-CRC, was not sampled during this time, but the following
week (Fig. 3-8g), it showed peak movement south at 8 indiv/min. Numbers were low
along the eastern transect during all the weeks sampled, except for large numbers (23
indiv/min) during week 9 (trip 2, Fig. 3-8h) through INT-PLK. All peak movement
through the season occurred within this segment, except for week 9 (trip 1, Fig. 3-8h)
where GNV-HAW had nearly 15 indiv/min.


73
4. WILD WOOD-NEW SMYRNA
5. NOBLETON-BUSHNELL
E. BROOKSVILLE-HILL N DALE
7. BAYONET POINT-SAN ANTOMO
8. WESLEY CHAPEL-ZEPHRHILLS
#. LAKE LAND-HA INES CITY
10. BARTOW-LAKE WALES
11. FORT MEADE-FROSTPROOF
12. ZOLFO SPRINGS-SEBRING
13. BEE RIDGE-LAKE PLACID
o NORTH BIAS
O O NO BIAS
GNV LAT (29.65)
NONE SEEN
7 MM =20 KM


% TOTAL NET CATCH % TOTAL NET CATCH % TOTAL CATCH % TOTAL NET CATCH
101
a)
VALDOSTA TRAP 1
VALDOSTA TRAP 2
WEEKS
GAINESVILLE PTP3 (WALKER,1991)
WEEKS
LEESBURG TRAP 1
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
LAKE ALFRED TRAP 1
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
d)
WEEKS
GAINESVILLE PTP5 (WALKER,1991)
50 i
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
LEESBURG TRAP 2
WEEKS
LAKE ALFRED TRAP 2
h)
LAKE PLACID
1 2 3 4 5 6 7 8 9 10111213
WEEKS
Ox50
o
H 40
o
t30
z
H
*
-
*
f\
N=6
X
Ji
*
1 2 3 4 5 6 7 8 9 1011 12 13
WEEKS


96
decreased greatly and there was no significant migration recorded south of these areas.
Flight trap catches
The weekly percentages of seasonal total net trap catch of Agraulis vanillae
along a north-south transect through five sites from Valdosta, Georgia to Lake Placid,
Florida during fall 1987 are shown in Fig. 4-8 a,b,c,d,e,f. All statistically significant
(chi-square, p < .05) net movement is southward. The midpoint of migration occurred
in Valdosta (Fig. 4-8a) no later than week 5. The lack of data for the early weeks has
probably shifted the midpoint forward and there has been no adjustment made for this.
The pattern seen from the two Gaineville sites suggest a buildup before the mode; thus
the Valdosta mode during week 4 may also represent the median. Significant southward
movement occurred at Valdosta as late as week 11. With the same trapping effort,
significant southward migration occurred during weeks 5, 6 and 7 at Gainesville (Fig.
4-8b) with mid-migration seen from Walkers (1991) traps (Fig. 4-8c) as occurring
during week 5. Net numbers of gulf fritillary individuals were comparable in Valdosta
and Gainesville, numbering 72 and 69, respectively, unlike the large discrepancy
described for the cloudless sulphur at these sites. The Valdosta mid-migration point of
week 4 or 5 and Gainesville mid-migration point of week 5 suggest that the same
individuals may travel through south Georgia to north central Florida in a timely manner.
It is also possible that throughout this general area, the initiation of migratory flight is
coordinated. Although numbers captured in Valdosta and Gainesville were similar, the
net movement at Leesburg (Fig. 4-8c) seemed low. In actuality, the total numbers
captured there, 102, were higher than any other site, but there was almost equal


79
(L.) (Hesperiidae), the long-tailed skipper. Another important migrant in Florida is
Precis coenia Hbner (Nymphalidae), the buckeye, but most of its migration occurs
northward in the spring with the southward fall movement being relatively small. This
pattern is the reverse of what is seen for the other species listed. Together, these four
species make up more than 85% of Walkers (1991) total trap catch. Walker (1991)
captured at least 100 individuals of five other species in his ten years of trapping. Pieris
rapae (L.) (Pieridae) and Vanessa virginiensis (Drury) (Nymphalidae) showed significant
northward movement in the spring, but were seldom caught in the fall. Eurema lisa
(Boisduval & LeConte) (Pieridae) had significant southward movement through
Gainesville in the fall, but was never caught in the spring. Another pierid, Eurema daira
(Godart) moved south in the fall of one year, but showed no statistical bias in either
direction over all years combined. Finally, the pierid, Eurema nicippe (Cramer) was
found in some years to move northward in the fall, a seemingly inappropriate direction.
The monarch passes primarily along the gulf coast to Texas in its migrations south
(Brower et ai, 1985) although, a few individuals do move through north central Florida
(Walker, 1991). Since this species flies at a higher altitude than others (Gibo, 1986),
they are seldom captured in flight traps.
Although the movements of these species have been well documented in
Gainesville by Walker, migratory activities in the rest of the state are virtually unknown.
In particular, the timing, or phenology, of the migration south of Gainesville is not
known, nor are the characteristics of the migratory front, as far as density and species
composition. Arbogast (1966) and Balciunas and Knopf (1977) determined the flight


62
a)
>
O
*
H-
LU
Z
9
Q
<
b)
>
Q
Z
4
h
lil
Z
3
o
<
WEST
WEEK 4
WEEK 6
EAST
o
1 -
2 -
3 -
4
26 SEP
'I
STN CRC TRN
NWB
GK V
4 OCT
XXX
HAW ITL PLK HST CRB
i 1 1 1 1 1 1 1 1 1 1 1 r
0 16 32 46 M 80 M 112 128 144 160 176 192
JN
ts
WEST
KILOMETERS
WEEK 7
WEEK 8
EAST
17 OCT
6
10
15
18 OCT
STN CRC TRN NWB GNV HAW ITL PLK HST CRB
i i i i r
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-9 a,b.-Profile estimates of Agraulis vanillae migration (net number of
individuals/minute, adjusted for time of day, sample size and season) across a west-east
transect at the latitude of Gainesville, Florida (29.65) from 26 September to 18 October
1987. Counts were made on nine transect segments defined by 10 sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


71
Using his numbers for these other two species, the total migration of these four principle
migrants is given as 175 million individuals annually.
Fall 1987 drive counts of Phoebis sennae net migration across west-east transects
south of Gainesville, Florida, during fall 1987 are given in Fig. 3-14 a,b,c,d,e,f. The
net numbers observed in these drive counts were adjusted only for time of day. Visits
to the same transects were approximately 4 weeks apart and week 9 was the only time
that the full north-south range of transects was travelled. Statistically significant
southward movement is evident south to the Crystal River-Leesburg transect (3) during
week 4 (Fig. 3-14a) and continued through week 9 (Fig. 3-14c). It is not known whether
individuals are continually moving southward throughout this time or if this is a
secondary wave of migrants from subsequent generations. From the pattern observed at
Gainesville, there was probably a continual, and gradually diminishing, net southward
movement throughout this period. By week 15 (Fig. 3-14f), this northern area no longer
showed any significant migration. The only time statistically significant migration was
noted south of Crystal River-Leesburg was during week 7 at the Zolfo Springs-Sebring
transect. As is seen with results of trap catches (discussed in Chapter 4), there was no
evidence of large scale migratory movement through south central Florida. Unless the
migratory track deviates from the mean flight direction observed at Gainesville, the
migration appears to diminish south of Leesburg. Small numbers of individuals may be
moving farther south, perhaps with a second generation augmenting populations around
the Zolfo Springs-Sebring area. Drive counts made during fall 1987 depicting the net
migration of Agraulis vanillae net migration across various west-east transects south of


3e
ARCADIA
ALLIGATOR POINT
BARTOW
BELLE GLADE
CROSS CITY

O

O

CLERMONT
Q
CRESCENT BEACH
O
CLEWISTON
o
EVERGLADES CITY

EASTPOINT
O
FROSTPROOF
FISHCREEK GLEN ST. MARY
HAWTHORNE
HASTINGS
GOO
IMMOKALLEE INTERLACHEN JASPER LAKE CITY
O O
o
LEESBURG LAKE ALFRED TABELLE LAKE PLACID NEWBERRY

o o
OLGA OKEECHOBEE PALATKA PERRY STEINHATCHEE
Q CD
Q G
ST. GEORGE TRENTON TALLAHASSEE WHITE SPRINGS YEEHAW JUNCTION
O Q O O
Fig. 2-6.--Azimuths of Urbanus proteus flight directions during fall migrations, 1985-
1988. The circle diameter represents the number of visits (5 mm = l); the length and
width of the line represents the number of individuals (2 mm=l); the orientation of the
lline represents the flight direction.


120
long tailed skipper and buckeye were included, the total number of butterflies migrating
south through the Florida peninsula was estimated to be 175 million individuals annually.
The phenology of the migration was studied by trapping the four major migrant
species using directional flight traps at five stations, from south Georgia to south central
Florida, during the fall of 1987 and 1988. All statistically significant net movement was
southward, and extended to the south central site of Lake Alfred, about 192 km south of
Gainesville. There was no statistically significant movement recorded for the buckeye,
but numbers captured were low. There was also no evidence from flight trap catches or
drive counts at the Lake Placid latitude (27.45) that large numbers of migrants regularly
move into extreme south Florida to overwinter.
The midpoints of migration revealed through fall 1987 and 1988 flight trap
catches suggest several possibilities for the pattern of movement along the Florida
peninsula. The movement along the southward route was closely coordinated in time for
all species and usually occurred within the first three weeks of October. This time span
does not indicate a step-wise generational flight that occured over short distances. One
possibility is that individuals fly long distances along the migratory route. Another
possibility is that populations increase throughout the state during late summer and fall
migration is more or less synchronous over a large area. In this scenario, individuals fly
relatively short distances, followed by a second generation. For two cloudless sulphurs
marked in Gainesville, the duration of travel was longer than expected from the
knowledge of ground speed. These individuals traveled only a distance of 13 and 51 km
in 10 and 14 days, respectively. As an alternative to the two possibilities mentioned,


28
the mean vector (r) with sites labeled in bold letters indicating statistically significant
vector values for those sites with n >5 using Rayleighs z test (Zar, 1984). Gainesville
data is, as previously reported by Walker (1985a), 141. The observations made in this
study along the latitude of Gainesville also show a similar pattern of movement. The
inland sites south of Gainesville correspond to the expected flight direction if migrants
were continuing along the same track observed at Gainesville. Although not statistically
significant, due to small sample sizes at many sites, the movement along coastlines
seemed confused and quite variable. Along the Gulf coast, butterflies usually followed
the coastline even in inappropriate directions. Previous reports of Phoebis sennae flight
directions are southeastward inland and coastwise when near the coast, as reported by
a number of different observers over many years (Shannon, 1916; Williams, 1930, 1958;
Clark and Clark, 1951; Lambremont, 1968; Howe, 1975; Urquhart and Urquhart, 1976;
Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981; Walker, 1985a).
Phoebis sennae is particularly sparse on the east coast considering the southeast
bearings of inland migrants. It has been observed by Walker (1985a) that even 90 km
from the coast, butterflies in south Georgia fly more southerly than further inland, as
though they can detect the coast. This trend is also seen here and no individuals of any
migrant species were observed at Crescent Beach, the easternmost site of the transect.
The reason for an absence of migrants along the east coast portion of the transect is not
clear. Migrating butterflies do tend to follow coastlines possibly using a different
orientation mechanism used from inland flight. If north or south coastline flight is not
distinguished, then by avoiding the east coast altogether, northward flight during the fall,


Fig. 4-14 a,b,c,d,e,f,g,h,i.--Weeklypercentage of total net trap catch of Urbanus proteus
at Valdosta, Georgia; Leesburg, Lake Alfred and Lake Placid, Florida during 28 August-
26 November 1988. Data for Gainesville, Florida are from Walker (1991). Solid bars
represent statistically significant southward biased movement (chi-square test, p < .05).
Shaded bars show nonsignificant (p >.05) southward movement. Clear bars show
nonsignificant (p > .05) northward movement. The total net catch (southbound minus
northbound) for the period is shown in the upper right comer. The midpoint of
migration is designated by an asterisk, x =trap not operating.


Fig. 3-15 a,b,c,d,e.f.Drive counts of Agraulis vanillae migration (net number of
individuals/minute, adjusted for time of day) across various west-east transects south of
Gainesville, Florida during fall 1987. Transects with a southward or northward net
migration of statistical significance (chi-squared test, p < .05), are shown with half filled
circles weighted in the appropriate direction. Net movement, either north or south, that
is not statistically significant (p < .05), is depicted with open circles. Lines with no
circles represent transects where there were no individuals seen. Number of individuals
and percentage flying south are given in parentheses after each transect.


for field assistance with traps, Dr. Nigg for a tour of the Green Swamp and Dr. Deyrup
for help during hurricanes and otherwise.
My first interest in butterflies and their migration through the Florida Peninsula
was inspired by the late Dr. Dennis Leston, in whose memory this work is completed.
His enthusiasm for nature and love for insects will remain with me always. I am greatly
indebted to my parents, Lucjan and Stanislawa Lenczewski, who have encouraged my
education. My sister, Eva Lenczewski, despite her dislike of insects, has sometimes
even helped with this perversity of mine and she deserves recognition for her courage.
Finally, and most importantly, I thank my husband, Joseph Jowers, for a great deal of
assistance and companionship in the field. He and my son, Justin, held down the home
front to really make my work possible.
IV


86
Fig. 4-2.-Summary of pole counts of Phoebis sennae (average net number of
individuals/meter/minute, adjusted for time of day) through Gainesville, Clermont,
Arcadia, LaBelle, and Immokalee, Florida during 11 September through 8 November
1986. Numbers below the bars represent total number of counts at that site.


41
Crescent Beach, a distance of about 200 km. All results of pole counts were summarized
by averaging the net numbers of individuals/m/minute from samples made at each site
throughout the season.
Migration Profiles
During the fall migrations of 1986, another technique was devised to estimate
the net migration across west/east transects. Counts of north and south flying butterflies
were made while driving along selected roads running west to east. This method covered
entire transects rather than the 15 or 45 m samples made during the pole counts.
Transects with low vegetation and therefore good visibility were selected. The major
transect was at the latitude of Gainesville, through the same sites as the pole counts.
Ideally, the entire transect was travelled roundtrip once per week, with half the counts
completed on a particular day. Only cloudless sulphurs and gulf fritillaries were included
in the drive counts because they were easily identifiable at a distance. To be counted,
a butterfly had to at least have reached the centerline of the road before the car passed.
A speed of approximately 97 kph was maintained and counts were recorded for every 16
km of transect from which the net number of indiv/min was calculated. Standard time
was noted at the beginning and end of each segment of transect, as was weather.
Weather conditions recorded were the percentage of blue sky visible, the appearance of
the sun (b=bright, h=hazy, o=obscured, p=disk obscured, but position easily
determined 5), air temperature (C), wind direction and speed (m/sec with a hand
held anemometer).


84
a)
GAINESVILLE
a)
WEEKS
LABELLE
b)
d)
J
g)
CLERMONT
WEEKS
ARCADIA
XX XX
10
WEEKS
WEEKS
Fig. 4-1 a,b,c,d,e,f.--Pole counts of Phoebis sennae movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September through 7 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p <.05). x=no counts made.


GEORGIA
FLORIDA
A Arcadia LE
B Bartow LA
C Cross City |_B
CB Crescent Beach LP
CL Clermont N
G Gainesville P
H Hawthorne S
HA Hastings j
I Interlachen v
IM Immokalee
Leesburg
- Lake Alfred
- LaBelle
- Lake Placid
- Newberry
- Palatka
Steinhatchee
- Trenton
- Valdosta
Fig. 1.1-Location of pole count and flight trap sampling locations used during the fall of 1986, 1987 and 1988.


115
a)
c)
WEEKS
LEESBURG TRAP 1
WEEKS
LEESBURG TRAP 2
g)
1 2 3 4 5 6 7 6 9 1011 1213
WEEKS
LAKE ALFRED TRAP 1
WEEKS
LAKE ALFRED TRAP 2
1 23456789 1011 1213
WEEKS


Fig. 3-8 a,b,c,d,e,f,g,h.--Profile estimates of Apraulis vanillae migration (net number
of individuals/minute, adjusted for time of day, sample size and season) across a west-
east transect at the latitude of Gainesville, Florida (29.65) from 13 September to 28
October 1986. Counts were made on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW=Hawthorne, ITL=Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. x=unsampled sites.


46
numbers of individuals were counted through the central sites of Newberry, Gainesville
and Hawthorne, averaging about .45 indiv/m/min. The highest count was through
Gainesville at .57 indiv/m/min and numbers declined toward both the Gulf and Atlantic
coasts.
Gulf fritillarv
Pole counts taken of the gulf fritillary migration during fall 1986 are shown in
Fig. 3-2. All net movement was southward and the peak migration was concentrated
slightly more to the west, passing through Trenton, Newberry and Gainesville. Peak
numbers were recorded in Trenton at .58 indiv/m/min followed by Newberry at .48
indiv/m/min. Whereas the cloudless sulphur had little or no migration on either coast,
the west coast migration for the fritillary was relatively strong (.3 indiv/m/min). Few
individuals were counted at Hastings, which was also the only site lacking statistical
significance to the southward net movement. At Crescent Beach however, directly along
the coast, movement was recorded at .13 indiv/m/min. The greater presence along the
coastlines of this species corresponds to the gulf fritillary buildup along the panhandle
coast.
Long tailed skipper
The summary for fall 1986 pole counts of long tailed skipper net movement
across the west-east transect at the latitude of Gainesville is shown in Fig. 3-3. The
greatest number of individuals move south through the central sites of Gainesville-
Hawthome (.21 indiv/m/min), but migration of Urbanus proteus is also substantial along
both the Gulf and Atlantic coastlines (. 14 and 13 indiv/m/min, respectively). Centrally


16
ranging throughout tropical America, from Argentina north to the southern United States
(Howe, 1975) and throughout the West Indies (Riley, 1975). The eastern subspecies,
nigrior. occurs from Florida, west to Louisiana, north to North Carolina and is also
found on Bermuda. Although this species has been reported as far north as New York
(Howe, 1975), it is not commonly found above southern Virginia (Clark and Clark,
1951) on the eastern coast. Unlike the cloudless sulphur, this species is not usually seen
during winter months (December-February) in northern Florida. However, there was an
exceptionally warm winter in Gainesville during 1986-87 and adults were seen flying in
January (pers. obs.).
Like the other heliconiids, the larvae of Agraulis vanillae feed only on species
of Passiflora. In Georgia and north central Florida, the caterpillars are commonly found
on the passion flower, Passi flora incamata L. (pers. obs.). The larva has a black body,
with three pairs of dorso-lateral, orange-brown stripes. There are six rows of branching
tubercles on the body and a pair of longer tubercles curving back on the head. The adult
is distinctively colored with bright orange upper wings marked with black and the
hindwing underside covered with metallic silver spots and they are particularly attracted
to the flowers of Spanish needle, Bidens pilosa and lantana, Verbena spp. According to
Klots (1951), there are three or more broods in the south. In Gainesville, Florida, the
butterflies arrive and begin to oviposit as soon as plants are available in early spring,
continuing to breed until the winter months, usually late November (pers. obs.).


N
A=Arcadia n=9,r=81,168
AP=Alligator Point n=0
B=Bartow n=8,r=.66,156
BG=Belle Glade n=0
C=Cross City n=12,r=.82,147
CB=Crescent Beach n=0
CL=Clermont n=3,r=.82,147
CW=Clewiston n=0
E=Everglades City n=6j=.48.121
EP=East point n=0
F=Frostproof n=2,r=.94,143
FC=Fishcreek n=2,r=.64,318
G=Gainesville (Walker, 1985a)
GS=Glen St. Mary n=0
H=Hawthome n=6,r=.89,157
HA=Hastings n=3j=.31,338
I=Interlachen n=3,r=.67,135
IM=lmmokalee n=ll j=.68,177
J=Jasper n=l,r=l,152
KB=Keeton Beach n=22,r=.28,168
L=Lake City n=0
LE=Leesburg n=12j=.89,145
LA=Lake Alfred n=2j=l,347
LB=LaBelle n=7j=.53,215
LP=Lake Placid n=l,r= 1,277
N=Newberry n=l,r= 1,280
0=01ga n=5,r=.87,168
OK=Okeechobee n=6,r=.97,186
P=Palatka n=o
PE=Perry n=2,r=l,162
S=Steinhatchee n=23,r=.91,120
SG=SL George n=0
T=Trenton n=3,r=.98,152
TA=Tallahassee n=0
W=White Springs n=0
YJ=Yeehaw Junction n=13j=.55,144
Fig. 2-5.--Summary of mean flight directions at various sites throughout Florida of Urbanus proteus during fall migrations of 1985-
1988. (The direction of the arrow=mean flight direction, value given in key for each site; the length of line=mean vector r; bold
letters = significant r value). w


82
period. Walker and Lenczewski (1989) estimated that semi-weekly service recovered at
least 90% of all individuals entering the traps collecting cages. After testing other
construction materials in the spring of 1988, the traps were modified, primarily to
increase durability. The backs of eight polyester traps were replaced with monofilament
shrimp netting. Two of the polyester netting traps were left intact. An additional four
traps were built of the same design but with all monofilament shrimp netting. All these
traps were found to be similar in efficiency (Walker and Lenczewski, 1989), but the
monofilament shrimp netting was more durable and easier to work with. In 1988, traps
were in place from 28 August through 26 November.
The sites were 157, 100, 70 and 92 km apart, respectively, a total of 419 km
from northernmost to southernmost site. The types of traps used each year, at each site,
are as follows:
1. Valdosta State College, Valdosta, Georgia (30.85 lat, 83.29 long).
1987 two all polyester traps, 1988 four polyester traps with shrimp
net backing.
2. Green Acres Farm, University of Florida, Gainesville, Florida (29.65
lat, 82.44 long). 1987 two all polyester traps. 1988 no traps at this
location, data used from Walker (1991).
3. IFAS/AREC, University of Florida, Leesburg, Florida (28.90 lat,
81.54 long). 1987 two all polyester traps, 1988 four polyester traps
with shrimp net backing.
4. IFAS/CREC, University of Florida, Lake Alfred, Florida (28.00 lat,


7
migratory route for birds (National Geographic, 1979). It seems to function this way,
in some respects, for migrant butterflies also. Large numbers of monarchs, gulf
fritillaries and long tailed skippers are often seen feeding at Baccharis halimifolia along
the Gulf coast, particularly in the St. Marks or Appalachicola area (pers. obs.) during
the fall. Cloudless sulphurs are also present, but not in such great numbers as the other
species and there are some indications (Walker, pers. comm.) that they avoid coastlines.
Despite the intriguing presence of these large aggregations of migrant butterflies which
are obvious to the most casual observer, the subsequent movements of these individuals,
except for the monarch, remain largely unknown. These species often exhibit erratic
flight near the coastline, thus making it very difficult to find a preferred pattern. There
are reports of movements west, following the coastline (Urquhart and Urquhart, 1976),
but I have seen migration in both directions along the coast and also south across the
water, as well as north back to land. The migrants seen massed along the Gulf are
spectacular, and understandably, have attracted the most attention. Away from the Gulf
coast, most north Florida residents realize that the numbers of these species increase in
the fall, but are generally unaware of their persistent southward flights. Once the
butterflies pass into the Florida peninsula, we know virtually nothing of the migration
beyond Gainesville (Walker, 1991).
Most migrants pass through north peninsular Florida during September through
November and it has been presumed they join resident populations in subtropical south
Florida. Although the migratory flights are so conspicuous through north Florida during
fall, this does not seem to be the case in the southern parts of the state. Although the


65
b)
KILOMETERS
1987
0 16 32 48 64 80 96 112128144160176192
KILOMETERS
1988
STN CRC TR NWB GNV HAW ITL PLK HST CRB
i i "T 1 T ri T~"T 1 i i
0 16 32 48 64 80 96 112128 144 1 60 1 76 1 92
KILOMETERS


4
Migrating butterflies can maintain their compass direction and therefore must
have some orientation mechanism. It is possible that a time-compensated sun compass
is used (Walker, 1985a; Baker, 1978), but although Baker (1968a) has shown possible
azimuth orientation to the suns position, by Pieris rapae. this remains to be proven
(Able, 1980). That fall migrants can maintain their migratory direction on overcast days,
although not many fly then, weakens, but does not eliminate, this theory (Verheijen,
1978). Another possibility is that they have a magnetic compass. Evidence of
geomagnetic orientation has been found in the underwing moth, Noctua prnuba L.
(Baker and Mather, 1982), and is well known in honey bees (Lindauer and Martin,
1972). Magnetic particles have been detected in the monarch butterfly, although their
function has not been demonstrated (MacFadden and Jones, 1985). Jungreis (1987)
tested two of the Florida migrant butterflies, Phoebis sennae (L.) and Agraulis vanillae
(L.), and found no evidence of magnetic particles.
Migration requires a high degree of coordination among behavior, morphology
and metabolism to function as an adaptive life history strategy. Migrant insects must
undergo physiological changes that can sometimes be used to identify their migratory
tendency. One metabolic preparation for sustained flight that has been studied is energy
storage. A primary source of energy is fat, or lipids (Beenakkers, 1965), and increased
levels have been found in migrating locusts and monarch butterflies (Johnson, 1974).
It is known that monarchs build up these lipid stores prior to fall migration and deplete
them during long flights to Mexico (Beall, 1948; Brower, 1977). Whether they
replenish, or slow, the depletion of these stores by nectar feeding enroute has still not


24
There are at least eight species of migrating butterflies that pass throughnorth
central Florida (Walker, 1978, 1980, 1985ab, 1991) each fall. The four most obvious
and abundant species are Phoebis sennae (L.) (Pieridae), the cloudless sulphur; Agraulis
vanillae (L.) (Heliconiidae), the gulf fritillary; Urbanus proteus (L.) (Hesperiidae), the
long-tailed skipper; and Precis coenia (Hbner) (Nymphalidae), the buckeye. Outside
of Walkers studies (1978, 1980, 1985ab, 1991; Walker and Riordan, 1981), mostly
based in Gainesville, Florida, there is very little known about the flight directions of
these species. The conspicuously yellow cloudless sulphur, especially, has been reported
travelling in southeasterly fall flights throughout the southeastern states (Lambremont,
1968; Shannon, 1916; Williams, 1930, 1958; Clark and Clark, 1951; Howe, 1975;
Urquhart and Urquhart, 1976; Muller, 1977; Gaddy and Laurie, 1983; Pyle, 1981).
Although there are reports of inappropriate northern flights (Gaddy and Laurie, 1983;
Muller, 1977), or variable movements, especially along coastlines (Urquhart and
Urquhart, 1976), the majority of individuals appear to be navigating toward the Florida
peninsula (Walker, 1985a). Walkers trapping and observational data (1978, 1980,
1985ab, 1991) clearly demonstrate that the migrants pass through Gainesville, in north
central Florida, each fall and maintain an average flight path approximating 141.
However, for most of the rest of the state, flight paths are unknown. My study was
undertaken to investigate previously unknown migratory flight directions of the four
principal migrant species at various sites throughout the Florida peninsula.


19
temperature from below 16C to above 16C and from above 27C to below 27C and
not with rainfall, as previously suspected. Scott (1975) investigated the issues of male
territoriality and migratory tendency which seem to be at theoretical odds. To the casual
observer, buckeye males seem highly territorial, dashing from their resting places, to
chase intruders or meet females. Scott (1975) concluded that it is a "psuedo-territorial"
tendency and is not maintained for any significant length of time at a particular site.
The adults are often seen in open country, especially sitting along dirt roads near
disturbed, weedy fields. Three broods occur in the southeast (in Virginia from late May
to late fall). Although the butterflies of the last brood may be seen on warm days
through the first half of winter, they are not found in the spring in Virginia (Clark and
Clark, 1951). This species is considered a summer breeding resident in Virginia and
there are no records of overwintering individuals. However, Clark and Clark (1951)
report having seen ragged, overwintering individuals in March, April and early May in
Washington, D.C. Howe (1975) describes the adults as hibernating and sometimes
migrating. In southern Florida at least, the subspecies, P. coenia coenia. is reported for
every month of the year (Lenczewski, 1980).
The larva has a dark gray body, striped or spotted with orange-yellow. There
are a number of short, branching spines on the body and one pair on top of the head.
The larval foodplants are Ruellia (Acanthaceae); Plantago (Plantaginaceae); Antirrhinum.
Buchnera. Gerardia harperi. Linaria. Mimulus. Scrophularia lanceolata and others
(Scrophulariaceae); Ludwidgia (Onagraceae); Sedum (Crassulaceae); Verbena prostrata


74
Gainesville, Florida are shown in Fig. 3-15 a,b,c,d,e,f. A pattern similar to P. sennae
is seen for this species which also demonstrated a statistically significant southward
movement from Inglis-Belleview through Crystal River-Leesburg during week 4.
Subsequently, the farthest significant southward movement seen occurred during week
7 through Bartow-Lake Wales. Again, since the more northern transects were not
sampled at this time, it is not known if southward migration continued there during this
period, but presumably it did. Significant migration was still seen through the north
central transects during week 9 and, from the patterns seen along the Gainesville transect
(Fig. 3-11), it is likely that this continued. No individuals were observed on any trips
south of Gainesville (Fig. 3-15 d,e,f) after this time and Bee Ridge-Lake Placid also
showed no significant migration throughout the season for this species.
The area known as the Green Swamp is located between transects 5 and 7. It
is a large area maintained as wilderness, with some grazing areas, for cattle. The wet
areas along the Withlahochee River support a variety of flowering plants, especially in
the disturbed habitats where cattle are allowed to graze. Large numbers of migrants have
been sighted there during some years, particularly in late November (H. Nigg, pers.
com.). This largely unexplored area may contain many resources and provide an ideal
environment for adult winter survival. There is evidence from Gainesville (Chapter 4),
that some cloudless sulphur individuals will remain in an area with nectar sources for
several weeks.


ACKNOWLEDGEMENTS
A number of people were helpful during this research. Foremost, I would like
to thank my major professor, Dr. Thomas J. Walker, for supplying the materials and
workshop used in construction of the portable traps and for providing expertise and labor
during their construction. The funds for travel and other expenses were arranged by Dr.
Walker and he provided a great deal of enthusiasm and encouragement for this project
throughout many years. I am also indebted to Dr. J. J. H. Frank, who, through the
Mole Cricket Biocontrol Project, provided the opportunity for a graduate assistantship
and facilitated the loan of travel vehicles. Dr. Lincoln P. Brower provided many
opportunities for "butterfly" socializing, and the use of his property for mark-release-
recapture studies, during which he kindly made many valuable observations. Dr. Brower
and Dr. Frank Slansky generously allowed me the use of their laboratories and
equipment. I also thank the other members of my committee, Drs. J. E. Lloyd and J. L.
Nation, for their helpful suggestions and a critical review of the manuscript.
The following people made the trapping study feasible by providing locations
for, and assistance with, flight traps. I am particularly grateful to Dr. J. Whitesell,
Valdosta State College; Dr. G. W. Ellstrom and his assistant, Ms. Annette Chandler,
IFAS/AREC Leesburg; Dr. Mohammed Abou-Setta and Dr. H. N. Nigg, IFAS/CREC;
and Dr. Mark Deyrup, Archbold Biological Station. Special thanks go to Dr. Whitesell
in


Table 1-1. Calendar dates of sampling and the corresponding assigned week number for
all sampling methods used during 1986-1988.
1986
1987
1988
WEEKS
1
01-07 SEP
AUG 30-05 SEP
AUG 28-03 SEP
2
08-14
06-12
04-10
3
15-21
13-19
11-17
4
22-28
20-26
18-24
5
29-05 OCT
27-03 OCT
25-01 OCT
6
06-12
04-10
02-08
7
13-19
11-17
09-15
8
20-26
18-24
16-22
9
27-02
25-31
23-29
10
03-09 NOV
01-07 NOV
30-05 NOV
11
10-16
08-14
06-12
12
17-23
15-21
13-19
13
24-30
22-28
20-26
14
29-05 DEC
27-02 DEC
15
06-12
03-10
months. Generally, the species occurs in the Mississippi Valley to Central Illinois and
along the Atlantic coastal plain to New Jersey (Calhoun s jd, 1990). It is known to
reproduce as far north as west central Illinois (Sedman and Hess, 1985) and Virginia
(Clark and Clark, 1951). In Florida, this species has been recorded for every month of
the year throughout the state, but individuals are most common during the late summer
and fall in north central Florida (Lenczewski, 1980; D. Baggett, pers. comm.). This
subspecies is also resident in the states adjacent to Mexico (Howe, 1975). During some
years, individuals are frequent in eastern Kansas and Missouri where they may be seen
flying in a southeast direction in late summer (Howe, 1975). Similar flights have been


104
c)
e)
Z-2
to
I:
i -6
o a
S 8
* 1
tjl.2
BARTOW
l
1
WEEKS
*
Fig. 4-11 a,b,c,d,e,f. Pole counts of Urbanus proteus movement (net number of
individuals/meter/minute, adjusted for time of day) along a north-south transect through
Gainesville, Clermont, Bartow, Arcadia, LaBelle and Immokalee, Florida from 11
September-8 November 1986. Solid bars represent a statistically significant net
movement (chi-square test, p > .05), shaded bars represent net movement which was not
significantly biased (chi-square test, p < .05), x=no counts made.


WEST
N
f
S
KILOMETERS
WEST
WEEK 5 : 20 SEP
0
2
4
5
8
10 k
TRIP 1
I
I I
gpispc TF|M
TRIP 2
| I
0 16 32 48 64 80 06 0 16 32 48 64 80 06
KILOMETERS
b)
I
a
z
*
t
<)
EAST
WEEK 4 : 25 SEP
TRIP 1
TRIP 2
I"
INVHAWm. PLK HST CRB
1 ""
JNV HAW m. PLK HST CRG
112120136152168184 200112128144160176192
KILOMETERS
EAST
WEEK 6:12 OCT
+
KILOMETERS
Fig. 3-10 a,b,c,d. Profile estimates of Agraulis vanillae migration (net number of individuals/minute, adjusted for time of day,
sample size and season) across a west-east transect at the latitude of Gainesville, Florida (29.65) from 18 September to 12 October
1988. Counts were made on nine transect segments defined by ten sites: STN=Steinhatchee, CRC=Cross City, TRN=Trenton,
NWB=Newberry, GNV=Gainesville, HAW=Hawthome, ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.
x=unsampled sites.


45
STNCRC TRN NWB GNV HAW ITL PLK HST CRB
0 16 32 48 64 80 96 112 128 144 160 176 192
KILOMETERS
Fig. 3-1.--Summary of pole counts showing Phoebis sennae migration (sum of time
adjusted net number of individuals/meter/minute/number of visits) in north central
Florida along the latitude of Gainesville (29.65). Solid bars represent statistically
significant (chi-square, p < .05) southward movement and open bars were sites with no
statistical bias (chi-square, p <.05) in net movement. Counts were made during 1
September to 2 November 1986 at ten sites: STN=Steinhatchee, CRC=Cross City,
TRN=Trenton, NWB = Newberry, GNV = Gainesville, HAW = Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach. The number
of visits to each site are shown above the bars.


40
observation period. Cloudless sulphurs were easily observed at 45 meters and therefore
individuals flying between the first and third poles, were counted. Long-tailed skippers,
gulf fritillaries and buckeyes were less visible and were counted if within 15 meters
(between the first and second poles). Counts were made at least once every three weeks
from 1 September to 2 November 1986 at most of the ten sites (see Table 1.1 for
assigned weeks and Fig. 1-1 for location of sites). This period has been shown by
Walker (1991) to include >95% of the total migration through Gainesville.
Observations were adjusted for time of day differences by using a "time-of-day"
factor, T, from Walkers (1985a) observations of each species flight variation throughout
the day. The average for each half hour was calculated from three days of observation
(3 October 1982 and 4, 11 October 1983) by Walker of percent individuals flying
throughout the day. This average percent of individuals flying was then multiplied by
the T factor which would adjust up to 100% for that time period. A 15 minute
adjustment to T was made by averaging two adjacent half hourly observations. For
example, if 6.4% of the days total was the average seen flying at 1000 HRS (EDT)
(counts made between 0945-1014) then T would be 15.6. An average of 5.9% flying at
1030 HRS (counts between 1015-1044) results in a T of 16.9. To adjust for 15 minute
periods, the average percent flying for both half hourly observations was averaged with
a result of 6.15, the resulting T factor for observations at 1015 HRS would then be 16.3.
Observations were grouped according to their closest time period in 15 minute intervals
and multiplied by the appropriate T factor. The west-east transect sites were
approximately 23 kms apart and extended the width of the state from Steinhatchee to


99
rate from egg to adult at 28.5-29.5 C for this species is approximately 23 days. The
observed situation at Leesburg may indicate that adult butterflies, presumably having
arrived from Gainesville north, have ceased migrating and are utilizing available
hostplants for oviposition. From known estimates of developmental rates and flight
speeds, it is possible that a second generation travels southward to Lake Alfred.
Although a few individuals were seen at Lake Placid, there was no significant migration
recorded during the fall and only one individual was captured heading south during week
14.
The percent of the total net trap catch of gulf fritillaries during fall 1988 is
shown in Fig. 4-9 a,b,c,d,e,f,g,h,i. As seen in 1987, all statistically significant (chi-
square, p < .05) net movement throughout the season was southward at each site. Also
as seen the previous year, mid-migration at Valdosta trap 1 (Fig. 4-9a) was during week
5 and this was also the peak for trap catches during the 6 weeks sampled with trap 2
(Fig.4-9b). Walkers (1991) data during 1988 are given for Gainesville (Fig. 4-9 c,d)
and the midpoint is shown as week 6. Leesburg traps 1 and 2 (Fig. 4-9 e,f) had
corresponding midpoints also during week 8. Whereas in 1987, the only southward
biased movement was during week 5, 1988 catches from traps 1 and 2, respectively,
revealed a full 8 and 4 weeks of significant net southward movement. Movement during
1988 through Leesburg was less bidirectional than the previous year and the migration
midpoints were much closer to the values for Lake Alfred (Fig. 4-9 g,h). There was still
a three week difference in the midpoint of migration between Valdosta and the mid
central site of Lake Alfred, and this year, also Leesburg. Again, this spacing of


ADJ NET # INDIV/MIN SOUTH ADJ NET INOIV/MIN SOUTH
57
KILOMETERS
KILOMETERS
c)
KILOMETERS


110
VALDOSTA TRAP 1
3 4 5 6 7 8 9 1011 1213
WEEKS
VALDOSTA TRAP 2
2 3 4 5 6 7 8 9 1011 1213
WEEKS
C) GAINESVILLE PTP3 (WALKER, 1991)
WEEKS
WEEKS
9) LAKE ALFRED TRAP 1
WEEKS
d) GAINESVILLE PTP5 (WALKER,1991)
WEEKS
WEEKS
") LAKE ALFRED TRAP 2
50
0 * spa is^a
1 2 3 4 5 6 7 8 9 1011 1213
WEEKS
WEEKS


Fig. 3-7 a,b,c.--Migration profiles summarizing estimates of Phoebis sennae migration
across a west-east transect in north central Florida along the latitude of Gainesville
(29.65). Yearly migration profiles were derived from the median estimate (net number
of individuals/minute, adjusted for time of day, sample size and season). Counts were
made from 13 September-28 October 1986, 26 September-18 October 1987 and 18
September-12 October 1988 on nine transect segments defined by ten sites:
STN=Steinhatchee, CRC=Cross City, TRN=Trenton, NWB=Newberry,
GNV = Gainesville, HAW = Hawthorne, ITL = Interlachen, PLK = Palatka,
HST=Hastings, CRB=Crescent Beach. Solid bars represent statistically significant (chi-
square, p < .05) southward movement and open bars were sites with no statistical bias
(chi-square, p < .05) in net movement. Total number of visits made to each site are
shown above the bars.


130
Southwood, T.R.E. 1962. Migration of terrestrial arthropods in relation to habitat.
Biol. Rev. 37:171-214.
Stinner, R.E., C.S. Barfield, J.L. Stimac and L. Dohse, 1983. Dispersal and migration
of insect pests. Annu. Rev. Entomol. 28:319-335.
Taylor, L.R. 1958. Aphid dispersal and diurnal periodicity. Proc. Linn. Soc. (Lond.)
169:67-73.
Taylor, R.A.J. and D. Reling, 1986. Preferred wind direction of long distance
leafhopper (Empoasca fabael migrants and its relevance to the return migration
of small insects. J. Anim. Ecol. 55:1103-1114.
Urquhart, F.A. 1960. The Monarch Butterfly. Univ. of Toronto Press, Toronto. 361
pp.
Urquhart, F.A. and N.R. Urquhart, 1976. Migration of butterflies along the gulf coast
of north Florida. I. Lep. Soc. 30:59-61.
Verheijen, F.J. 1978. Orientation based on directivity, a directional parameter of the
animals radiant environment. Pp. 447-458 in Animal Migration, Navigation,
and Homing, K. Schmidt-Koenig and W. T. Keeton (eds.) Springer-Verlag,
Berlin. 462 pp.
Walker, T.J. 1978. Migration and re-migration of butterflies through north peninsular
Florida: quantification with Malaise traps. J. Lep. Soc. 32:178-190.
Walker, T.J. 1980. Migrating Lepidoptera: are butterflies better than moths? Fla.
Entomol. 63:79-98.
Walker, T.J. 1985a. Butterfly migration in the boundary layer. Pp. 704-723 in
Migration: Mechanisms and Adaptive Significance, M. A. Rankin (ed.), Contrib.
Mar. Sci. 27 (Suppl.). 868 pp.
Walker, T.J. 1985b. Permanent traps for monitoring butterfly migration: tests in Florida
1979-84. I Lep. Soc. 39:313-320.
Walker, T.J. 1991. Butterfly migration from and to peninsular Florida. Ecol. Entomol.
16:241-252.
Walker, T.J. and B. Lenczewski, 1989. An inexpensive portable trap for monitoring
butterfly migration. J. Lep. Soc. 43:289-298.


2
These highly visible insects travel within the boundary layer, the layer of air near the
ground where wind velocity is less than the insects air speed, and flight direction can
be controlled by the individual. The thickness of this boundary layer is variable,
determined by wind velocity and the air speed maintained by the individual (Taylor,
1958; Pedgley, 1982). This determined choice of flight track is perhaps easiest to
observe in butterflies and demonstrates that migration can be more under the individual
control of these, and perhaps other, insects than previously suspected.
There has been considerable controversy in theoretically extricating dispersion
from migration. Baker (1978) has suggested that there is a continuum in the expression
of the migratory habit, ranging from random movement to the evolution of highly
specific destinations with correspondingly adaptive physiological changes. In a recent
summary, Dan than arayana (1986, p. 1) claimed a consensus has emerged on the
terminology used to describe insect movements. His definition of migration will be
accepted for the purposes of this study:
Non-migratory movements involve travel within the habitat associated with such
activities as feeding, mating and opposition ... In contrast, migratory
movements take insects beyond the habitat for the purposes of colonizing new
habitats, re-colonizing old ones, aestivation or hibernation ....
The defining criterion is travel outside what is, or had previously served, as the
sustaining "habitat" to more favorable conditions. The behavior and physiology involved
in getting to these new habitats fall somewhere on a continuum in the evolution of the
migratory habit. Some insect migrants, such as the monarch butterfly, may be
reproductively inactive during these movements and maintain long distance flights on
stored body fats, but in others, mating and feeding occur in conjunction with the


121
some individuals may have a facultative migration during which they follow a wave of
short-lived larval and adult resources as they are available. The bimodal pattern noticed
in the seasonal number of individuals caught in flight traps, two peaks separated by about
three weeks, suggests that a second wave of migrants, probably offspring, move along
later in the season.
While migrating, these species remain reproductively active. Most females are
mated and will oviposit abundantly on available hostplants. In north Florida, the larval
hostplants senesce shortly after the migration peaks and these, as well as adult nectar
sources, may be limited resources for these species. In fact, some cloudless sulphur
individuals marked in Gainesville remained near the same areas with abundant nectar
sources, at least through December. As mentioned previously, the cloudless sulphur and
buckeye are capable of regularly overwintering in north Florida and the adjacent coastal
plain. During unusually warm winters, a few gulf fritillaries may survive in north
Florida, but this is not common.
An overall picture of the butterfly migration through the Florida peninsula is still
elusive and requires more detailed studies in many areas throughout the state. The
question of just how far individuals migrate and at what rate and consistency is still
unclear. Although the trapping results of this study suggest that movement south occurs
in a closely coordinated manner, it does not rule out the possibility that migration is
synchronous over a large area. The availability of larval hostplants and adult nectar
sources seem to play an important role, particularly in the migration of the cloudless
sulphur. Hostplant phenology needs to be closely monitored throughout the year on a


42
During this study, the analysis of the data presented several problems incurred
by the distance and time required for sampling and daily and seasonal changes in the
density of migration. Walker and Riordan (1981) have shown that by sampling only on
days that meet certain weather criteria, day-to-day fluctuations in numbers can be
minimized. Driving periods were limited to weather conditions favorable to migration
and were terminated if more than 50% of the sky was obscured by clouds, the
temperature dropped below 21 C or wind speed exceeded 3 m/sec. There are however,
hourly differences in migration throughout the day with peak flight times usually from
1000 to 1400 HRS EDT (Walker, 1985). Since it was not possible to sample
simultaneously along the transect sections, it was necessary to accommodate the daily
changes in peak flight periods. Adjustments for time of day differences were made to
the actual net numbers observed during drive counts (D) by multiplying with the "time
of day" factor, T (see explanation for pole counts above).
As a result of time limitations, unpredictable weather changes and other
occurrences during sampling periods, not all segments along a transect were always
visited on a particular sampling trip. To equalize for the differences in sample number
from each transect segment, the time adjusted counts (DnT) were then transformed
relative to a time corrected "base" segment (DjT) which was always sampled during each
trip. For the western portion of the transect (Gainesville-Steinhatchee), this base was the
Gainesville-Newberry segment. For the eastern portion of the transect (Gainesville-
Crescent Beach), the base was the Gainesville-Hawthome segment. The net number of


129
Roer, H. 1962. Experimentelle Untersuchungen zum Migrationsverhalten des Kleinen
Fuchs (Agais urticae L.). Beitr. Entomol. 12:528-554.
Roer, H. 1968. Weitere Untersuchungen uber die Auswirkungen der Widening auf
Richtung und Distanz der Fluge des Kleinen Fuchses (Agais urticae L.) (Lep.
Nymphalidae) im Rheinland. Decheniana 120:313-334.
Roer, H. 1969. Zur Biologie des Tagfauenauges, Inachis io L. (Lep. Nymphalidae).
unter besonderer Berucksichtigung der Wanderungen im mitteleuropaischen
Raum. Zool. Anz. 183:177-194.
Roer, H. 1970. Untersuchungen zum Migrationsverhalten des Trauermantels Nvmphalis
antiopa L.) (Lep. Nymphalidae). Z. Angew. Entomol. 4:388-396.
Roff, D.A. 1975. Population stability and evolution of dispersal in a heterogeneous
environment. Oecolopia 19:217-237.
Schaefer, G.W. 1976. Radar observations of insect flight. R. Entomol. Soc. (Lond.)
7:157-159.
Schmidt-Koenig, K. 1975. Migration and Homing in Animals. Ser. Zoophysiology and
Ecology 6. Springer-Verlag, Berlin. 99 pp.
Scott, J.A. 1975. Movement of Precis coenia. a "pseudoterritorial" submigrant
(Lepidoptera: Nymphalidae). I. Anim. Ecol. 44:843-850.
Scott, J.A. 1986. The Butterflies of North America: A Natural History and Field Guide.
Stanford Univ. Press, Stanford, California. 583 pp.
Sedman, Y. and D.F. Hess, 1985. The Butterflies of West Central Illinois. W. Illinois
Univ., Ser. Biol. Sci. No. 11. 120 pp.
Shannon, H.J. 1916. Insect migrations as related to those of birds. Science Monthly
3:227-240.
Showers, R.B. Smelser, A.J. Keaster, F. Whitford, J.F. Robinson, J.D. Lopez and S.E.
Taylor, 1989. Recapture of marked black cutworm (Lepidoptera: Noctuidae)
males after long-range transport. Environ. Entomol. 18:447-458.
Solbreck, C. 1985. Insect migration strategies and population dynamics. Pp. 641-662
in Migration: Mechanisms and Adaptive Significance, M.A. Rankin (ed.),
Contrib. Mar. Sci. 27 (Suppl). 868 pp.


17
Urbanus Proteus
Urbanus proteus. the long tailed skipper, is a minor pest of cultivated beans,
known as the bean leaf roller in agricultural literature. This species is distributed from
Argentina through the United States and West Indies. The mainland populations
comprise the nominate subspecies, with all of the insular populations falling into a second
subspecies (Howe, 1975). In the United States, U. proteus proteus is found from
Connecticut, south to Florida and west to Texas, Arkansas, Arizona and California. The
larvae are green and have a dark mid-dorsal line with yellow lateral lines and green
stripes below these. The area between the stripes is dotted with black and yellow spots.
The head is large, reddish brown, with two yellow spots between the ocelli and the
mouthparts. The caterpillars come out to feed from leaf shelters that they construct on
their leguminous hostplants, moving to a larger one at each instar. The major foodplants
include such legumes as Bauhinia. Clitoria mariana. Desmodium. Phaseolus. Soja,
Vigna. Wisteria and Pueraria lobata (Lenczewski, 1980). In the adult, the thorax, wing
bases and hind wings are a conspicuous metallic green. There are four separate square
spots on the central band of the forewing and a larger square spot under the end of the
cell. The male has a costal fold, the antennal club is yellowish brown below and the
range of wingspread is 38-50 mm (Howe, 1975). Adults are found flying in open areas,
particularly fields and along roadsides in disturbed sites. Three generations are reported
in Florida but only representatives of the late-summer brood arrive in Virginia where this
species is considered a casual visitor (Clark and Clark, 1951). Although Howe (1975)
claims it occurs throughout the year, there are no records during the summer months of


68
O On i i i i" i i' i i i i i 1 r
< O 16 32 48 64 80 96 112 128 144 160 176 192 (KM)
1986-1988
Fig. 3-13.A migration profile summarizing Agraulis vanillae migration across a west-
east transect at the latitude of Gainesville, Florida (29.65). The profile is derived from
a median estimate (net number of individuals/minute, adjusted for time of day, sample
size and season) at each site during 13 September-28 October 1986, 26 September-18
October 1987 and 18 September-12 October 1988. Estimates were made along nine
segments of a transect defined by ten sites: STN=Steinhatchee, CRC=Cross City,
TRN = Trenton, NWB = Newberry, GNV = Gainesville, HAW=Hawthorne,
ITL=Interlachen, PLK=Palatka, HST=Hastings, CRB=Crescent Beach.