Fao RAs IQUE Bi BS Se HP au oe BM ease Ree: Re at: fre beter ots Re Re MaMa wares ds / Sh Ms a ee yur ae a ee Sig ek ae dew a pct eet, Rake Rete he RC ee DE ae es SGA AE Bie eet i ee : Se RL Pate DD oh ARSED AE secs! a 0°48 Sem Aste ai

eth he Peta 00? es CNS pe Peters ae AEE oy

eB Serge: © Ae ter ar ee she mo mane Baer rap ve BAS

-— PAE ot ee Per Re dts Fs oe Gs ee reno Se nea

Satie sete eR Oe Asura Te ee ae Pare De, See ATE Retin ring Re ee we ete Thad nage eae

. aad aE So TOS mt ore eS hh er air at Te yee te a St Bae spec eB tae 2 age Me pem. om serene ead me Fe ete Me Poe Re tie Fe

re oa See! pee POR ratiraccernier ty > Ace > oe RS ee ee oe ara ©

ee ee

; a - shee SP Sha suse

D2 a Re FEL ORAS | ee rp tee a Par

A lp hig oa ot Ew as Nise

Bip pe SE a eo

rep opaieets eed yee pees

pvp use dersee rn asap Hira ee er ,

‘eS ia fan Sone RE OK a eS

Pe Se oss Be eels ote Pare

vee wees oe Bee Se ae

eee fate s. 2S) % ae mee

oe Sri ota beste ; ant greene i sen eae Eee teary BR oR

Fe ee ahi pyre A eas ihe Se Se wT ja -

S E sey be se DE

s Hah eo Ty Tec

f ei eb ede ee oy yer ge KES Sass age Lap set esis


LAL, 54 |


Volume 50 ae 0) 1996 | Number 1 EN 7

ISSN 0024-0966


of the



26 April 1996



JOHN M. Burns, President ASTRID CALDAS, FREDERICK W. STEHR, Immediate Past Vice President President Pot DANG; ELAINE R. S. HopceEs, Vice President Vice President MICHAEL J. SMITH, Secretary Davip C. IFrner, Treasurer

Members at large:

John V. Calhoun Susan J. Weller Richard L. Brown Robert C. Lederhouse Jon H. Shepard Charles V. Covell, Jr. William E. Miller M. Alma Solis John W. Peacock


Rosert K. Rossins (Chairman), FREDERICK W. STEHR (Member at large) RoBert C. LEDERHOUSE (Member at large) LAWRENCE F. Ga. (Journal) WILLIAM E. MILLER (Memoirs) Marc C, and Maria S. MINNOo (News)


CHARLES L. REMINGTON (1966), E. G. Munroe (1973), ZDRAVKO Lorkovic (1980), IAN F. B. Common (1987), JoHN G. FRANCLEMONT (1988), LINCOLN P. BRowER (1990), DoucLas C. Fercuson (1990), Hon. MiriAM RoruHscHiLp (1991), CLaupE Lemaire (1992)

The object of the Lepidopterists’ Society, which was formed in May 1947 and _ for- mally constituted in December 1950, is “to promote the science of lepidopterology in all its branches, .... to issue a periodical and other publications on Lepidoptera, to facilitate the ex- change of specimens and ideas by both the professional worker and the amateur in the field; to secure cooperation in all measures” directed towards these aims.

Membership in the Society is open to all persons interested in the study of Lepi- doptera. All members receive the Journal and the News of the Lepidopterists’ Society. Prospective members should send to the Assistant Treasurer full dues for the current year, together with their full name, address, and special lepidopterological interests. In alternate years a list of members of the Society is issued, with addresses and special interests.

Active members—annual dues $35.00 Student members—annual dues $15.00 Sustaining members—annual dues $50.00 Life members—single sum $700.00 Institutional subscriptions—annual $50.00

Send remittances, payable to The Lepidopterists’ Society, to: Ron Leuschner, Asst. Treasurer, 1900 John St., Manhattan Beach, CA 90266-2608; and address changes to: Julian P. Donahue, Natural History Museum, 900 Exposition Blvd., Los Angeles, CA 90007-4057. For information about the Society, contact: Michael J. Smith, Secretary, 1608 Presidio Way, Roseville, CA 95661. To order back issues of the Journal, News, and Memoirs, write for availability and prices to Ron Leuschner, Publications Manager, 1900 John St., Manhattan Beach, CA 90266-2608.

Journal of the Lepidopterists’ Society (ISSN 0024-0966) is published quarterly by the Lepi- dopterists’ Society, % Los Angeles County Museum of Natural History, 900 Exposition Blvd., Los Angeles, CA 90007-4057. Second-class postage paid at Lawrence, KS. POSTMASTER: Send ad- dress changes to the Lepidopterists’ Society, % Natural History Museum, 900 Exposition Blvd., Los Angeles, CA 90007-4057.

Cover illustration: The lacewing butterfly, Cethosia biblis (Nymphalidae), occurs in China and southeast Asia. Pen and ink drawing from photograph by Margarette Mead. Original drawing by Martie Clemons, Dudek & Associates Inc., 305 Third Street, Encinitas, California 92024.


Volume 50 1996 Number 1

Journal of the Lepidopterists’ Society 50(1), 1996, 1-20


RICHARD K. WALTON Cape May Bird Observatory, Cape May Point, New Jersey 08212, USA


LINCOLN P. BROWER Department of Zoology, University of Florida, Gainesville, Florida 32611, USA

ABSTRACT. The results of two different sampling methods are presented for sum- mer resident and migrant monarch butterflies, Danaus plexippus L. (Nymphalidae: Dan- ainae), in the northeastern region of North America during the period 1991-1994. Esti- mates of the relative numbers of butterflies obtained by the Xerces/NABA Fourth of July Butterfly Counts and by a newly instituted fall census in Cape May, New Jersey are correlated. Monitoring of the relative annual abundance of monarchs of the eastern pop- ulation and its predicted decline therefore appears feasible. Our Cape May data, as well as historical records, indicate that the monarch’s annual fall migration along the Atlantic Coast is a normal, and not an “aberrant” phenomenon. An alternative migration model, based on avian research, is presented as a possible adaptive explanation for the fall mi- gratory movements of the monarch butterfly along the Atlantic Coast.

Additional key words: censusing, transect methods, Fourth of July Butterfly Counts, migration, Atlantic Coastal Plain, Cape May, New Jersey.

Monarch butterflies, Danaus plexippus L., remigrating from Mexico annually colonize large areas of North America east of the Rocky Moun- tains. Successive broods (Malcolm et al. 1993) move northward as far as southern Canada and the northeastern United States where multiple generations are produced each summer (Cockrell et al. 1993). During late summer and early fall, the year’s final generation(s) of monarchs migrate southwestward (Urquhart 1960), with survivors ultimately reaching the overwintering sites in the Transverse Neovolcanic Belt of mountains in central Mexico (Urquhart 1976, Urquhart & Urquhart 1976, Brower 1977, 1985, Brower et al. 1977).

Although the monarch is a well studied organism, quantitative data


are lacking on its fall movements, including such basic factors as the phenology and the relative sizes of the migrations. Subsequent to the discovery of monarch overwintering sites in Mexico, a model for the autumnal movement of the eastern North American monarch popula- tion i.e., the population that breeds east of the Rocky Mountains, was proposed by Urquhart & Urquhart (1978, 1979) and Urquhart (1987). One aspect of this model dealt with the subset of monarchs found along the east coast—they maintained that migrants heading south along the Atlantic coast in the fall are largely wind drifted and off course. Ur- quhart described such migrants as “aberrant” and suggested that they ultimately end up in Bermuda or, via the Yucatan, in Central America (Urquhart 1987:138-—143).

In reviewing monarch migration, Brower (1995) addressed the spe- cific issues dealing with monarchs migrating to Bermuda, the Bahamas, and the Yucatan Peninsula and concluded that these are essentially “fail- ures of the fall migration to Mexico.” Citing the historical record as well as our studies at Cape May, Brower (1995) hypothesized that most mi- grants east of the Appalachians either move southwestward through the mountains, or follow the coast south to northern Florida, where they turn westward to follow the Gulf Coast on their way to the Mexican overwintering grounds.

Williams (1930) cited approximately 100 observations of migratory movements in North America, of which more than 25% refer to local- ities along the Atlantic Coast from Hampton, New Hampshire to Charleston, South Carolina (Williams 1930:141—149). These accounts extend back to the nineteenth century and indicate that southern New Jersey, and Cape May in particular, have long been known as concen- tration areas for southbound monarchs in the fall. For example, an ac- count from Cape May stated “habitually seen according to Holland” (Williams 1930:149). Hamilton (1885) characterized the September 1885 monarch migration at Brigantine, New Jersey as “almost past belief

. millions is but feebly expressive ... miles of them is no exaggera- tion.” Roger Tory Peterson (pers. comm.) recalled from his visits to Cape May in the early 1930s trees so completely covered with monarchs ~ that they were “more orange than green.” During the last quarter cen- tury, reports by hawk counters from the New Jersey Audubon Society's Cape May Bird Observatory have regularly mentioned substantial mon- arch flights (P. Dunne, pers. comm.). The collective force of these an- ecdotal accounts suggests that a fall migration of monarchs along the east coast is a normal, recurrent phenomenon.

The purpose of this paper is to present and evaluate a quantitative methodology for estimating the relative size of the monarch’s annual fall migrations. We compare two sets of data obtained over four years, from


1991 through 1994. The first was obtained during the summer breeding seasons by the Xerces/NABA (North American Butterfly Association) Fourth of July Butterfly Counts (4JBCs). The second set of data was obtained during the fall migrations in Cape May, New Jersey using a procedure we developed and describe here.

Our results support the hypothesis that a migration along the Atlantic coast is part of the monarch’s normal fall migration. More importantly, the estimates of the relative numbers of butterflies obtained by the two very different methods are significantly correlated. A long term study that combines these two independent estimates should therefore allow us to monitor the relative annual abundance of monarchs of the eastern population.


The Fourth of July Butterfly Counts. The 4JBCs, initiated by the Xerces Society and now administrated by NABA, have been held an- nually since 1975. During the 1991—1994 seasons, the number of counts held throughout North America ranged from 145 to 249 (Opler & Swengel 1992, 1994, Swengel & Opler 1993, 1995). Each 4JBC covers a circular area with a diameter of 15 m (24 km). Participants conduct a one day census of all species of butterflies seen in their count area. The annual census reports include location, date, weather conditions, general habitat descriptions, land use, and notes on unique habitats as well as any significant changes in land use. Observational effort for each census is indicated by the number of observers, the number of field parties, total party hours, and total party miles. Relative abundance data can therefore be calculated as a function of the numbers of butterflies per species seen per party hour.

In her review of the “issues, problems, and opportunities” of the 4|]BCs, Swengel (1990) enumerated a variety of potential pitfalls of us- ing the data to analyze fluctuations in population sizes of various but- terfly species. She concluded that the monarch, however, “is well-qual- ified for such a study because it is widespread, abundant, easily iden- tified, and a habitat generalist” (Swengel 1990:398—399). In her original and subsequent analyses of monarch populations on both continent- wide and regional bases, Swengel (1994, 1996) used the mean numbers of monarchs per party hour from several of the counts to compare trends in various North American monarch populations.

We used the 4]BC database to extract the 1991-1994 monarch data for the summer breeding censuses made at the 68 sites listed in Table 1 and mapped in Fig. 1. Our data are thus a subset of the 4JBC censuses for the northeastern region, encompassing the area of the


TABLE 1. Fourth of July Butterfly Counts listed from north to south by latitude, and secondarily from east to west by longitude. Included are site names, site numbers (Ref), latitude and longitude coordinates (lat.N, long.W), census dates, the number of party hours (ph), the total number of monarchs seen (Dp), and the calculated values for mon- archs per party hour (Dp/ph). The calculated means of all the censuses for each year are at the bottom of the table. Data extracted from Opler & Swengel (1992, 1994) and Swen- gel and Opler (1993, 1995).

Location Ref lat.N long.W “Date, ph. | Dp Dp Onin Compton Co., QU il 45.20 71.45 IW 8.00 10 125 Parc Du Mont Oxford, QU DY Alls) 9X0) TAM 14/7 15.80 34 DUS) Missisiquoi, QU 3 45.08 72.48 ae ss —- North Hero, VT APA AN 5) 73.09 21/6 BAD) 2 0.38 Essex Co., VI aia ale) 71.44 23/6 7.50 if 0.13 Colchester, VT 6 44.32 73.10 Waa 10.00 SYD) 3.50 Camel’s Hump, VT (er AAIG 721k V7 9.00 5 0.56 LaPlatte Marsh, VT 8 44.24 73.14 28/7 4.00 9 Dh ONS Pinkham Notch, NH 9 44.16 TA ALS) eo acct ae athe Lake Placid, NY NON 44s 73.59 aie aot eu Ferrisburg 1B .9 War Ii 44.13 TOMS 14/7 4.00 24 6.00 Ferrisburg W., VT OS TAAaS TOMS ey'7 3.00 39 13.00 Elizabethtown, NY ESAS 73.36 ae a = Hiram, ME VA Aaa» 70.44 = West Rutland, VT ASOD 73.09 aan = ae aut Danby, VT WG 4k IEA 73.00 We 6.50 4 0.62 Saratoga Co. Arpt., NY Wy ALS} OS} Too a a = Antioch, NH SMEAR aii 2G = == Vischer Ferry, NY LOW ADA 73.49 a = = N. Berkshire Co., MA 0) ADS TS.OW sos == C. Franklin Co., MA 21 42.34 Wc 17/7 32.50 184 5.66 Concord, MA OA 26 TQS 16/7 7.00 5D) 7.43 C. Berkshire Co., MA DE A) Oye T3A5 am Na S. Berkshire Co., MA 24 42.09 73.20 ei ee ars Lower Pioneer Vy., MA DS ALD OS T238) YT 5.20 0 0.00 Foxboro, MA WE AOS 71.15 = a Salisbury, CT DY ADO TQ = —= Cumberland, RI W385 41-59 71.29 _— a = Sherman, CT 2) Ocala ill 73.30 e/a 4.50 6 133 Storrs, CT 30) 41GA4S Wola! Wt 6.00 32 Bi} Bristol Co., MA Ol 41.38 70.58 21/7 6.00 5 0.83 Rocky Hill, CT oy Aull ie) T2288) oo Prudence Island, RI 83° ALB Tals Voluntown, CT vb AIL Bis) Valet a= a= Stormville, NY Bi) | ZL BI5) 73.45 = = Monticello, NY 88 2 ZUL Syl 74.38 = a Aquidneck Island, RI 87 9 Algo 71.16 == Meriden, CT Be) GUL CW 72.48 —_— Great Swamp, RI 39 41.26 71.34 E. Fairfield Co., CT 40 41.19 TSP E. Frfld/W. New Hav. Cos., CT 41 41.16 12:55 9/7 4.75 1% 3.58 Fairfield Co., CT AN) Alls MY, Voumke DT 1.00 1 1.00 N. Westchester Co., NY AS ALY) Tora 14/7 44.00 230 5.23 Block Island, RI AVAL ZETA) TASS = =o Westport/Fairfield, CT Ai ALL IN 73.19

Springdale, NJ 46 41.05 74.49 =


Date ph Dp Dp/ph a OO QO 0.00 12/7 9.50 QO 0.00 23/8 5.25 O 0.00 14/6 10.00 QO 0.00 13/7 9.50 ee OL:

‘BYTE Aaa O 0.00 19/7 2.50 One 0200 18/7 3.00 OF 0:00

Wi 3.00 O 0.00 S77 10:25 O 0.00 12/7 6.60 ey OLS 19/7 3.00 O 0.00 MOF 24:00 11. ..0.46 LS - 10s50 O 0.00 11/7 34.00 oe OL09 18/7 9.00 Ibo @3hi QO | .OO.20 67 O17 QO 4.08 Pa OAS IW BSHESD) S) Oyu! 29/7 8.00 Ome 0200 18/7 4.25 QO 0.00 19/7 9.00 Dey? 19/7 5 5) QO 0.00

ot = MAO) ln OlO9

Wm 1350 QO 0.00 20/7 5.00 QO 0.00 18/7 2.00 QO 0.00 ii a4 001 24") 044 16/7 SOUR LOW els

af 3100 2 0:06

TABLE 1. Extended. 1993

Date ph Dp _Dp/ph 13/6 8.25 BD - Ovel 23/6 5.00 O 0.00 18/7 7.00 ee OPI y/ 7 B50) OR Or00 4/7 3.00 QO 0.00 8/7 7.00 OI! 14/6 6.50 O 0.00 17/7 4.50 ee OR22; A/7 9.75 Le Oso 26/09 20100 le Ola NOG eZ Te O0 el OrA4 10/7 L7G IG OL 10/7 LOO as 1.20 WY AOC Ba OBS IeVvie) DSO 9 BS Wall 18/7 OOP sel 222.0 20/7 O50 Jul 1.69 WHT 5.00 DBD OBS 3/7 4.00 ii 10825 Wa 3.00 O 0.00 SIZ 1.50 O 0.00 NOAl = a OOr WiSeemeg Ss ea D0) OY 4LO@) IW 7 S200 Le Oo


NS) ol | e

_ ePwWNMbW -~

|} eal | es

bo [iS

NS) ~]


TABLE 1. Continued.


Location Ref lat.N long.W \ "Date. phy =) pin aIDe ania

West Milford, NJ A7 41.05 (Ae 28/1 = «+ 8.00 9 AS Long Pond, PA 48 41.03 L520 = = suet Greenbrook Sanc., NJ 49 40.54 73.56 20/7 7.00 2) 0.29 Brooklyn/Queens Cos., NY 50 §=©40.52 73.54 6/7 24.00 246 10.25 Muttontown, NY 51 40.51 We 20/7 33.00 106 Ball Great Swamp, NJ 52 40.47 74.28 = = Bronx/Manhattan Co., NY 53 40.41 TDA 29/6 2.7.00 Daal 0.78 Mariton Wildl. Sanc., PA 54 40.4] 75.19 15/6 5.00 2 0.40 Western Suffolk, NY DS 40.40 12345 30/7 32.00 60 1.88 Staten Island, NY 56 40.35 74.09 22/6 25.00 18 0.72 Hawk Mountain, PA BY 40.35 Do —_ = Pool Wildl. Sanc., PA 58 40.33 TO Gl 15/6 1.50 0 0.00 Raritan Canal, NJ 59 40.25 74.34 11/7 28.00 126 4.50 Hendricks, PA 60 40.20 Howls 10/8 3.50 14 4.00 Furnace Hills, PA 61 40.13 76.18 _— = Se Bucks! Con RA 62 40.11 74.54 _— = se Bryn Mawr, PA 63 40.02 75.19 ey 7 15.00 74 4.93 White Clay ClagRx 64 39.42 aA 11/7 5.00 29 5.80 Galloway Township, NJ 65 39.29 74.34 TWH 4.00 9 As) Cumberland Co., NJ 66 39.20 512 15/6 6.00 34 5.67 Belleplain, NJ 67 39.15 74.56 14/7 32.50 105 Oel0 Cape May, NJ 68 39.01 74.52 6/7 “| 31750 lige autem

No. Sites and Mean Dp/h n = 36 3.09

northern Appalachian Mountains and eastward, including southeast- ern Quebec, eastern New York and Pennsylvania east to the Atlantic coast, and thence southward through New England to southern New Jersey. Note that site 68 is Cape May, New Jersey where the fall censuses also were made. Table 1 presents the data for the 1991— 1994 censuses.

The Cape May Census. Sutton et al. (1991) estimated that 10,106 monarchs passed their lookout at East Point, New Jersey during Sep- tember through November 1990, with peak flights occurring on 28 Sep- tember (1,500 monarchs) and 6 October (1,000 monarchs). On 27 Sep- tember 1990, Walton recorded 618 monarchs moving southwest along the dunes at Cape May Point State Park in one 8-minute period. The following day at the same location the flight averaged 536 monarchs per hour between 0832 h and 1541 h (EST). These informal observations confirmed Cape May Point as having excellent potential for annual mon- itoring.

Systematic quantitative data collection was initiated at Cape May Point (hereafter called Cape May) during September-October 1991 by Walton and colleagues, under the aegis of the Monarch Migration As- sociation of North America (c/o 7 Concord Greene No. 8, Concord,


TABLE 1. Extended continued.

1992 1993 1994

Date ph Dp Dp/ph Date ph Dp Dp/ph Date ph Dp Dp/ph 19/7 5.00 O20 2776 10.00 OF 0,008 26/6 16.50 A ORO: 11/7 6.50 Sy OL00 18/7 6.00 QO 0.00 WT 6.00 QO 0.00 16/7 6.50 3 0.46

Sie 2.00 SO .09 ai 16.00 39 2.44 Se 34200). 50) «1.47 Wwe SO.CO- U7. 4a lo? 8700 isis Bs a -—— Din 600 9 0135 17/6 26.00 ie OL04 26/6. 43,00 On On 2 s20/Onmn 2100 QO 0.00 20/6 2.00 Oe OL002 23/6 1.50 DNS VES 1.50 OF0,00 28/6 45.00 QO 0.00 wt 43,00 Jub. O25 WY SBAOO iin OFS PAW TE 5.00 QO 0.00 18/7 S100 Reet Ome 2.002676 8.75 8) = 8 yea OO Ow 0L00 oY I 9.50 Ile = Osa 9/7 ILA XO) LZ OKOS 13/6 1.50 QO 0.00 19/6 2.00 QO 0.00 18/6 2.00 QO 0.00 Sie 225 QO 0.00 Lae PAL > BA ILI LW 18.50 See ORS

12/8 2.50 QO 0.00 26/7 9.25 3 O32 A/7 9.80 Se O82 == = 26/6 4.00 O 0.00 16/7 9.50 8) OS) ean 2.00 QO 0.00 28/6 5.00 Se) teh) 5y/7/ 625 Le Ose 10/7 4.75 QO 0.00 Wai 6.50 Ons 14/7 5.00 Ae ONS 18/7 3.50 Sme29 5/7 Doo 16 =—3.00 Syl ASO OPAL iy Cee LOO WoO leo: 9/7 28.00 AS alenel A ole 0 3S @U0 BOS se) 24 Oo Soo 00 Son le00 28/6 =21.50 AO iO S/Ome2 OL OO eames Ont 25/6 38.00 DIOLS n = 48 0.14 n = 42 0.87 n = 47 1.03

Massachusetts 01742, USA); the censusing was continued in the falls of 1992-1994. Based on methods developed by Pollard (1977), data were collected along a census route through a variety of habitats, including southern hardwood forest, agricultural fields, brackish wetland meadow, suburban neighborhoods, and coastal dunes along the Atlantic Ocean and Delaware Bay. While the 1992-1994 route covered a single 8 km transect, the 1991 census employed three shorter transects, albeit of approximately the same total distance and over much of the same ground covered in 1992-1994 (Fig. 2).

Each transect census was made by a single observer driving a car at approximately 32—40 km per hour, with all monarchs counted along the route. No stops were allowed to count specific concentrations of but- terflies. Monarch totals, starting and elapsed times, and local weather conditions were recorded during each census. These census data are summarized in Table 2.

Statistical analyses. We used Statview II, version 1.03 (Feldman et al. 1987) to run regression analyses that relate the four annual averages of monarchs seen per census hour at Cape May to the average number of monarchs seen per party hour in the four 4JBC censuses. In these regressions, we defined the Cape May data as the dependent variable,


80° 75° 70°

ONTARIO fe Na ee 45°

wa Champiain





ol CAPE MAY 68

100 200 300 scale : km

Fic. 1. Geographic distribution of 68 monarch butterfly census locations of the Fourth of July Butterfly Counts in the northeastern USA, made during the summers of 1991 through 1994. Site 68 is the location of the Cape May road census. See Table 1 for exact locations and yearly census data.

on the rationale that the number of migrants is dependent upon the size of the summer breeding populations.


Table 1 summarizes the 1991-1994 4JBC data for the Northeast in the area from the Appalachian mountain region east through the coastal


Fic. 2. Cape May, New Jersey, showing the census routes over the four years. The 8 km route used in 1992-1994 is indicated by arrows. The crosses at the west end of Sunset Boulevard indicate the end of the truncated transect segment traversed in 1991. The base map is reproduced with the permission of P. Sutton of The Cape May Bird Observatory.


Ferry Road

Intracoastay jy), Ve

A 13 miles to

wih N ae, Lewes, Delaware

—-——— = ae _—

—_—_ ee



Bayshore ROA

Lighthouse Ave.


Ss # CapeMay Migratory “°F By, “4 Point Bird Refuge : rt State Park (TNC)

Atlantic Ocean


CV 1GG SOG = SS S Ic v9l GS 61 i er ILL OL IS) & O0O0'VS SI 0G G des 98ST len cc G LG 16G LIG tsi ce I if SV 00 SV Sil OG S does 6z y9'GG 1S cc e GG GOL 66 VS LNGEG VG SS iS 00°09 OL OL CG. dag gz 00 GI ai SIS ci 69 OL 9) VE OH ss cs HESS) s 00 ST OL G dos 1z aes = a Se 68 SL 192 VS & 00 08 8G ONS) S 00°0 0 OL it das 9z GV 8G LG LS & 99° C6I QE VS © re =e = ings ae re Ree og dog cz te) Gs) Si, VS S O€'S9G LEG Sg S eL$ 8 . &S © 00 VSI OV cI G dos 72 00'0G 9 ST if IL 09T OST 9S o CESS IG VS & O0'GTI G OT I des ¢% rs =: ae ae GV SOG S6I JS OV'G G OS 00 VS9 601 OT T does Z 16GS OL SIS S 9G GG 8 61 it IV'6 8 is c 00 €6E Git OG dog 1Z GE OF TV JES 69 SST Sill 6S (SE All OI VS S 00° G6E S6 cI G dog 0z cy L6V 9SV CSS OL V6G 666 19 IS 67 VV cS & 00°89 Ve OF V dos 61 cL GST OVI gS S IGVL LV SE S si S S| eS OO FTI GI ‘SI I des gs cI TI qT GS GLI8 62 8S ICT V gs Ss a Fea ze = dog LT 09 OS © 6L ES GS 8S 689 fe 19 S 00°09 S| Ss I das 91 OG SS OV OS S cc’sI LI qc CB L fe SS © 00°96 8 ¢ I des ¢] OV EIG Sl Gs o Pegs {S} 6 ©9 OO'LT pall 09 © 00 S9T VI S) ll dos FI I6 VCE LIS S| S We Sill OL 19 S 00'S SG 09 S ae ei Pay a= dog ¢T LL O&G O0OG cS c VV 9G 9G 6S S 18 Ge JS 69 S =e 7 =e ~e des ZI OV VGG a Sie OS 09 ST el OS S $96 OL 69 Fw 4 = i ae dos TT 80° S8T SOM 26S OGL 9 SV 1S) o) IS) S ee Sieg ae ae dog oT se Asi 611i GS © OSS Il LI COG G IS = = == aa dos 6 Or SST 691 SS S CV OL 8 OV 1G9 9 SS a a oe: as das g SG VLL GS © 6S OL 6 Ig ZOO 9 VS = a ae =e dag 1 GG 6SG G66 ss SVL (SG 61 SV S 000 0 tS) = es = a des 9 09°69 8S OS Seve Sill CV LOG G 8S = = Se = dag ¢ CG EG OG IS S Oe T I OF S cl 6 6 6S S a Fu =a a das 7 O8'GG 61 OS SES S SV 2) OGL 9 OS S =: al as a dog ¢ G6 99 8S GS iS ce Il 6 SV 64 TT Il 9S c = = =e ee dog Z SEgs SV GS g9 CT GL OV S We Sit OT 19 © rE was aie = das | wdq dq uIUI Io ywdq dq UIUL Io wdq dq UU a) wdq dq uIUI Io a1eq F661 €661 C661 I66I

‘oTqe} 24} Jo Wi0}0q JY} ye ole ead yovo 10} noy snsued tod syoIeuOU JO LOqUINU OY} 1OF SUBS PozJE[Nyeo oy, ‘(y/dq) Aep 1od Moy, snsuoo 10d usas syoreuoul JO toquinu IBLIOAL OY} 1Of SonyeaA peyepnoyeo oy} pue ‘(dq) peAtoesqo syoieuoUl JO toquinu 9y} *(UTUL) UOT}PAIOSGO JO SOINUTUL [e}O} BUY} *(19) sunt snsusd jo Joquinu 94} ‘soyep snsued oY} o1e pepnyouy ‘vyep snsuso uoneisiul YOIeVUOUL [][eF Aasio[ MON ‘ACI ode F6-I661 24} JO Areutumns ‘SG ATGV LE






ae Sa on 000 6 GE G 000 g GE G 000 9) LI i 000 s LY i 000 OG Ge G LOG 8 CE G 000 L ce G 6G PF I VE G 00°0 9 9 G 00°0 ine aay SF 00°0 g ee G 00°0 VI GE G 000 91 I SE 8 9G ce G cv 9 OG IS 69 $I Il OS e L961 L ee G SECS ca id = Te 6E LG IS WHE SIL SS Zi I VV 9G cV 6P GEG LQ LV VEL 66 OS c LG’ LE 6P VIP LE VS WY SOG VS [ie SV LE G 6S OL LT LI I LL G1 6 tg 9Oy dq UTUI Jo wdq

‘ponunuoy %G AIdVL

G g I G


oy al


) GD AQ & Sy! Sqr! Ga) I A] C9 qi! CONG OD CAN OOO EC SCQNQoTtoooe 2S



w 12)


00ST 00'9 009 OO FG 00 06 00 SL


OO'CTLE 00 OF6 00°99 OS L0€ 00 6EE 00 c6I OO TS OS 8G VV VG 00°08 00ST 00 OL




Sil I 2) Ge) Ne) ae) (<2) te) ae) (co) Taal Ns | Taehiceli a (oeNice). |) (ce)celise) (oe) allan ae




PO $


12) &)



PO ©


12XQ) I



Mean = 142 per hour

1992 2007) Mean = 10 per hour


Mean = 63 per hour

Mean number of monarchs per census hour

1994 500 Mean = 86 per hour 400 300 200 100 0 A 26h 0; is 20 25) a0 40,°. 15%, | 201 es eae SEPTEMBER OCTOBER

Fic. 3. The daily migration pattern of monarchs along the Atlantic Coast at Cape May, New Jersey during the 1991—1994 fall migrations, based on data in Table 2. Note that: (a) the migration extends over a period of approximately 8 weeks during September and October; (b) there is a greater than fourteen-fold difference in size between the minimum (1992) and maximum (1991) migrations; and (c) there appears to be more than a single pulse of migrant butterflies, which may represent migrants of successive gener- ations (see also Fig. 4). In 1991, no counts were made on 1-13, 17 and 25 September, and on 13, 14, 17, 20, and 25-30 October. Zeros on the 1992-1994 graphs include both


plain. A total of 36 sites reported in 1991, 48 in 1992, 42 in 1993, and A7 in 1994 (Table 1). The calculated figures for the mean numbers of monarchs seen per party hour were: 3.09 for 1991, 0.14 for 1992, 0.87 for 1993, and 1.03 for 1994.

Table 2 summarizes the daily census results at Cape May, New Jersey. During the 1991 season, 93 census runs were conducted between 14 September to 24 October and a total of 1,759 monarchs was counted in 11.5 hours of observation. During the 1992 season, 160 census runs were conducted between 1 September to 30 October and a total of 565 monarchs was counted in 47.2 hours of observation. During the 1993 season, 145 census runs were conducted between | September to 30 October and a total of 2,857 monarchs was counted in 40.9 hours of observation. During the 1994 season, 148 census runs were conducted between 1 September to 30 October, and a total of 4,103 monarchs was counted in 42.1 hours of observation. The mean numbers of monarchs per party hour were 142.2 for 1991, 10.4 for 1992, 62.9 for 1993, and 86.3 for 1994.

Figure 3 presents histograms of the daily numbers of monarchs per census hour for each year at Cape May. Because so few monarchs mi- grated in 1992, Figure 4 shows the same 1992 data on an expanded scale. Figure 5 is a plot of the four seasonal averages obtained by the Cape May censuses (y axis) against the 4JBC censuses (x axis). The regression (y = 41x + 23) indicates that the data for the two censuses are significantly correlated (r2=0.905, F=19.0, p=0.049). We also ran an additional regression that included a hypothetical 0:0 pair of values to the four years’ data. It is not unreasonable to do this, because if there were no breeding in the northeastern area, it is likely that there would be virtually no migration through Cape May. This regression (y = 45x + 14) was also significant (r7=0.913, F=31.4, p=0.011).


Correlation of the 4JBCs and the Cape May counts. The Fourth of July Butterfly Counts and the Cape May, New Jersey censuses are very different measures of monarch populations in the northeastern region of North America. Besides the methodological differences, the A4|]BCs are a measure of the relative size of the summer breeding pop- ulation, while the Cape May transects are a measure of the migrating population in the fall i.e., the two record different parts of the monarchs’

rain days when no censuses were made and days when no monarchs were observed (see Table 2).




Mean = 10 per hour

Mean number of monarchs per census hour Ol So

40 30 20 10 0 1 5 10 15 20 5 30 5 10 15 20 25 30 SEPTEMBER OCTOBER

Fic. 4. The same Cape May, New Jersey data for 1992 as shown in Fig. 3, but scaled up to clarify the phenology of the migration.

annual cycle. We had originally thought that a correlation between the two would at best be weak because of the 4] BC methodologies (Swengel 1990). However, the implication of the significant correlation is that the size of the fall migration along the Atlantic coast can be predicted from the magnitude of the summer breeding population in the northeastern United States. Although more years are needed to confirm this rela- tionship, the data suggest that the 4JBCs and the Cape May census procedures are both valid techniques for estimating the relative size of the monarch population from the Appalachians eastwards.

While our major goal for the Cape May census is to compile at least a ten-year data base, the results of the first 4 years include several additional findings of interest. Of particular note is the major decline in 1992 detected by both measures. This decline corresponds with ob- servations of severe mortality in the Transverse Neovolcanic Belt over- wintering sites in Mexico during the preceding winter i.e., in February 1992 (Brower in Culotta 1992). Swengel’s (1993) summary of all 4JBC data for monarchs east of the Rocky Mountains for 1980-1993 indicated that the summer of 1992 was an average year. Compared to 1991, how- ever, the numbers in the midwestern states were down about 65% in the summer of 1992. If only the northeastern 4]BCs sites are considered (i.e., those in Table 1), the decline from 1991 to 1992 was 95% (= {3.09-0.14/3.09}). This drop closely agrees with our Cape May data, which showed a 93% decline (= {142.16—10.41/142.16}).

Data from both counts in future years should thus continue to pro- vide an annual measure of the regional fluctuations in monarch numbers


150 = @ 1991 a > © = © 100 2. i) y = 41X + 23 O R42 = 0.91 Se = TOL

P 0.049


(Mean monarchs per census hour)

Fall migrants

0 1 2 3 4 Summer populations: northeastern region (Mean monarchs per party hour: 4th of July censuses)

Fic. 5. Relationship of the yearly means of the 1991-1994 transect censuses made during the fall migration at Cape May, New Jersey and the mean Fourth of July Butterfly Counts made during the summer breeding season for the same four years in the north- eastern USA. The Fourth of July Butterfly Count data are from the sampling sites in Fig. 1 and Table 1. The Cape May data are the yearly means of the annual fall migrant census counts from Table 2. The two estimates, made during different seasons and using different methods, are correlated (r2=0.91) and the regression of the Cape May data on the 4JBC data is statistically significant (F=19.0, p<0.05). The correlation is evidence that both methods are valid estimates of the relative annual abundance of monarchs in the north- eastern region, and that the 4JBC estimate of the summer breeding population in the northeast is a good predictor of the relative size of the fall migrant population along the Atlantic coast.

east of the Appalachians, as well as furnishing data for comparison with other regions of the country.

The Atlantic coast migration is not aberrant. During the four- year period of our study we have consistently recorded large numbers of monarchs at Cape May, New Jersey, ranging from 565 to 4,103 in- dividuals (Table 2). We have also regularly observed migratory behaviors including: (1) mass movements along beach and dune lines; (2) a high degree of directionality of monarchs observed both during point counts and during counts of the butterflies crossing Delaware Bay (Walton et al., unpubl. data); (3) roost formations; and (4) significant build-ups and exoduses on consecutive days. Another notable characteristic has been


the timing of the fall movement. In each of the four years studied, the numbers of migrants peaked during the third week of September (Figs. 3-4). A second annual peak also is apparent in 1991, 1992, and 1994.

Our Cape May observations argue in favor of describing the Atlantic coast migrants as routine constituents in the monarch’s fall migration. The numbers and behavior of monarchs observed leave little doubt that a significant migration has occurred at Cape May in each year of our censuses. Aspects of the timing of the migration, in particular the re- curring September peaks, also indicate a routine passage of monarchs. Such consistent timing of the peaks would be unlikely if they were caused solely by weather conditions such as cold fronts, because the latter do not occur at the same time each year. Finally, the correspon- dence of the Cape May and 4]BC data sets suggests that the number of monarchs passing through Cape May is representative of northeast- ern breeding populations as a whole. If this correlation holds in future years, it will strengthen the hypothesis that the Cape May migration is representative of the population of northeastern monarchs, rather than comprising an “aberrant” group displaced by atypical weather condi- tions, as hypothesized by the Urquharts.

It is instructive to examine briefly the data and inferences that un- derlie the Urquharts’ position. Maps based on release/recapture data (Urquhart & Urquhart 1978, Urquhart 1987) depict their contention that eastern monarchs normally proceed in a southwest direction in fall, directly (more or less) on course to the Mexican wintering grounds. This would, therefore, take most northeastern monarchs on an inland path away from the coast. But their release/recapture data (Urquhart & Urquhart 1978) also indicate that a substantial number of migrants move east—or southeastward and this, they say, can be accounted for by strong westerly and northwesterly winds. Furthermore, they main- tained that these “aberrant” monarchs probably end up in locations other than the known wintering sites (Urquhart 1987:141). There is substantial evidence (Gibo 1986, Schmidt-Koenig 1993), including data from the Urquharts (1979), that migrating monarchs are affected by wind drift during migration, and so we consider it unreasonable to con- clude that most wind-drifted individuals are necessarily off-course, or that their destination must differ from the main cohort of eastern mon- archs. Point to point release/recapture data establish relatively little, if anything, about either the actual tracks taken by individual monarchs or the mean track of the whole migrant population. It seems just as reasonable to hypothesize that large