PROCEEDINGS OF THE UNITED STATES NATIONAL MUSEUM
issued |^b(v vl 0?^i ^y ^^
SMITHSONIAN INSTITUTIONU. S. NATIONAL MUSEUM
Vol. 103 Washington: 1953 No. 3315
DISTRIBUTION, GENERAL BIONOMICS, AND RECOGNITIONCHARACTERS OF TWO COCKROACHES RECENTLY ESTAB-LISHED IN THE UNITED STATESBy Ashley B. Gurney^
IntroductionIn contrast to the large numbers of cockroach species that occurin a great variety of natural habitats in most tropical countries, onlyabout 55 are now recorded from the entire United States. With theexception of those occurring in a few localities, mainly in the South-western States, our roaches are well known to entomologists. In thenorthern third of the country there are few native roaches living infields or woodlands, and most of the better-known species are estab-lished adventives which are rather closely associated with humanhabitations, having been brought there by artificial means. The recentestablishment in the Northeastern States of two potentially importantroaches new to this country is therefore of decided interest.Beginning in 1948 and continuing each subsequent year, specimensof Ectohius livens (Turton), the spotted Mediterranean roach,^ havebeen collected at Falmouth, Mass., under outdoor conditions, comingto houses at night and also occurring on fresh vegetables. Becauseof these habits and the wide natural occurrence of this roach in Europeand adjacent areas, it seems quite probable that a slow spread in theUnited States may take place, and that from time to time questionsregarding its identity and importance will arise.
1 Bureau of Entomology and Plant Quarantine, Agricultural Research Administration,U. S. Department of Agriculture.2 No generally accepted common name seems to be applied to E. livens in Europe, andthis name is considered suitable for American use in view of the species' appearance andthe region of its dominant Old World occurrence. 39228354?53??1
40 PROCEEDINGS OF THE NATIONAL MUSEUM vol. io3A second species not previously known to breed in the United Statesis LeiLCophaea maderae (Fabricius), the Madeira roach. Unlike thespotted Mediterranean roach, L. maderae is well known because it isfrequently transported with articles of commerce in tropical countries.It has many times been intercepted by quarantine inspectors atUnited States ports and has occasionally been carried to inland cities,where it failed to survive. In the autum of 1950 it was found tobe established in the basements of New York City buildings occupiedby people who had come from Puerto Rico. Although its nativehome is evidently West Africa, it apparently was introduced to theWest Indies more than 100 years ago. The Madeira roach couldprobably live outdoors in tropical Florida and perhaps at the othersouthern extremities of the United States, but it is more likely to beimportant as a domestic pest in the well-heated buildings of our largercities. One has only to recall the increasing importance of the brown-banded roach Supella supellectiliuTii (Serville) in this country duringcomparatively recent years to realize the potential importance ofLeucophaea maderae. Apparently a native of Africa that has beenspread by commerce to most of the warmer parts of the world, S.supellectilium was first reported from the United States in 1903, basedon specimens from Key West, Fla. It was not noted as a pest hereuntil found in a private home in Nebraska in 1929, but since thenhas rapidly gained recognition as one of our principal pest roaches(see Back, 1937, and Gould and Deay, 1940)
.
This paper is intended to aid in the identification of these tworoaches and to supply such available information on distribution andhabits as is likely to be helpful to subsequent studies of these speciesin the United States. A third species, Nauphoeta cinerea (Olivier),was found to be established in Florida after this manuscript wasprepared. It is briefly noted on page 46.Genus Ectohius Stephens, 1835Ectobius livens (Turton), 1800? (spotted Mediterranean roach)Figure 10, a-d: Plate 2, Figures 3-6Blatta livida Fabricius, Entomologia systematicae, vol. 2, p. 10, 1793. (Pre-occupied by Blatta livida De Geer, Memoirs pour servir a I'histoire des In-sectes, vol. 3, p. 538, pi. 44, fig. 6, 1773.)Blatta livens Turton. A general system of nature . . . , vol. 2, p. 529, 1800?(New name for Blatta livida Fabricius.)EctoMus lividus Fabricius, Stephens, Illustrations of British entomology, Mandi-bulata, vol. 6, p. 48, 1835.Ectohius livens (Turton) , Ramme, Mitt. Zool. Mus. Berlin, vol. 27, pp. 34-45, pi. 10,fig. 3, 1951.Nomenclatural coTnments.?^Until 1951 Ectobius lividus was usuallyapplied to this species. The name E. livens has been resurrected by
COCKROACHES NEW TO UNITED STATES?GURNEY 41Kamme (1951, p. 34) who says that Dr. K. Princis of Lund, Sweden,informed him that livida Fabricius was preoccupied and that Turtonproposed livens on page 526, vol. 2, of his 1806 edition of Linne's Sys-tema Naturae. Turton published two editions of an English transla-tion of the thirteenth (Gmelin) edition of this work. These editionsare listed in the 1903-1915 Catalogue of the Books, Manuscripts, Mapsand Drawings in the British Museum (Natural History) (vol. 3, 1910,p. 1128). I have verified their dates and contents by an examinationof the entire 7-volume 1806 edition in the Library of Congress and ofvolume 2 of the 4-volume first edition in the library of the U. S. Na-tional Museum. The first Turton edition appeared from 1800 to 1802.Volume 2, containing the genus Blatta^ is dated 1800 and is reasonablysure to have appeared no later than 1802. The second Turton editionappeared in 1806, the first 4 volumes being re-issues of the formeredition with slight changes of the title pages. On the page indicatedby Ramme (p. 526) Forftcula livida^ an earwig, is treated.Prior to Ramme's monograph, the identities of the British species ofEctohius were confused. Thus the species treated by Lucas (1920, p.78) as pers-picillaris Herbst is in reality livens. This has been clarifiedby Blair (1934, 1935). Princis (1936) has explained that a Laplandrecord of lividus (i. e., livens) actually was based on material oflapponicus (Limiaeus). Hebard (1943, p. 12) states that Australianrecords of lividus are incorrect. Taxonomists now differentiate thespecies of Ectobius on rather minute characters, especially on the basisof the dorsal abdominal gland and genitalia of males. Ramme ( 1949
)
has described E. siculus from Sicily, differing from livens in havingblack legs and antennae. E. finoti Chopard of Algeria (see Chopard,1943, p. 21) is much like livens, having the small tegminal spots betweenthe veins and with the dorsal gland bearing a rounded tubercle. Thegenus EctoMus (genotype: lappoQiicns) in a modern sense is muchmore restricted than formerly. In the older literature, species ofEctoMus were reported from many parts of the world. Certain ofthese are now known to belong to other genera. For instance, manyyears ago several Nearctic roaches were referred to Ectobius. Rehn(1931, pp. 306-374) and Hebard (1943) have discussed the genericlimits of Ectobius: Rehn (p. 308) has explained the nature of thetoothed tarsal claws ; the teeth were evidently overlooked by Hebard(p. 12).Recognition features.?The most important features enabling rec-ognition of Ectobius livens are: (1) Small size, over-all length about8 to 9 millimeters ; (2) small, dark spots on the veins of the front wing(tegmen) ; (3) conspicuous intercalated triangle at apex of wing (itr,fig. 10, b) ', (4) claws of unequal length, the posterior claw of eachpair much longer than anterior one; (5) in males the characteristicshape of the dorsal abdominal glandular depression and the sub-
42 PROCEEDINGS OF THE NATIONAL MUSEUM vol. losgenital plate (fig. 10, ?, c) ; (6) conspicuous spines on ventroposteriormargin of each femur.Of the foregoing features, the second and fifth are specific for theseparation of this roach from other species occurring in the UnitedStates and the others are highly confirmatory when in combination.The native Nearctic species most likely to be mistaken ^ for E. livensare Chorisoneura texensis Saussure and Zehntner, of Texas and theSoutheast, and members of the genus Cariblatta. Both have claws ofequal length, and Cariblatta lacks a well-developed, intercalated tri-angle. C. texensis has a large intercalated triangle, but has no con-spicuous spines along the posterior margins of the femora. The best-known outdoor roaches of the Eastern States belong to the genus Par-coblatta; all are larger than livens and lack an intercalated triangle.Identification of the Massachusetts specimens of livens was aidedby comparisons made at the Academy of Natural Sciences of Phila-delphia with authentic French and Swiss material, given to the Acad-emy by the Swiss entomologist Henri de Saussure.Description.?Adults : General form as in plate 2, figure 6 ; tegminaand wings covering abdomen ; width between eyes on vertex about 1^times length of first antennal segment (female), slightly less in male;all femora with conspicuous spines on ventroposterior margins ; ven-troposterior margin of front femur with three or four well-spaced,strong spines along part slightly basad of middle, the more apical oneoften smaller than others, followed by 10 to 15 small, delicate spines(some of which are frequently broken) extending in a regular rownearly to apex, a long curved apical spine closely preceded by a straightsubapical one about three-fourths as long ; tarsal segments one to foureach with a tiny, round pulvillus (membranous pad) ; segment 5 withlarge, broad apical arolium between conspicuously unequal tarsalclaws, anterior claw about two-thirds length of posterior one, longerclaw with about four spinelike inner teeth along basal half, teeth onsmaller claw poorly developed. Tegmen with costal veins simple,cubital veins (discoidal sectors) strongly oblique, apex rather sharplyrounded. Wing with distinctive intercalated triangle (itr, fig. 10,General color pale yellowish; tegmina and marginal area of pro-notum transparent, marked with reddish brown spots, on the tegmenthese spots located on the veins ; eyes dark brown to black ; face yellowto reddish orange ; vertex and upper part of face usually with severalspots of darker orange; coxae and femora unspotted; tibiae withdark brown at bases of some spines; venter of abdomen with poorly
*The standard taxonomie reference on Nearctic Blattarla Is that of Hebard (1917). Arecent key to genera is by John W. H. Rehn (1950). The European species of EctoMuswere reviewed by Ramme (1923).
COCKROACHES NEW TO UNITED STATES?GURNEY 43developed longitudinal rows of dark brown siibmarginal and medianspots.Measurements (in millimeters) of representative specimens: Over-all length, including folded tegmina, female 9, male 8.5; pronotum,female 2.1, male 1.9 ; tegmen, female 7, male 7.2. Width of pronotum,female 3, male 2.9. No significant size variation has been noted.Nymphs : Nymphs (pi. 2, figs. 3, 4) ranging in body length from2.5 to 4.5 millimeters have been examined. They are yellow, withconspicuous dark reddish brown spots on the dorsal surface and on thevertex and upper part of the face. Dark submarginal longitudinalstreaks appear on the thorax. The venter of the abdomen is darkbrown submarginally and medially.Oothecae: Two oothecae have been examined (pi. 2, fig. 5). Theyare dark brown and measure 2.3 millimeters in length. The convexmargin is weakly crenulate and there are faint transverse indicationsof the eggs contained.Distribution.?Three American localities are known for Ectobiuslivens, all in Massachusetts: Falmouth, Manomet, and Plymouth.Falmouth, the first site discovered here, is situated near Woods Holeand the Cape Cod Canal, on the southern side of the base of Cape Cod.In October 1951, after this manuscript had been prepared, I ex-amined 3 males and 4 females of livens collected at Manomet, Plym-outh County, Mass., by Mrs. S. P. Graeff and submitted by Dr. Ells-worth H. Wheeler of the University of Massachusetts. Manomet ison the coast of Cape Cod, about 20 miles north of Falmouth. Thespecimens were collected before August 13, 1951, apparently justprior to that date. Mrs. Graeff wrote, "Our summer cottage is prettywell overrun with them. Painters told me they were even on the roof.The cottage is in the woods and the underbrush is close. I think thisis the third summer we have had them. I suspect they like dampspots, they scurry under shingles."An adult of livens was submitted July 28, 1952, by a Plymouthhousewife who reported that her house and shrubbery were becominginfested.This roach is widely distributed in the southern and central portionsof western Europe, occurring in Germany, the Netherlands, Belgium,France, Switzerland, Italy, Portugal, Spain, and southern England.It has also been reported as far eastward in the Mediterranean areaas Asia Minor, but the records I have seen were published prior tothe modern work dating from Ramme's 1923 monograph of Ectobiusand are not dependable. Chopard (1943, p. 21) has reported it fromAlgeria and Tunisia. I have found no records of its occurrence inthe Azores. Fabricius originally described the species from France,and, in the absence of type material, Eamme in 1923 designated a male
44 PROCEEDINGS OF THE NATIONAL MUSEUIVI vol. io3neotype from Montpellier in southern France. Two varieties oflivens have been recognized, chopardi Adelung described from Francein 1916, and minor Kamme described from the Island of Elba in 1923.These varieties are based primarily on the amount of dark color onthe lower surface of the abdomen, minor being quite dark except nearthe lateral margins, only the submarginal dark spots being well de-veloped in chopardi^ and the submarginal spots being poorly or moder-ately developed in typical livens. The species of Ectohius are wellknown for color variation, the color of tegmina and abdomen oftenbeing recessive (pale) or intensive (dark). In view of this situation,it remains to be proved whether chopardi and minor are valid sub-species or if they are forms without real significance. The center ofdistribution for the genus Ectohius^ as outlined by Rehn (1931, pp.314?318), was apparently the Mediterranean Basin, with a secondarycenter developed in Africa south of the Sahara Desert.Biology.?The original Massachusetts specimens of Ecto'hius livensexamined consist of 2 males, 16 females, 9 nymphs, and 2 oothecae.The males were taken June 21, 1948, and the other specimens in lateAugust and September 1950.^ It may be significant that males werefound only in the spring and early summer. Lucas (1928) reportedthat in England males of E. lapponicus disappeared about midsum-mer, the females persisting longer than males, and nymphs appearingin late summer and fall preparatory to hibernation. No detailedlife history studies of livens in Europe have come to my attention.Brown (1952) has reported observations on the related E. panzeriStephens in England, which occurs there mainly on sand dunes nearthe coast. Adults die in the fall, the males disappearing first, butnymphs overwinter and do not reach full maturity until early Augustof the following year.The Falmouth collections were made on a farm located beside asmall salt river about a mile from the shore. Mr. Flint found speci-mens in a large variety of situations, specimens occurring most con-sistently beneath loose lichens on oak trees, and crawling on trees andhouses at night?the latter in the vicinity of lights. Some werefound under baskets, in buckets, or on Swiss chard and other vege-tables, or were taken by general sweeping (see Flint, 1951).In England E. livens occurs only in the extreme southern counties
?
there found on trees, among bracken ferns, under dead leaves, andflying actively in hot sunshine (Lucas, 1920, p. 80; Burr, 1936, p. 44).Chopard ( 1947, p. 37) says livens is very common in the French woods,and he gives a colored illustration (pi. 2, fig. 28)
.
* AU of this material was collected by Oliver S. Flint, Jr., a student at the Universityof Massachusetts. I am much indebted to Mr. Flint for the enthusiasm, with which hesought additional material once the character of his Initial captures was pointed out tohim, as well as for his diligence in seeking clues to the origin of the introduction.
COCKROACHES NEW TO UNITED STATES?GURNEY 45Source of introduction.?The Massachusetts colony of Ectohiuslivens is probably the result of an accidental introduction from westernEurope or the Mediterranean area. A good many people of Portu-guese ancestry live in Falmouth and the vicinity, and visits to Portu-gal and the Azores are frequent. Seeds and occasional shipmentsof fruit are brought back, but quarantine regulations prevent wide-spread movements of plant products. Since livens occurs in Portugal,as reported by Seabra (1942, p. 19), the possibility of an introduc-tion from that part of the Mediterranean seems the most likely.Probable importance.?Ectohius livens is not likely to develop intoa major pest or to live in buildings except occasionally. As a nuis-ance pest in the vicinity of gardens and dwellings, it may be expectedto attract attention at intervals and to be submitted to entomologistsfor identification and advice, as is now true of male specimens ofParcohlatta which are attracted to lights and which occasionally re-main in houses for short periods.So far as known, species of Ectohius are not economically impor-tant in Europe. E. lapponicus has often been stated to be a pest ofdried fish in Swedish Lappland, but Gaunitz (1935, 1936) foundthat laffonicus does not occur indoors and probably feeds onvegetable matter.Genus Leucophaea Brunner, 1865Leucophaea maderae (Fabricius), 1793 (Madeira roach)Figure 10, e-h; Plate 2, Figxjbes 1, 2Blatta maderae Fabricius, Entomologia systematicae, vol. 2, p. 6, 1793.Leucophaea maderae (Fabricius), Rehn, Trans. Amer. Ent. Soc, vol. 29, p. 283,1908.Nomenclatural comments.?The name Rhyparohia maderae is cur-rently used by certain workers, especially Europeans, as exemplifiedby Chopard (1943, p. 45) in his monograph of North African Orthop-tera. Leucophaea was proposed by Brunner (1865, pp. 272, 278)as a subgenus of Panchlora.^ with four included species. One of thefour, maderae^ was designated type of Leucophaea by Rehn (1903,p. 282). Rhyparohia was proposed by Krauss (1892, p. 165), mono-typic for maderae. Thus Rhyparohia is an isogenotypic synonym ofLeucophaea. The use of Rhyparohia was doubtless encouraged by theaction of Kirby (1904, pp. 150-151) who, evidently unaware of Rehn's1903 designation, designated Blatta surinamensis Linnaeus type ofLeucophaea. As explained by Hebard (1917, p. 309), Kirby also ap-parently objected to the name Pycnoscelus Scudder, 1862, to whichsurinamensis belongs, because it was originally based on an immaturespecimen. Kirby's latter view is not supported by the rules of nomen-
46 PROCEEDINGS OF THE NATIONAL MUSEUM yol. los
clature but, regardless of that aspect of the matter, the validity ofLeucophaea^ with maderae its type, is clear.Recognition features.?The most important features in the recog-nition of Leucophaea maderae are : (1) Large size, overall length about38 to 53 millimeters; (2) pronotal shape, and general color markingsas illustrated (pi. 2, figs. 1, 2) ; (3) no strong spines along posteriormargins of femora except one short apical spine on each hind andmiddle femur; (4) in males the characteristic shape of the specializedorgan on the dorsal surface of the second abdominal segment and ofthe subgenital plate (fig. 10, e, h.).Relatively few Nearctic roaches are as large as L, maderae., and noneof them is likely to be confused with it. Blaherus craniifer Burmeister,illustrated by Hebard (1917, pi. 8, fig. 6), is the largest roach in theUnited States, being much larger than maderae. The species ofPerijylaneta are quite differently marked, and have numerous heavyspines on the posterior margins of the femora. Both Eurycotisfloridana (Walker), of the Southeast, and Hemihldbera tenehricosaRehn and Hebard,^ of the Florida Keys, have short, lobate tegminaand wings which are vestigial or lacking.A relative of L. maderae recently established in Florida isNauphoeta cinerea (Olivier). It superficially resembles maderae.^though it is smaller, its average length being about 23 millimeters.The femora are proportionately stouter and the tegmina scarcely ex-tend beyond the apex of the abdomen, while the posterior margin ofthe pronotum is transversely subtruncate, a median projection beingscarcely evident. Zimmerman (1948, p. 94) and Rehn (1945, p. 274)have given photographs of cinerea. This species is apparently a na-tive of East Africa now widely distributed in the Tropics by com-merce and often intercepted at United States ports. Its potentialimportance is suggested by Illingworth (1942), who described its oc-currence in alarming numbers about the feed room of the poultryplant at the University of Hawaii. He also mentioned its occurrencein heated buildings in England and Germany, though no records ofpermanent establishment in those countries have come to my attention.In early December 1951 material of cinerea was received from Jo-seph Gross of Tampa, Fla., who stated that it was widely and thor-oughly established about Tampa, especially in feed mills. Later, inAugust 1952, Mr. Gross contributed a series of 34 adults and 32 nymphs
s IlemiMaiera tenebricosa was described from the West Indies, including Nassau in theBahamas, by Rehn and Hebard (1927, p. 271). Caudell (1931) recorded an adult paircollected at Key Largo. Fla., by E. A. Popenoe in 1896. Caudell's unpublished notes dis-close subsequent discussions of this record with Mr. Popenoe's son, who explained that hisfather was collecting insects primarily for Kansas State College during the 1896 trip andthat in view of conditions at Key Largo at that time this roach was probably an establishedspecies rather than an intercepted adventive. A number of roaches are now known tobe native to the northern Bahamas and extreme tropical Florida, so the occurrence oftenehricosa in the Florida Keys Is not illogical.
COCKROACHES NEW TO UNITED STATES?GURNEY 47
Figure 10.?Structural details of Ectobius livens (Turton) and Leucophaea maderae (Fab-ricius). a-d, E. livens: a. Dorsal view of apical half of cleared male abdomen; Z?, wing;c, ventral view of subgenital plate and associated cerci, male; d, apical portion of genitalhook, male, e-h, L. maderae: e, Dorsal view of second abdominal segment, male, clearedand in alcohol;/, dorsal view of apical half of cleared male abdomen; g, genital hook ofmale, enlarged; h, ventral view of subgenital plate and associated structures, male, drypreparation. Drawn from specimens as follows: a-d, from Massachusetts specimen; e,from Puerto Rico specimen; / and g, from specimen collected in a Tampa, Fla., warehouse;A, from New York City specimen. (A-9, apodemes of sternum 9; ca, oblique carinae ofglandular specialization; gh, genital hook; gld, glandular depression; itr, intercalatedtriangle; me, mesal extension; st, stylus; sap, supra-anal plate; sgp, subgenital plate.)228354?53 2
48 PROCEEDINGS OF THE NATIONAL MUSEUM vol. io3to the U. S. National Museum. I inquired especially as to the apparentpermanence of the infestation, and on June 24, 1952, a colleague of Mr.Gross', W. B. Gresham, Jr., replied, "It seems evident to us thatNauphoeta is in Florida to stay. They appear to be well adapted andquite numerous in the locations where we have noted them. All in-festations noted have been within commercial buildings engaged in themanufacture of animal feeds. I myself have not noted any out ofdoors." Notes on the Florida infestations by Ratcliffe (1952) andGresham (1952) have appeared.Description.?Adults : General form as in plate 2, figure 1 ; tegminaand wings extending beyond apex of abdomen about one-fifth theirlength ; width between eyes on vertex about two-thirds length of firstantennal segment ; a shallow, scarcely wrinkled transverse depressionin interocular area where eyes are closest ; ventroposterior margin ofmiddle and hind femur each with short, apical spine; conspicuousspines on posterior femoral margins otherwise lacking; ventropos-terior margin of front femur with continuous row of delicate, seta-like spines, these slightly shorter toward apex of femur and a fewnoticeably longer ones near base ; front tibia with two elongate groupsof setae along margins; pulvillus on each tarsal segment, elongateon segments 1 and 5; claws equal, unarmed ventrally, with sparsesetae dorsally ; pronotum obtusely angulate laterally, posterior marginwith rounded median production; tegmen with cubital veins (dis-coidal sectors) oblique; wing (see J. W. H. Eehn, 1951, pi. 9, fig. 102)with well-developed axillary and cubital fields, intercalated trianglelacking. Dorsal surface of abdomen simple in female, specialized inmale; second tergum of male (fig. 10, e) with median glandular de-velopment consisting of paired oblique carinae (ca) and a brief longi-tudinal carina centered on a trifid papilla which largely covers asmall glandular depression (gld) ; supra-anal plate transverse, withmedian emargination of posterior margin, general shape more quadratewhen dry (fig. 10, h) than when preserved in alcohol (fig. 10, /) ; sub-genital plate unspecialized in female, slightly asymmetrical in male,with recurved lateral hooks near styli ; cleared male abdomen withapodemes of sternum 9 (subgenital plate) (A9) united anteriorly,genital hook (gh) and additional sclerotized elongate rod associatedwith an irregular genital mass; genital hook (fig. 10, g) enclosed in amembranous sheath.General color pale brown, the tegmen and pronotum marked withdark brown as illustrated (pi. 2, fig. 1) ; remainder of body and ap-pendages largely dark brown ; clypeus and most of labrum pale ; ab-dominal sterna with oblique submarginal marks of brownish-black;terga with more heavily pigmented areas as shown (fig. 10, /). Baseof genital hook with pigmented transverse lines, the apical portionmuch darker.
U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 103 PLATE 2
1, 2, Leucophaea maderae (Fabricius); 1, Adult female, collected in Manila, PhilippineIslands (length, 47 mm.): 2, last nymphal stage, intercepted at Washington, D.C.3-6, Ectohius livens (Turton), all from Falmouth, Massachusetts: 3, 4, Nymphal stages;5, cotheca; 6, adult female (length, 9 mm.).

COCKROACHES NEW TO UNITED STATES?GURNEY 49Measurements (in millimeters) : Overall body length, includingfolded tegmina, female 45 to 53, male 38 to 45 ; pronotum, female 9 to10.3, male 7.5 to 8.5 ; tegmen, female 37 to 46, male 32 to 38.5. Widthof pronotum, female 13.3 to 16, male 11.5 to 14.Nymphs: Nymphs (pi. 2, fig. 2) ranging from 7.5 to 32 millimetersin body length have been examined. The dorsal surface of the thoraxand abdomen is sharply rugose, with short, microscopic spines, theseespecially conspicuous along the posterior margins of segments;much smaller and less evident spines on ventral surface of abdomen.General color reddish brown ; each tergum with a submarginal darkspot near the base, the basal half of the exposed lateral margin black-ish and the apical third yellowish.Distnhution.?In the United States Leucophaea maderae is knownto be established only in the Harlem section of New York City.*' Manytimes each year inspections made at shipping centers detect the specieswith plant products and other articles coming both from the AmericanTropics and the Old World. The distribution of maderae in warmcountries is very wide, but somewhat irregular, depending upon thevagaries of commerce. The specific and the common names are de-rived from the fact that this roach was first described from Madeira,a small island now belonging to Portugal and located in the AtlanticOcean about 400 miles west of Morocco. Eehn ( 1937, pp. 56-58 ; 1945,p. 273) has reasoned (by a detailed analysis of the present occurrenceof Tnaderae and its congeners) that West Africa is the native home,from which commerce (in early centuries?that of the slave trade inparticular) has carried it to Madeira, the West Indies, Brazil, andelsewhere. Ecuador, Colombia, Venezuela, the Guianas, Brazil, andArgentina are known to have established colonies of maderae^ as wellas Panama, Costa Rica, and most of the West Indian islands. OldWorld distribution outside of Africa is more spotty and, to my knowl-edge, the species is not yet recorded from India or Australia?eventhough it is known from Java, the Philippines, Hawaii, and Fiji.Doubtless, 7naderae now breeds in manj^^ coastal areas from whichthere are no records and it will continue to spread as conditions permit.Biology.?In New York City Leucophaea maderae has been "foundin some abundance in the basement of an apartment of the Harlemsection," and an observer also writes ^ "From their abundance and thepresence of all stages of the insect .... I would conclude that they
?Davis (1940) reported at a meeting held December 15, 1938, that a specimen ofL. maderae taicen in the Bird House of Bronx Park, New York City, was the first recordof tliis species in New York State. In the absence of further information, there is a stronglikeliliood that the specimen in question was an escaped adventive.
' Specimens from the described infestation were submitted by Ralph E. Heal. TechnicalDirector of the National Pest Control Association. Grateful acknowledgment is made ofDr. Heal's cooperation in obtaining information on the Harlem Infestation.
50 PROCEEDINGS OF THE NATIONAL MUSEUM vol. io3
are breeding in this house." It was also noted that the Harlem roachesappeared rather sluggish in their movements.In the absence of biological studies of L. maderae in this country itis instructive to review the observations of life history and habitsthat have been made elsewhere, particularly the papers of Illingworth(1915), Sein (1923), and Pessoa and Correa (1928).As is true of species of PeHplaneta^ growth in the Madeira roachis slow. The first molt may occur 1 to 4 months after birth. Pinnednymphs reared by Illingworth in Hawaii and now deposited in theU. S. National Museum suggest that there are six nymphal stages,though the number is likely to vary, and in cultures it may sometimesbe difficult to observe because exuviae reportedly are eaten immedi-ately after each molt. Pessoa and Correa report only four "meta-morphoses," but that number of molts is almost certainly too smallfor a normal growth cycle. The body length in millimeters of speci-mens of various ages collected by Illingworth is as follows: 1 day,7.5; 1 month, 7.5; 4 months, 16; 5 months, 18.5; 6 months, 24.5; 7months, 29 ; 8 months, 30.7 ; 9 months, 32. Food and environmentalconditions are doubtless important influences on growth. Sein notesmaturity being attained in 220 days, but that a year is often required,while Pessoa and Correa indicate that adults are obtained in a year,with 16 to 18 months the longest observed time. In Sein's culturesthe first nymphs appeared 4 months after their parents matured.Inmiediately following birth, nymphs usually hide beneath themother during the day. According to Pessoa and Correa, this makesthe adult restless and active in contrast to its usual slow gait.In a general paper emphasizing the adaptability and value ofLeucophaea Trmderae as an experimental animal, Scharrer (1951)comments that it thrives on a diet of apples, carrots, and dogfood.She reports that there are 30 to 35 young every 3 months, that thereis an average of 8 molts, and that life expectancy is up to 2V^ years.Dr. Scharrer has observed that tumors which resemble malignantcancer of higher animals develop in various parts of the roach bodyfollowing the removal of the corpora allata and the corpora cardiaca.That such an injury to the nervous system has produced tumors hasvery interesting and practical implications for further research.The name "woodroach" that Dr. Scharrer applied to maderae is aquestionable choice, since the name is most often used for nativeNearctic species of Parcohlatta.Leucophaea is one of several roach genera which are viviparous;that is, the eggs are enclosed in a delicate membranous sac which nor-mally ruptures before extrusion from the mother's body, or immedi-ately thereafter, so that, in effect, the young are born alive. This isin contrast to the habit of most roaches which form a heavily-sclero-tized, dark-colored capsule or ootheca, usually of characteristic shape,
COCKROACHES NEW TO UNITED STATES?GURNEY 51which may be carried protruding from the body for some time priorto hatching, or be deposited loosely or surrounded with a matrix.The viviparity exhibited generally takes a special form known asovoviviparity. This occurs in several well-known genera, especiallyPanchlora^ Pycnoscelus^ Nauphoeta^ and Leucophaea^ and Chopard(1938, p. 218) notes six subfamilies of roaches (according to the long-used classification) in which viviparity occurs. J. W. H. Rehn (1951)'has recently proposed a new classification, based on wings, and it issignificant that many of the viviparous genera fall in what he termsthe epilamproid complex. In addition to Chopard, Shelford (1907),Karny (1924, pp. 3-10), and Rau (1941) have discussed viviparityamong roaches. Hagan (1941; 1951) has described the female re-productive system of the viviparous Diploptera^ and Chopard (1950),that of Gromphadorhina.lUingworth and Sein both recognized that Leucophaea maderae isviviparous, but it appears that Pessoa and Correa observed unusualor abnormal instances of the Qgg sac being deposited, as they wrote ofa capsule being placed in the darkest corner of a rearing box and thefirst young appearing 20 days later. In the following paragraphs Ihave attempted to explain this apparent lack of agreement.Several preserved adult females of L. maderae received for identifi-cation during recent years have exhibited an elongate sac about 20millimeters long protruding from the end of the abdomen. In cer-tain cases the eggs (varying up to about 40 in number), directedtransversely with respect to the mother's body and arranged in tworows, were undeveloped, but in others nymphs ready for hatching werevisible?in fact, rupturing of the sac had sometimes occurred. Rehn(1937, p. 62) has described his experience in the Belgian Congo withthe related L. grandis (Saussure) . A female confined in a bottle gavebirth to 20 living young. The number of young of tnaderae producedat one time, according to published records, ranges from 25 to 32.The related Pycnoscelus surinamensis (Linnaeus) is known to beviviparous, but the literature concerning its habits suggests a lack ofuniformity in birth or hatching. Watson (1929, p. 58) commented onan egg capsule being deposited but the eggs failing to hatch. Caudell(1925) also maintained cultures of P. surinamensis, and mentionedyoung born alive and oothecae seldom if ever protruding from thefemale abdomen. In a somewhat fuller account, Zappe (1918) gavethe opinion that young of surinamensis are either born alive or hatchfrom eggs within 24 hours. Eggs were often laid in soil, but were notobserved to hatch. When females were injured or excited, they oftendeposited poorly developed egg masses that did not hatch. Similarly,lUingworth (1942) reported that the eggs of Nauphoeta cinerea wereusually kept in the body until living young appeared, but that confinedindividuals sometimes produced aborted egg masses that were ex-
52 PROCEEDINGS OF THE NATIONAL MUSEUM vol.103truded but which dried up without hatching when the delicate enclos-ing membrane was exposed to the air. These experiences suggest thatsome of the protruding eggs of L. niaderae that I have noted are theresult of handling when the specimens were collected and that occa-sional Qgg masses may be deposited under certain conditions by thisordinarily viviparous species.In his important recent paper, Chopard (1950) has discussed theanatomy and development of Gromfhadorhina laevigata Saussureand Zehntner, of Madagascar. He observed, in the course of rearingsconducted in Paris, that the female of this viviparous species extrudesthe soft ootheca nearly to its full extent, then it is drawn into anincubating pouch where the eggs undergo incubation for approxi-mately 70 days prior to the appearance of the young. Occasionallythe ootheca is completely extruded, with the result that desiccationoccurs and no hatching takes place. I am indebted to my colleagueR. E. Snodgrassfor pointing out the significance of Chopard's obser-vations. They explain how the ootheca is transferred from the uterusto the incubating chamber in the case of Gromphadorhina. Possiblythe same habit occurs in certain other viviparous genera, and someprotruding oothecae that have been seen may represent a stage in thisnormal act. In discussing the transfer of the developing eggs inthe viviparous Difloptera^ Hagan (1951, p. 299) states that theoocytes pass from the ovarioles to the lower end of the common ovi-duct, where "they are directed by the ovipositor from the genitalchamber ventrally into the open end of the uterus."In Brazil, mating of Leucofhaea maderae occurs mainly during thewarm and rainy season, according to Pessoa and Correa, who say thatcopulation may occupy 20 to 30 minutes and takes place with the pairend to end facing in opposite directions. Prior to mating, the femaleis described as opening her wings and drawing them along the ground,at the same time producing a sound by vibrating them. The scentgland on the dorsum of the male presumably is attractive to the femaleat mating time, but its function is not definitely known.lUingworth stated that when disturbed the Madeira roach stridu-lates very noticeably, and he believed the sound is produced by rubbingthe posterior margin of the pronotum over the mesonotum. Accord-ing to Chopard (1938, p. 286), several roaches stridulate delicately byrubbing the border of the pronotum upon the mesonotum or upon thestrongly denticulate base of the costal vein of the tegmen. To test thepossibilities of stridulation, I have relaxed dry specimens of both sexesof L. maderae^ then manipulated the body parts with my fingers. Alow, squeaking sound is consistently obtained in either sex by rubbingthe lateroposterior margin of the pronotum on the basal costal marginof the tegmen. Both surfaces are heavily sclerotized, turned to oppose
COCKROACHES NEW TO UNITED STATES?GURNEY 53each other, nonpubescent, and are finely rugose so as to make stridula-tion possible.In tropical regions where it is established, Z. maderae is definitely adomiciliary species, though, like certain other roaches often associatedwith man, it is capable of living apart from him in a purely wild state.From available records, however, it is not clear how frequently it hasbeen collected entirely unassociated with man-made surroundings.According to Alfken (1904, p. 565) more than 50 years ago a Germancollector, H. Schauinsland, in 1896 or 1897 found mfiderae in nativehuts on the Hawaiian Island of Molokai, commerce already havingcarried it to certain Pacific areas. However, in spite of this long estab-lishment in the Hawaiian group, no mention of its occurrence in fieldsis made by Williams ( 1931 ) . In July 1950, in the Venezuelan State ofAragua, Dr. Ernst Schwarz ^ collected it coming to lights in fair num-bers, but whether it was breeding in the adjacent forest or associatedwith nearby habitations is not known. Likewise, specimens collectedby the late C. F. Baker in the Philippines before 1927 bear the collect-ing label "Mt. Maquiling," and they may have bred under entirely nat-ural surroundings. However, as I saw during my visit to Mt. Maquil-ing in 1945, at the base and on the lower slopes there are ample oppor-tunities for this roach to remain associated with man. The speciesfrequently flies actively and has often been taken at lights, includingthose on porches, in field camps, or of automobiles. On the other hand,Sein states that Leucophaea maderae does not fly much when indoorsat night as part of an infestation. L. puerilis Rehn, considered itsclosest relative, is a strictly endemic, forest type in West Africa.Source of introduction.?The New York City infestation is reportedto be localized in apartment buildings occupied by people from PuertoRico and, since there has been a good deal of movement?much of itby air?from that island in the past few years, there is little doubt thatPuerto Rico is the source of the infestation.Probable importance.?In Puerto Rico the Madeira roach most oftenoccurs in fruit stores and markets. It is especially fond of grapes. Itis considered very gregarious and develops large, localized colonies.Wolcott (1950, p. 43) records about a bushel having been swept fromone store. Warehouses and other buildings are often infested. Whenhandled or otherwise disturbed, Leucophaea maderae produces an odordescribed as especially offensive. Houses infested by the species usu-ally do not contain Periplaneta or Blattella. There is a strong possi-bility that maderae., if allowed to spread, will gradually develop into aserious pest in our larger cities. In the light of its wide occurrence inthe Bahamas and other West Indian islands, tropical Florida wouldprobably be a suitable habitat, either outdoors or in buildings withoutcentral heating.
? Associated with the Venezuela Plague Mission (Commander J. M. Amberson, D.S.N.,Dr. Ernst Schwarz, and Mrs. Schwarz).
54 PROCEEDESTGS OF THE NATIONAL MUSEUM vol.108ReferencesAlfken, Johann D.1904. Beitrag zur Insectenfauna der Hawaiischen und NeuseelandischenInseln. Zool. Jahrb., System., vol, 19, pp. 561-628, 1 pi.Back, Ebnest A.1937. The increasing Importance of tlie cockroach, Supella supellectiliumServ., as a pest in the United States. Proc. Ent. Soc. Washington,vol. 39, pp. 205-213, 2 pis., 2 figs.Blaib, Kenneth G.1934. A note on the British species of EctoUus Steph. Ent. Monthly Mag.,vol. 70, pp. 157-159.1935. [British species of EctoUus. '[ Proc. South London Ent. Nat. Hist.Soc, 1934-35, p. 5.Brown, E. B.1952. Observations of the life-history of the cockroach EctoMus panzeriStephens ( Orth., Blattidae ) . Ent. Monthly Mag., vol. 88, pp. 209-212.Beunneb de Wattenwyl, ChaklesI860. Nouveau syst^me des Blattaires. 426 pp., 13 pis., Vienna.BuBE, Malcolm1936. British grasshoppers and their allies. 164 pp., 6 pis., 56 figs., 40 maps,London.Caudell, Andrevv^ N.1925. Pycnoscelus surinamensis Linnaeus (Orthoptera) ; on its nymphs andthe damage it does to rose bushes. Proc. Ent. Soc. Washington, vol.27, pp. 154-157, 2 figs.1931. Notes on Blattidae, adventive to the United States. Ent. News, voL42, p. 204.Chopard, Lucien1938. La biologie des Orthopt&res. Encyclopedic entomologique, ser. A, vol.20, pp. 1-541, 4 pis., 453 figs. Paris.1943. Orthoptferoides de I'Afrique du Nord. Faune Empire Fran^ais, vol. 1,pp. 1-450, 658 figs., Paris.1947. Atlas des AptSrygotes et Orthopt^roides de France. Nouvel atlasd'entomologie, vol. 2, pp. 1-111, 12 pis., 12 figs. Paris.1950. Sur I'anatomie et le developpement d'une blatte vivipare. Proc. 8thInt. Cong. Ent., Stockholm, 1948, pp. 218-222, 6 figs.Davis, William T.1940. [Leucophaea recorded.] Bull. Brooklyn Ent. Soc, vol. 35, p. 35.Flint, Oliver S.1951. A new cockroach record from the United States. Bull. Brooklyn Ent.Soc, vol. 46, p. 53.Gaunitz, C. B.1935. Till frSgan om EctoMus lapponictis L. formente skadegSrelse inomhusi Lappland. Ent. Tidskr., vol. 56, pp. 138-150.1936. EctoMus lapponicus L. als Vorratsschadling in Lappland, eine altesicher unrichtige Vermutung in neuer Beleuchtung. Konowia, vol.15, pp. 162-166.Gould, George E., and Deat, H. O.1940. The biology of six species of cockroaches which inhabit buildings.Purdue Agr. Exp. Stat. Bull. 451, pp. 1-31, 13 figs.Gresham, William B.1952. [Note on 2VaM2)/ioe/a in Florida.] Florida Ent, voL 35, p. 77.
COCKKOACHES NEW TO UNITED STATES?GURNEY 55Hagan, Haeold R.1941. The general morphology of the female reproductive system of a vivip-arous roach, Diploptera dytiscoides. Psyche, vol. 48, pp. 1-9, 1 pi.,2 figs.1951. Embryology of the viviparous insects. 472 pp., 160 figs., New York.Hebard, Morgan1917. The Blattidae of North America north of the Mexican boundary.Mem. Amer. Eut. Soc, No. 2, pp. 1-284, 10 pis., 1 fig.1943. Australian Blattidae of the subfamilies Chorisoneurinae and Ecto-biinae. Monogr. Acad. Nat. Sci. Philadelphia, No. 4, pp. 1-129,14 pis.Illingw^obth, James F.1915. Notes on Hawaiian roaches. Proc. Ent. Soc. Hawaii, vol. 3, pp. 136-140.1942. An outbreak of cockroaches, Nauphocta cinerea (Olivier), in Hawaii.Proc. Ent. Soc. Hawaii, vol. 11, pp. 169-170.Kabny, Heinkich H.1924. Beitrage zur malayischen Orthopterenfauna. Treubia, vol. 5, pp. 1-234,1 pi., 85 figs.Kbauss, Hermann1892, Systematische Verzeichnis der Canarischen Dermapteren und Orthop-teren mit Diagnosen der neuen Gattungen und Arten. Zool. Anz,,vol. 15, pp. 163-171.Lucas, William J.1920. A monograph of the British Orthoptera. Ray Soc, Monogr. 104, 264pp., 25 pis., 25 figs., London.1928. Notes on British Orthoptera, including Dermaptera, in 1927. Ento-mologist, vol. 61, pp. 78-81.Pessoa, Samuel B., and Correa, Clovis1928. Notas sobre a biologie de Rhyparohia maderae, Fabr. Rev. Biol. Hyg.,Sao Paulo, vol. 1, pp. 83-87, 1 pi., 3 figs.Princis, K.1936. Erganzungen und Berichtigungen zur Orthopterenfauna Lettlands, I.Folia Zool. Hydrobiol., vol. 9, pp. 90-92.Ramme, Willy1923. Vorarbeiten zu einer Monographie des Blattidengenus EctoMus Steph.Arch. Naturg., Berlin, ser. A, vol. 89, pp. 97-145, 2 pis, 55 figs.1949. Ein neuer EctoMus vor Sizilien (Blatt.) Opuscula Ent., vol. 14, p. 158,1 fig.1951. Zur systematik faunistik und biologie der Orthopteren von Siidost-Europa und Vorderasien. Mitt. Zool. Mus. Berlin, vol. 27, pp. 1^31,39 pis., 134 figs.Ratcliffe, .Tack1952. Lobster roach. Pest Control, vol. 20, pp. 44, 54.Rau, Phil1941. Cockroaches: The forerunners of termites. Ent. News, vol. 52, pp.256-259.Rehn, James A. G.1903. Studies in American Blattidae. Trans. Amer. Ent. Soc, vol. 29, pp.259-290.1931. African and Malagasy Blattidae, Part I. Proc. Acad. Nat. Sci. Phila-delphia, vol. 83, pp. 305-387, 5 pis.1937. African and Malagasy Blattidae, Part III. Proc. Acad. Nat. Sci.Philadelphia, vol. 89, pp. 17-123, 4 pis.
56 PROCEEDENGS OF THE NATIONAL MUSEUM vol. loaRehn, James A. G.?Continued1945. Man's uninvited fellow traveler?the cockroach. Sci. Monthly, voL 61,pp. 265-276, 11 figs.Rehn, James A. G., and Hebakd, Morgan1927. The Orthoptera of the West Indies, No. 1. Blattidae. Bull. Amer.Mus. Nat. Hist., vol. 54 (Art. 1), pp. 1-320, 25 pis.Rehn, John W. H.1950. A key to the genera of North American Blattaria, including estab-lished adventives. Ent. News. vol. 61, pp. 64-67.1951. Classification of the Blattaria as indicated by their wings. Mem.Amer. Ent. Soc. No. 14, pp. 1-134, 13 pis., 5 figs.Schabkee, Berta1951. The woodroach. Sci. American, vol. 185, no. 6, pp. 59-62, 5 figs.Seabba, a. F. de1942. Contribuicoes par o inventario da fauna lusitanica. Insecta Orthop-tera. Mem. Mus. 2k)ol. Univ. Coimbra, Portugal, vol. 127, pp. 1-26.Sein, Francisco1923, Cucarachas. Puerto Rico Estac. Exp. Insul., Circ. 64, pp. 1-12, 9 figs.Shelford, Robert W. O.1907. Viviparity amongst the Blattidae. Trans. Ent. Soc. London, 1906, pp.509-514.Watson, John R.1929. [Rearing of Pycnoscelus.} Univ. Florida Agr. Exp. Stat., Ann. Rep.1929, p. 58.Williams, Francis X.1931. Handbook of the insects and other invertebrates of Hawaiian sugarcane fields. 400 pp., 41 pis., 190 figs., Honolulu.WoLCOTT, George N.1950. The insects of Puerto Rico. Journ. Agric, Univ. Puerto Rico, vol. 32(1948), pp. 1-224, illus.Zappe, Max P.1918. A cockroach pest of greenhouses. Connecticut Exp. Stat. Bull. 203,pp. 802-313.Zimmerman, Elwood C.1948. Insects of Hawaii, vol. 2, pp. 1^75, 228 figs., Honolulu
U. S. GOVERNMENT PRfNTINS OFFICE: 1993