A revision of the fire ants, Solenopsis geminata group (Hymenoptera: Formicidae, Myrmicinae).

59 Pages·1991·1.4 MB·English

by  Trager, J. C.

^Z.<^oH-
May29, 1991
Madein UnitedStatesofAmerica
Reprinted from
JOURNALOFTHENEWYORKENTOMOLOGICALSOCIETY
Volume99(2), 1991
A REVISION OF THE FIRE ANTS,
SOLENOPSIS GEMINATA GROUP
(HYMENOPTERA: FORMICIDAE: MYRMICINAE)
James C. Tracer
1
/. New YorkEntomol. Soc. 99(2):141-198, 1991
A REVISION OF THE FIRE ANTS,
SOLENOPSIS GEMINATA GROUP
(HYMENOPTERA: FORMICIDAE: MYRMICINAE)
James C. Tracer
Entomology and Nematology Department, University ofHonda,
Gainesville, Horida 3261
Current address: Shaw Arboretum ofthe Missouri Botanical Garden,
P.O. Box 38, Gray Summit, Missouri 63039
Abstract.—The subgeneraand satellitegeneraofSolenopsisare reviewed. Synonymyunder
Solenopsisofall subgeneraandofthegeneraBisolenopsis, Synsolenopsis, Paranamyrma, and
Labauchena is confirmed. Certain placement ofthe genusLilidris, known from a single alate
female, willrequirestudyofadditionalmaterial,butLilidrisappearstobedistinctfromSole-
nopsis. The fire ants and their close relatives, S. substituta and S. tridens, are collectively
designated as the S. geminata species group, which together with the parasitic S. daguerrei
group(Labauchena)formamonophyleticlineage. 5. virulens, aspeciespheneticallysimilarto
minors ofS. saevissima, is also includedin this revision, though it probablydoes notbelong
inthe5. geminatagroup. Fournativespecies, 2introducedspecies,and2 hybridformsoccur
inNorthAmerica. SeventeenspeciesareknownfromSouthAmerica.Akeytomajorworkers
andillustrationsofallspeciesareincluded.Notesontheidentificationofqueensareprovided
wherethesearesufficientlydistinctive.ThenativeNorthAmericanspeciesare5.aureaWheeler,
S. amblychilaWheelernewstatus, S. geminata(Fabricius), andS. xyloni(MacCook), andthe
introducedspeciesareS. invictaBurenandS. richteriForelfrom SouthAmerica. S. xyloni x
geminataandS. invicta x richteriare the hybridforms. S. invicta x richteriisabundantand
highly fertile in parts ofAlabama, Mississippi and northwestern Georgia in North America,
butthishybridhasnotbeenobservedinSouthAmerica, eventhoughtheparentspecieshave
overlappingdistributionsinatleastSantaFe Province, Argentina. In SouthAmerica, onlyS.
bruesiCreightonnewstatus,S. gayi(Spinola),S. geminataandS. weyrauchinewspeciesoccur
in ortothewestoftheAndes. Thelatterisunusualinthatitoccursatelevationsupto 3,500
m or higherin the PeruvianAndes. To the east ofthe Andes are found S. electro Forel new
status, S. geminata, S. interrupta Santschi, S. invicta, S. macdonaghi Santschi new status, S.
megergatesnew species, S. pusillignisnewspecies, S. pythia Santschi, S. quinquecuspisForel,
S. richteri, S. saevissimaF. Smith,S. substitutaSantschinewstatus,andS. tridensForel.With
the exception ofS. geminata, S. saevissima, and S. invicta (atleast part ofwhosegeographic
range is in the rain forest region) most ofthe South American species are endemic to the
monsoontropics orwarm temperate regions ofthe southern partofthecontinent. S. virulens
(F. Smith) new status, probably not a member of the S. geminata group but superficially
resemblingthem, inhabitstheforests ofAmazonia.
"Fire ant" is the English name used by entomologists and many laymen for a
groupofformicidspeciestraditionallyplacedinthesubgenusSolenopsis{Solenopsis).
In my experience, most English-speaking non-entomologists simply call them "red
ants,"andtheequivalent"fiormigaColorado." isaprevalentnameforthem insome
Spanish-speakingpartsofCentralandSouthAmerica.Othernamesinclude"hormiga
brava" (fierce ant) in Spanish, and ^'formiga defogo" (fire ant), "formiga lava-pe"
142 JOURNALOFTHENEWYORKENTOMOLOGICALSOCIETY Vol. 99(2)
(wash-foot ant, referring to what one might do to remove the stinging insects after
stepping into a mound) and "formiga toicinheira" (lard ant) in Portuguese. Un-
doubtedly, S. geminata has acquired many vernacular names where it has become
established in tropical Africa, Asia and Polynesia.
Certain fire ants, particularly S. invicta, are considered serious pests in the south-
eastern United States. In areas ofhuman habitat modification, fire ants may form
dense populations. Theybuild conspicuous earthen mounds, which are aggressively
defended by the painful stinging ofoften great numbers ofworkers. S. invicta and
its relatives have similarhabits in much ofSouth America, but they are apparently
viewed as no more than minor pests there. This is probably more indicative ofa
generally lowertolerance ofthepresence ofinsectsamongNorthAmericansthan of
a real difference in impact ofthe ants.
The concern about fire ants has created recognition ofthe need for a thorough
revision ofthefireants, whichisthepurpose ofthis paper. SinceCreighton's(1930)
revision ofSolenopsis, there has been no attemptto revise the subgenusSolenopsis,
nor even to define it. More recent revisionary works are those ofWilson (1952),
Snelling (1963) and Buren (1972). In these papers, coverage was limited to only a
few species, and the authors (and Creighton, for most taxa) did not examine type
specimens ofthe Solenopsis taxa described by Forel and Santschi, who described
most ofthe named forms in the group. For this study, I have had the opportunity
to study at least part of the syntype series of most taxa, those of S. pylades, S.
interruptaand ofsome ofthe synonyms ofS. geminatabeingthe exceptions. Types
ofthe former two were examined by W. F. Buren in 1974, and I have made use of
his notes in my assessment ofthesetaxa. Also, I haveon loan from USNM a series
ofS. interrupta with the same collection data as the type series, which I believe to
be workers from the same colony. BarryBolton (BMNH) hasexaminedtypes ofthe
taxasynonymizedwithS. geminatain thispaperdescribedbyFrederickSmith, and
I accept hisjudgement that they do not differ specifically from S. geminata.
METHODS AND TERMINOLOGY
Recognition ofmajorworkers
Thisrevisionisbasedprimarilyonthemajorworkersoffireants,foritisgenerally
in this subcaste that species-specific characteristics are best expressed. It is thus
appropriate here that I discuss the term "majorworker." In an attemptto arrive at
a definition ofthe term, I made plots ofmaximum head width vs. head length of
workers ofall sizes for several fire ant species, pooling conspecific specimens from
variouscoloniesand localities. Theresultingplotsareweaklydiphasic, i.e., showing
a slightlydifferentslopein the upperportion(Wilson, 1971, p. 141). One mightcall
anyworkerfallingwithinthe upperphaseofsuchaplotamajor. Thoseexperienced
with fire ants in the field (Tschinkel, 1988a; Wojcik, unpubl.) note that the major
workersofless populouscoloniesarenotaslargeasthose ofverylargecolonies, but
are clearly recognizable as major workers, especially by their characteristic head
shape. Ontheotherhand, workerpopulationsfrompolygynecoloniesorveryyoung
coloniesdonot,oncasualobservationcontainanyreadilyrecognizablemajorworkers
(but see Tschinkel, 1988a), yet the largest workers ofsuch populations may work
outinthekey. Thus, while"majorworker" cannotbedefinedinabsolutesizeterms.
1991 SOLENOPSISGEMINATA GROUP 143
forpurposes ofthis paper, majorworkers may be loosely defined as the upper 'A to
'A ofthe worker size distribution for mostcolonies.
Characters and possible pitfalls
Characters used for describing and identifying fire ants are ofthe sort commonly
used in ant taxonomy. Some traditional characters, such as pilosity and surface
sculpture patterns, are oflimited use in fire ants because ofthe great homogeneity
in these features across species in the group. The most useful characters are major
workerhead shapeandcolorpattern. Theseandsomeothercharactersarediscussed
below.
Head shape ofminor workers, males, and queens is nearly uniform throughout
the fire ants. However, the head shape ofmajor workers is often diagnostic. The
differencein shapebetweenheads ofcloselyrelated species maybe subtle, soI have
made every attempt to carefully illustrate the typical head shape, and to provide
metric clues to recognizing it. The user ofthis revision should be aware that not
every specimen examined will look "typical," so that some isolated specimens will
not be identifiable by this suite ofcharacters.
It is virtuallytraditional in ant taxonomy to warn readers ofthe dangers ofover-
reliance on color as a means for identifying ants, and color variation in some fire
ant sijecies can be vexing, especially since it may alterthe superficialappearance of
othercharacteristics. Morphologicalanalysisbyacolorneutralmethodsuchasscan-
ning electron microscopy might do much to avoid this problem. Much local and
regional color variation, superimposed on broader clinal patterns, is characteristic
in the species S. saevissima, S. invicta, S. geminataand S. xyloni, all abundantand
widelydistributedspecies. Itmaythen seemcontradictorythatcolorpatternsofthe
remaining, less widely distributed species provide generally reliable characters for
speciesrecognition,andtheyareofsomeutilityevenintheabove-mentionedspecies.
Thus, color is frequently used in the descriptions and key to species. The danger in
thisisnotthatthecolorcharactersarenotuseful,butthatthereissomuchvariation
in interpretation ofcolor terms. I have thus limited myselfto using English color
names. I have found the Munsell system ofnaming soil colors useful as a general
model fornamingantcolors, but have notfollowed itexactly. Allcolordescriptions
arebasedonobservationsofspecimensat25x underbrightincandescentillumination.
Keepinmindthatsmallerworkersandoccasionalmajorswithaberrantcolorpatterns
will be difficultto identify by color.
It is expected that genetic and chemical characterization offire ant species will
help overcome the deficiencies ofa taxonomy based upon strictly morphological
features, though it is safe topredictthatnewandunforeseendifficultieswillariseas
thesetypesofdataaccumulate.Forexample,wehaveatthemomentnoclearconcept
oftheecologicalorevolutionarysignificanceofthevariationinalkaloidcomponents
seen in fire antworkervenomby R. K. vanderMeerandcolleagues (USDAfireant
project, unpubl.). Their data leave considerable room for varying taxonomic inter-
pretations.GeneticdataonfireantspeciesbeingdevelopedbyK. G. Ross(Rossand
Trager, 1991) seem to me inherentlyless susceptible to misinterpretation, butthere
isnodoubtboth setsofdatawiUgreatlyenhanceourgeneralabilitytofurtherrefine
the taxonomy ofthese ants.
144 JOURNALOFTHENEWYORKENTOMOLOGICALSOCIETY Vol. 99(2)
Geographical distribution is another characteristic that may help distinguish fire
ant species, but one must exercise some caution in its use, since fire ant species are
readily transported to new localities where they may become established. The suc-
cessfulestablishmentofS. invicta, S. richteriandS. geminatainlandsfarfromtheir
native ranges is well known, but it appears that some disjunct populations ofthese
and other species within South America may be the result ofintroductions.
Measurements
Measurementswere made at40x or 50x onaNikon SMZ-10 orWild M5 stereo
dissecting microscope, respectively. For polymorphic species, approximate ranges
(includingabsolute maxima) arereportedonlyforworkers withHL > 0.99 mm. For
monomorphicspecies, thefullknownsizerangeisgiven. Whenasingleindividualof
a species was notably larger than all others measured, or when specimens from a
singlelocalityaveragedlargerthanthe othersmeasured, dataforsuchspecimensare
reported in parentheses following the "normal" range. Measurements ofholotypes
ofnew species are listed separately. Abbreviations and definitions for these mea-
surements and indices calculated from them (and other abbreviations used in the
text)aregivenbelow. Othermeasurementsandproportions,bothformajorsandfor
othercastes are defined as necessary in the text. Measurements and indices:
HL—Head length; in full face view (defined below), the distance along the
sagittal axis ofthe head between the anterior midpoint ofthe clypeus
(exclusiveofmedianclypealtooth)andtheposteriormarginofthehead
or, ifposterior margin concave, between the clypeal margin and a line
tangent to the two most posterior points ofthe rear margin.
HW—Head width; in full face view, the maximum width ofhead behind the
eyes.
SL—Scape length; length ofshaft ofantennal scape, exclusive ofbasal artic-
ulation.
EL—Eye length; maximum diameterofcompound eye.
PW—Pronotum width; maximum width ofpronotum in dorsal view.
AL—Thorax (alitrunk) length; distance from anterior base ofpronotum (ex-
clusive ofanterior "cervical flange," which is often hidden from view)
to posterioredge ofmetapleuron.
HW
Cl-CephaUc index; x IQO/HL.
SI-Scape index; SL x 100/HW.
OI-Ocularindex; EL x 100/HL.
Abbreviations forviewingorientations:
ffV—Full face view of the head, whereby which one obtains the greatest
straight-line distance between the midpoints ofthe clypeal border and
the vertex (posteriorborder). Viewingaxis is approximately perpendic-
ular to the surface ofthe frons.
Iv—Profile or strict lateral view.
pdv—Posterodorsalview,usefulforexaminationofsomefeaturesofthethorax,
and ofthe petiole and postpetiole.
1991 SOLENOPSISGEMINATA GROUP 145
Depositories ofspecimens:
AMNH—American Museum ofNatural History, NewYork.
BMNH—British Museum (Natural History), London.
FSCA—Florida State Collection ofArthropods, Gainesville.
IML—Instituto Miguel Lillo, Tucuman, Argentina.
LACM—Los Angeles County Museum ofNatural History.
MCZ—MuseumofComparativeZoology,HarvardUniversity,Cambridge,MA.
MZSP—Museu de Zoologia, Universidade de Sao Paulo, Brazil.
USNM—United States National Museum, Washington, DC.
RESULTS
Inthissection, IfirstdiscussclassificationwithinthegenusSolenopsisandofsome
supposedly separate but closely related groups. I then characterize the fire ants in a
general description ofmorphology ofmajor workers, followed by briefsummaries
from the literature ofother characteristics. A key to workers follows, and this is
followedbyaccountsofthespeciescomplexesandsubcomplexesandtheircontained
species. Minorworkersofmostspeciesarenearlyorindeedindistinguishable.Males
are poorly known and in most cases can reliably be sorted only to species complex,
andarethusnotkeyedordescribed. The queens,whilemorereadilydistinguishable
thanmales,areformanyspeciespoorlyrepresentedincollections,andinidentifying
them,onemustusuallyrelyonassociatedmajorworkers.Inafewcaseswherequeens
arethemostmorphologicallydistinctivecaste,diagnosticallyusefulqueencharacters
are pointed out in the notes accompanying species descriptions.
Synonymic notes on the subgenera and satellite
genera ofSolenopsis
The subgenera ofSolenopsis, and the related genera (sometimes considered sub-
genera) Bisolenopsis, Synsolenopsis, Lilidris, Labauchena, and Paranamyrma were
synonymized into Solenopsis with little explanation by Ettershank (1966). Bolton
(1987) presentedargumentssupportingthesesynonymies,especiallythatofthesub-
genus Diplorhoptrum. Ettershank recognized three "natural" groups in Solenopsis:
the fire ants, the "small species," and the socially parasitic forms. Aside from the
fireants,whichareanatural(thoughparaphyletic)speciesgroup,Ibelievethegroup-
ings in Solenopsiswill sort out very differently from Ettershank's when thegenus is
subjected to phylogenetic analysis.
Forexample,thereseemtobeseverallineagesof"smallspecies."Creighton(1930)
grouped the subgenus Diplorhoptrum (here called the S.fugaxgroup) into 5 species
groups which may better account for the diversity within the "small species." The
phylogenyofCreighton's(1930) 5 groupsisunresolved,andthe 5 speciesgroupsdid
not, furthermore, include all of the diversity represented by Ettershank's "small
species." Additionalgroups ofsmall Solenopsisnot containedinthe S.fugaxgroup
include the subgenera Euophthalma, Granisolenopsis, Diagyne, and Oedaleocerus,
and the so-calledgenera Bisolenopsisand Synsolenopsis.
The distinctions between these taxa are anything but clearly defined, as the fol-
146 JOURNALOFTHENEWYORKENTOMOLOGICALSOCIETY Vol. 99(2)
lowingexampleswillillustrate.(1)ThesingleknownspeciesofGranisolenopsisseems
to be a member ofa South American complex ofsmall Solenopsis spyecies charac-
terized by weakly polymorphic or dimorphic workers. Its major worker has head
morphology convergently developed to resemble a minuscule S. geminata, and its
queen has a narrow, permanently wingless thorax. Conditions approximating these
areseen insome othersmallSolenopsis(allclassifiedasDiplorhoptrum) fromSouth
America. (2) S. globularia (the type species ofEuophthalma) and its varieties and
subspecies are a group ofsmall Solenopsis species which share several, apparently
synapomorphicfeatureswiththeS. tenuiscomplexoftheS.fugaxgroupandshould
be considered related to it. (3) S. virulens is treated along with the fire ants in this
revision, but this is somewhat arbitrary, as the species has features suggesting rela-
tionship to S. globularia (see discussion ofS. virulens for details). (4) On the other
hand, a species described as a Euophthalma, S. huachucana is unmistakably a fire
ant,namelytheyoungqueenandnaniticworkersof5". aurea. (5)Diagynehasqueens
withdistinctivemandibulardentition, buttheworkersarenotdistinguishableatthe
species-group level from the globularia-like Euophthalma. (6) The Euophthalma
speciesoutsideoftheglobularia{-tenuis)groupareapparentlynotdifferentiablefrom
Kusnezov's(1957)generaBisolenopsisandSynsolenopsis. ThefewspecimensIhave
studied within these latter "genera" conform in some respects to my view ofwhat
ancestralSolenopsismayhavelookedlike.Theyaremarkedlysculpturedandsutured,
havepropodealprojections, relativelywell-developedeyes, andflagellarsegmentsof
intermediate length between those ofthe fire ants and ofthe S.fugaxgroup.
I also differ from Ettershank's classification by recognizing 2 distinct origins of
workerless inquilinism in Solenopsis, namely "Labauchena" and "Paranamyrma."
Both show strong signs ofphylogenetic ties to the species groups containing their
hosts, suggesting a recent common ancestry.
Itisclearthatthegenus-leveltaxa(includingsubgenera)inandaroundSolenopsis
are largely meaningless and inseparable. I thus reaffirm Ettershank's synonymy of
the free-livingtaxa Bisolenopsis and Synsolenopsis, and ofthe parasitic genera La-
bauchena and Paranamyrma under Solenopsis. And though there are a number of
recognizable speciesgroups in thegenus, I agree with Ettershank (1966) and Bolton
(1987)innotformallyrecognizinganyoftheabovegeneraandsubgenera,sincenone
appearto be clear-cut monophyletic groups.
Lilidris, represented by a single queen described by Kusnezov (1958) may not,
however,belonginSolenopsis.Itsantenna,though 10-segmented,bearsanapparently
3-segmentedclub. The wingvenation ofLilidrisis alittle differentfrom anyknown
Solenopsis, but venation is variable in Solenopsis, and very likely will encompass
that ofLilidris when studied in more species. The anterior metatarsal "brush" of
Lilidrisis also distinctive.
Based on the above, I referhenceforth to the fire ants andtheirclose relatives, S.
substituta andS. tridens, as the S. geminatagroup. Note that thisgroup isinformal
and paraphyletic (or even polyphyletic, ifS. virulens is included). A strictly mono-
phyleticformaltaxoncontainingthefireantsshouldalsoproperlyincludethespecies
in "Labauchena." The latter are derived, with modifications typical ofinquilinous
ants, from ancestors thatwould be placed within the saevissima complex. Revision
ofthese rarely collected inquilines is not attempted in this paper.
Within the S. geminata group, I refer to smaller groupings ofrelated species as
1991 SOLENOPSISGEMINATA GROUP 147
species complexes, and at a still lower level, subcomplexes. These are briefly char-
acterized at the beginning ofthe descriptions oftheir included species.
General description
A
general morphologicaldescriptionofworkersoftheS. geminatagroup follows,
provided to eliminate repetition of characteristics common to all species in the
treatments ofindividual species, to provide a common basis for comparison, and
forbetterunderstandingofthekeyanddescriptions.Abriefreviewfromtheliterature
ofothercharacteristics offire ants follows the morphological description.
Morphology
Thedescriptionproceedsanteriortoposterior.Theorientationnecessaryforproper
viewingisindicatedin parentheses. Features ofthe minorsare presentedparenthet-
ically, for comparison with the monomorphic workers ofS. virulens and ofthe S.
tridenscomplex.
Head (ffV) usually longerthan broad, usuallywidestbehindeyes; sides straightto
weakly convex in species with quadrate or trapezoidal head shape, more convex in
thosewithelliptical, ovateorcordateheadshape(minorheadshapeelliptical, ovate
or subrectangular, widest at or in front ofeyes); posteriorborder weakly to notably
concave, or less often with angular emargination (faintly concave to convex in mi-
nors); the concavity 1-2x as wide as the distance between apices offrontal lobes;
loweredgeofdistalborderofclypeusbearingalargemedian seta, thisusuallyborne
on a projecting triangular tooth, the latter reduced or lost in some species; clypeal
carinaedivergentdistad, usuallyprojectingasisosceles-triangularteeth, thesesome-
what to notably larger than median tooth and always much larger than paracarinal
teeth, which may be lacking (especiallyin smallerworkers), carinal and paracarinal
teeth more dorsad on clypeal borderthan median tooth; mandibles 4- or less often
3-toothed (teeth may be worn off in older specimens, but are always present at
eclosion);mandiblecurved, distalportionofouterborderusuallyataweaklyobtuse
angle to basal portion (about 100°, angle larger in minors); mandibular costulae 6-
10, complete throughoutlength ofmandible orobsolete medially, sometimesbifur-
cate distally; eye (Iv) ovate, elliptical or reniform, with fi-om 45-100 facets (20-60
inminors); scapes(ffv)curvedbasally,thickestsubapically; scapelengthsignificantly
less to a little less than length between basal articulation ofscape and most distant
portion ofposteriorborder, i.e., scape apex not surpassingposteriorborderofhead
(often as long or notably longer than this distance in minors and monomorphic
species);anterodorsalpronotalborder(pdv),weaklytonotablyconvex;anterolateral
pronotal comers variously developed, broadly rounded to distinctly angular and
bearing obliquely or transversely oriented, "humeral bosses"; promesonotal suture
chevron-shaped with a small dorsal projection at apex, or parabolic, or strongly
convex, rarely obsolete, and this only in smaller majors (commonly so in minors);
pronotum (Iv) usually with steep anterior declivity set offfrom dorsum, pronotal
dorsum forming an even convexity with mesonotal dorsum, or with a slight break
in outline at point ofanteriormesonotal projection; metanotal impression conspic-
uous, set offby steep, variously sculptured, posterior mesonotal and anterior pro-
podeal declivities, the former declivity often higher than the latter (metanotal im-
148 JOURNALOFTHENEWYORKENTOMOLOGICALSOCIETY Vol. 99(2)
pressionshallowerandlesssculpturedinminors);metanotalspiraclessmall,positioned
dorsolaterally; propodeum (pdv) with dorsal face concave, descending through an
even curve into declivous face; in profile (Iv) propodeum usually appears angular
becauseoflongitudinal,posterolateralbossesorshortroundedcarinae(bosseslacking
orweaklydevelopedinminorssopropodealprofilelessornotatallangular);petiolar
peduncle shorter than to slightly longer than base ofnode; profile ofpetiolar node
cuneate orthick-squamose to globular, with anteriorface straightto weaklyconvex
andposteriorfaceconvex,thefacesmeetingthroughstronglyconvexdorsalportion;
from behind outline ofpetiolar node (pdv) globular, subovate, orwith more or less
convexdorsalfacemeetingstraightorweaklyconcavesidesthroughroundedangles,
sides convergent ventrad; profile (Iv) ofpostpetiolar node in profile typically lower
thanthatofpetiole, appearingglobularornearlyso, withashortposteriorpeduncle;
from behind outline ofpostpetiolar node (pdv) globular to subtrapezoidal or sub-
rectangularwithdorsumconvex(alwaysmoreorlessglobularinminors);postpetiole
varying from slightly to notably widerthan petiole.
Integumentmostlysmooth;exceptforpiligerousfoveolae,andsculpture ofmeso-
metapleuron, propodeum, petiolar peduncle and rear face ofpostpetiole (sculpture
always lessdeveloped in minors); diameterand sometimesshapeofpiligerousfove-
olaevaryingindiagnosticallyusefulways(small,roundandinconspicuousinminors);
sculptureofmesometapleuronconsistingoflongitudinalstriaeorrugaewithvarying
levels ofinterstitial punctation; sculpture ofpropodeum and ofpostpetiole variable
and often diagnostically useful at species level; declivous face ofpropodeum with
weak transverse rugae or, more often, unsculptured on upper portion, on lower
portion with concentric semicircular rugae continuous with those ofmetapleuron,
but more widely spaced and usually lackingintersticial sculpture; petiolarpeduncle
usuallyfaintlyareolateorpunctate,thissculpturecontinuedposteradtobaseofnode
insomespecies; venterofpetiolewithlongitudinalmediancarinaandanteroventral
process consisting ofone to a few small teeth or a transparent flange, occasionally
absent; dorsum ofpetiolar and postpetiolar nodes often weakly scalloped or longi-
tudinallygrooved;petiolardorsumotherwiseunsculptured;postpetiolardorsumun-
sculpturedorwithsculptureresemblingbutweakerthanthatofpostpetiolarposterior
face; posterior face ofpostpetiole with varying, diagnostically useful combinations
and distributions oftransverse striae and punctation, especially on lower portion;
sides ofpostpetiole usually striate-punctate; venter ofpostpetiole usually coarsely
punctate with a few coarse longitudinal rugae.
Pilosity composed ofyellowish or reddish brown setae, these normally more or
lesscylindricalandtaperingdistally,ormoreprecisely,narrowlyconical;longersetae
curved;mesonotumusuallywithatleast20erectsetae(lessinminors);mesopleuron
with few aside from those on ventral edge; in most species pilosityvariesgreatlyin
length on a single specimen, longest hairs on thoracic dorsum usually at least 2.5x
length ofshortest; suberect pubescence present in a conspicuous patch on cervical
flangeofprothorax;lessoften,dilute,appressedpubescenceoftenpresentonanterior
face ofpetiolarnode, and rarely some on propodeal dorsum.
Color ranging from nearly uniform honey-yellow to brownish black, in lighter
shaded forms with at least posterior band oftergites usually notably darker; some
specieswithmoreorlessuniformcolorpatterninallsamples;othersspanningnearly
the entire range for the species group, though typically not within a single colony;
1991 SOLENOPSISGEMINATAGROUP 149
(minorworkersoftendarkerandmoreuniformlycoloredthanmajorsfromthesame
colony).
Sting morphology
Kugler(1978) published anextensive reviewofthemyrmicine stingapparatus, to
which the reader should refer for details. Kugler's analysis resulted in Solenopsis
genus group containing Megalomyrmex, Monomorium (including Chelaner), and
Oxyepoecus, known relatives oiSolenopsis (Ettershank, 1966; Bolton, 1987). Also
includedin thisgroupby KuglerwasapairofRogeria species. Theirrelationshipto
Solenopsisis contradicted by other lines ofevidence.
Malpighian tubules
Brown (1988) surveyed Malpighian tubule numbers ofants. Among the above
mentioned relatives ofSolenopsis, 2 Megalomyrmexspp. (perhapsthe most "prim-
itive" genus in thegroup) had 5 Malpighiantubules, while all the remainingspecies
(including1 or2spp.fromeachoftheother3generahad4tubules(asynapomorphy?).
The tubules are not cryptonephric.
Larvalmorphology
Wheelerand Wheeler(1960a) dividedwhat were then considered to be members
ofthetribeSolenopsidiniintosixgenusgroups.ThegeneraSolenopsis,Monomorium,
Oxyepoecus, and Megalomyrmex, (and Anergates, now thought to be a member of
theTetramoriini)aregroupedinthesolenopsidiformgenera; thosewithshort, stout,
superficiallystraightbodyform(butwithanteriorventralportionofthoraxcurved),
ends rounded, neck very short or lacking, and anus ventral. Later the Wheelers
(1960b) described the larvae ofS. picta, S. pergandei, and S. globularia littoralis
"
(which Iconsidertobelongto 3 distinct speciesgroups)as "similartoS. geminata,
differingin details ofsize and pilosity (confirmingthe close relationship ofall Sole-
nopsis).
Karyotypes
Taber and Cokendolpher (1988) reported the karyotypes ofS. xyloni specimens
from Texas and Arizona, synthesized their results with those from previous work,
and listed all pertinent references. Karyotypes ofthe 2 S. xyloni populations were
identical, and theirchromosome morphologyclosely resembledthat ofS. aurea, S.
invicta and S. saevissima, but differed from that ofS. geminata and S. richteri. The
diploidcomplementis32inallspecies,butasindicatedbytheirresults,chromosome
morphologyvariesinwaysthatappeartobeunrelatedtotaxonomicgroupingswithin
the S. geminatagroup.
Venom andcuticularhydrocarbon chemistry
Blum et al. (1985 and included references) have studied a variety ofSolenopsis
spp. andsomeinrelatedgenera. ItappearsthatSolenopsisspeciesexhibitachemical
synapomorphyofthepresenceof2-alkyl-6-methylpiperidinealkaloidsinthevenom.