The Upper Miocene aeshnids of Monte Castellaro, central Italy, and their relationships to extant species (Anisoptera: Aeshnidae)

32 Pages·1993·6.4 MB·English

by  Gentilini, G

Odonaiologica22(2): 147-178 June I. 1993
The Upper Miocene aeshnids of MonteCastellaro,
central Italy, and their relationships
to extant species
(Anisoptera: Aeshnidae)
G. Gentilin¹i andG.Peters²
1 ViaNazionaleAdriatica78,1-47046 MisanoAdrialico. Italy
2 Zoologisches Museumund Institutfiir SpezielleZoologie,MuseumfiirNaturkunde,
Humboldr-Universitat, Invalidenstrasse 43,D-10115 Berlin, Germany
Received April 14, 1992 /Revised andAcceptedSeptember25, 1992
30wingsand3abdominal segments from theLowerMessinian "Stratodegliinsetti"
nrPesaro are described, analysed and comparedwiththe extanttaxa. Anascryptus
sp.n., Aeshna ghiandoniisp.n., A.messiniana sp.n. andA. multicellulatasp.n. are
described and illustrated. 5out of10moreorlessentire Anaxwingsareindistinguish-
ablefromAnax imperator,butthemajorityoftheAnax andAeshnafragmentscould
notbeassigned toa definite sp. - The monophyleticnature ofthe Anax group is
emphasised,and the genus-group nameHemianax Selys, 1883 suppressed.
INTRODUCTION
The present paper deals with the review of the 6 million years old Upper
Miocene(LowerMessinian)adultdragonflies fromtheMonteCastellarodeposit,
referabletotheAeshnidae,andrecovered during 1974-1991.The dragonflies are
preserved as compression fossils in a well-cementedmarly layer called’’Strato
degli insetti” (GENTILINI, 1986), outcropping in the higherpart ofthefacies
’’Marnebituminose”.Inthisspecial levellarval stagesofdragonflies are absent.
The absence oflarvae, which are very abundantin the underlying strata, might
be dueto the conditions ofthe sedimentary basin, characterized by a variable
salinity, with an ichthyofauna dominatedby Aphanius crassicaudus possessing
traces ofpachyostosis.
Dragonfly remainsoccurwith numerousinsects, leaves,fishesandrarespiders,
molluscs andbirds, representing atemperateto warm-temperatehabitat.Thefirst
record of insect remains at Monte Castellaro was published by GENTILINI
148 G. Gentilini & G. Peters
(1975); descriptions of new arthropod species followed (GENTILINI, 1986;
1989).Thegeology oftheMessinianofthe Pesaroarea was studiedby SAVELLI
& WEZEL(1979) and thefish faunas by SORBINI & T1RAPELLE RANCAN
(1980) and by SORBINI (1988).
Fossil insects have been recovered from two sites of the Monte Castellaro
section, along theAdriatic Seain the vicinityofPesaro town (Marches, Central
Italy). The outcrop is included in the ’’Gessoso-solfifera” formation. Lower
Messinian, andcontainsfourfaciescomposed ofclayey marls, diatomite,bitumi-
nous marls and sandstones. Insect remains occur exclusively in fine-grained
bituminous marls, deposited in a coastal lagoonal environment surroundedby
wood-clad hills.The climatewas probably warm-temperatebut with atendency
to deteriorationtoward a moderately cooling phase, as suggested by insect and
plant remains.
Ofthethirteeninsectordersrepresented in theMessiniansedimentsofMonte
Castellaro, only fourfamiliesof Diptera (Limoniidae, Trichoceridae, Tipulidae
and Bibionidae) and two of Odonata (Libellulidae, Corduliidae) have been so
farstudied (GENTILINI, 1986, 1989, 1990).Numerousfossilinsects stillremain
unidentifiedat thefamily, subfamily and genus level. Only a listoforders, with
numberofspecimens andpercentageshasbeenpreviously recorded.Ontheother
hand the odonates are comparatively well known. Six families are represented
in the deposit: three in each of the suborders Anisoptera and Zygoptera. The
most commonspecies belong to the LibellulidaeandAeshnidaewithmore than
80% ofspecimens, whilethe Corduliidaeand Zygoptera are very infrequent.
Future findings may add more specimens to the materialsubstantiatedin this
review. Male hindwings and basal abdominal segments ofmales are urgently
neededto confirm the numberofaeshnidspecies inthe ’’Strato degli insetti”as
well as to allowtaxonomicalconsiderationsusing theargumentation ofphyloge-
netic systematics.
METHODICALREMARKS
From all specimens presented in this paper, careful drawings have been prepared in order to
calculate dimensional proportionsin wing venation, abdominal segments andappendices,important
for diagnosticalconclusions. Allabsolute measurements were taken directlyfrom the fossils. The
measured distances atthewingsurface have beenillustratedby PETERS (1991,p. 164, fig. I).The
main wingveins are named in accordancewith the evolutionary founded nomenclature ofRIEK &
KUKALOVA-PECK(1984).ItscomparisonwiththeTillyard/Frasersystemandalistofabbreviations,
customary in the terminologyofOdonata wing venation,ispresented in the paper onthe Monte
Castellaro libellulids (GENTILINI, 1989,p. 255).
ANAXLEACH, 1815
Material. — 12 forewings (Nos 154, 155, 156 I, 156r, 157A, I58A, I59A. 160 I, 160 r,
UpperMiocene Aeshnidae ofMonte Caslellaro 149
1209A, 1210, 1218; 9hindwings(Nos 185, 193, 197A, 198, 1211, 1215. 1216, 1217/1217A,1219;
3wing tips(Nos I47/I47A, 182, 195);2 fragmentsof6 abdomen (Nos210, 211/211A).Locality:
Monte Castellaro, Marches, Central Italy.Age: Upper Miocene (Lower Messinian); level ’’Strato
degliinsetti”, facies"Marne bituminose”;formation "Gessoso-solfifera”.
Somewingsarenearlytotallypreserved(Figs 1-3,14),the majorityonlypartly.Forthewingtips
adecision whether they are parts offore- orhindwings was impossible.Several wings and wing
fragmentsduetoplasticdeformationduringfossilisationand/ortectonicalburdenandstressafterwards
(synsedimentaryandsyndiageneticprocesses, C.Brauckmann,in litt.)arestretched orshortened and
inthelatter casemay show anundulatedhindmargin(Fig.6).
The materialofAnax specimensaswell asthe specimensreferredtothegenusAeshna(seebelow)
arestored in the Monte Castellaro workingcollectionofthefirst author.
diagnosis (newsynthesis). — Allabove listedspecimens offossil wings in
total or to a different extent (if fragmented) bear the complex of venational
cchhaarraacctteerrss qquuaalliiffyyiinngg tthhee AAnnaaxx group within the Aeshnidae: (1) anteriorpart of
Arc shorter than its posterior segment, (2) RP diverging near the midst ofthe
anteriorpart ofArc, (3) MA-fork andtheconjunction ofitsanteriorbranch with
RP3-4 very distinct, (4) MsplandRspl deeply arched,(5) betweenIR2 and Rspl
some distinct slanting rows of rarely less than 5 cells, (6) IR2 unforked (its
pseudofork starting beneathPtembracesonly 2rows ofcells),(7)RP, distinctly
curved towards Pt (Fig. 4a),(8) A2inhindwing running parallel toA3. Themale
abdominalappendices of No. 211 (Fig. 12) are constructed as typical for Anax.
On the otherhandthe fossils possess no combinationof characters by which
otherAeshnid groups (genera)are distinguished fromAnax. Themembersofthe
Anaciaeschnagroup(excluding Aeshnaisosceles) havethemajorityofthecharac-
ters mentioned in common with Anax aside oftheir peculiarities in the wing
,
baseandthe secondmaleabdominalsegment,and intheconfiguration ofhead,
they are quite different from Anax by the presence of a true IR2-fork which
embraces 3-4rows of cells (Fig. 4).
Itshouldbe mentionedthat incase offossil material,Anax males, dueto the
female-likeconstruction oftheirhindwing bases andsecondabdominalsegment,
can be detected only by the contour of appendices and (if preserved) by the
presenceofthefossa genitalis.
Five ofthe moreor less complete Anax wings preserved distinctcolourmar-
kings (GENTILINI, 1989),viz. forewing 157A,hindwings 185, 1209Aand 1210
are tinted greyyellowish or pale yellow, and hindwing 198is pale brown.Other
wings, for example forewings Nos 158A, 159Aand 160 as well as hindwing
No. 1219are hyaline inappearance. Colourtraces inthewing membranecannot
beinterpretedas indicativeoffemales.Freshly emerged Anaxspecimens ofboth
sexes bearfumose wings, and there are maturemales with pale yellow or also
greyish yellow tinted wings.
COMPARISON. — At first glance, the fossil Anax wings and their fragments
seemed indistinguishable from wings of the extant A. imperator. But the small
dimensionsoftheonly slightly deformedhindwing No. 1219,thelast discovered
150 G. Genlilini & G.Peters
specimen ofour series ofMonteCastellaro Anax fossils, caused us to pay close
attentiontothelength differencesandtocarefully calculatetheresultsofdeforma-
tion ofthe wings by comparison with the natural configuration of the wing
membranein extant species.
The relative breadth (maximal wing width in % ofits length) has proved to
be a usefulindexforestimating the degreeof deformationin fossil wings. InA.
imperator, ephippiger julius and parthenope, taken as representatives ofthe
,
group, therelative breadth oftheforewing varies between22.3 and24.3%. Out
ofthefive fossil Anaxforewings, only No. 160 (Fig. 5), withan index ofabout
Figs 1-3. ForewingsofAnax cf. imperator: (I)specimen No. 155; - (2)specimen No. 156r; -
(3) specimenno, 157A.
23%, fits the proportions ofrelative breadth in living Anaxspecies. Three other
specimens (Nos 155, 156 and 157A), by index values lowerthan 20%, express
clear signs ofbeing narrowed along the wing axis. The forewing No. 1209A
(Fig. 6), by its cellconfiguration andtheundulationofthehindmargin,revealing
its broadening in anterior-posterior direction, shows a corresponding high index
of 31.4% (Tab. I).
Plasticdeformationduringfossilisationand/orsyndiagenetic processes causing
stretching ofa wing, donot necessarily affectthe wholemembraneevenly. Due
to itsweaker consistence, the apical halfof wing will be less resistant against
stretching (lengthening) than its basal half;the same holds true for the frontal
side ofa wing in contrast with itsposterior part. In orderto test thisconjecture.
UpperMiocene Aeshnidae ofMonle Castcllaro 151
anotherindex was introduced: the ”post/antenodal index”. This ratioexpresses
thedistancebetween N andtheproximal endofPtin % ofthedistancebetween
the first antenodalcrossvein and N (measured along the wing border). In the
forewings of extantAnax species, the ”p/a-index”, being species-specific and
sex-dependent, varies from 62 to 82%. But in general, there is a steep positive
linear relationbetween the absolute N-Pt distance and its relative length (p/a-
-index). Therange ofthe indicatedratiovariationisoccupied by specimens with
N-Pt distances from 11.0-15.0mm in length. The fossilforewing No. 160 is in
line withthese conditions(12.1 mm; 63.4%). No. 157A(Fig. 3), in viewof its
Table I
Diagnosticallyvaluable measurements(inmm)anddistanceproportions(in %) offossil Anaxwings
fromMonteCastellaroin comparisonwithequivalentdataofsomeextant Anax species(meanvalues
orranges ofvariation)
CCoollll..NNoo..ooffffoossssiill lleennggtthhooffRReellaattiivveebbrreeaaddtthhpp//aarraattiioo DDiissttaanncceeNN--PPttiinn PPiilleennggtthhiinn%<* PPiilleennggtthhiinn''X* LLeennggtthhooffTT RReellaattiivvee bbrreeaaddtthh
oorrssppeecciieessnnaammee wwiinngg oolfwwiinngg ''KXoolfwwiinngglleennggtthh ooffddiissttaannccee ooffwwiinngglleennggtthh iinn%‘Xooffwwiinngg ooffTT
NN--PPtl lleennggtthh
FFOORREEWWIINNGG
11-5555 5511..44 1199..66 6611--6655 2244..44 4466..44 1111..33 1144..44 2233..99
I1V5»6rr 6622..00 1188..55 6677 2266..00 3399..88 1100..33 1144..88 1155..66
115377AA 5522..99 1199..88 6677 2266..11 3399..99 1100..44 1155..11 2200..77
100 4477..00 2233..11 66.33..44 2255..77 4400..99 1100..55 1155..11 2266..44
I1220099AA 4422..11 3311.4.4 6633 2255..00 4411..22 1100..33 1155..99 3311..77
iimmppeernaitluorrLLeeaacchh 4488..55--5522..55 2222..33--2233..22 6666--7766 2244..99--2288..55((2266..88)) 3333..66--4444..88((3333.M8)) 99..00--1111..66((1100..55))1133..33--1144.4.4 2255..44--2299.1.1
ppanhaetutrpethenope((SSceli..)) 4444..11--5511..22 2233..22--2244..33 7766--8822 2266..55--2299..33((2288..11)) 2299..00--.3366..66((3322..44)) 88..00--1100..22((99..11))1133..77--1155..11 2233..99--2299..44
jJuulliiuussBBrr, 4499..99--5533..662233..55--2233.9.9 6666--7700 2244..66--2277..22((2255..66)) 3355..77--4444..88((4400..77)) -- - -
e17pthhiippppiiggeerr((BBuurnnnt,.iI 4433..55--4499..55 2222..44--2233..66 6622--6699 2244..00--2277..44((2255..88)) 3377..44--5555.4.4((4455..22)) 99..88--1133..33((1111..55))1155.7.7-1-177.3.3 1199..88--2222..55
ppupauepnsuisensis((BBuurrmnt..)) 4488..44--5533..1| -- 6622--6677 2255..44--2288..11 .3344..99--4444.3.3 99..88--1111..66((1100..88))1155..33 2222..88--2233..55
HHIINNDDWWIINNGG
•185 4499..22 .3344..33 9966..11 3311..11 .3355..88 IILIII 1111..88 4400..44
I•9977AA 3511..33 22.33--2255 110066..77 3344..11 2299..22 1100..00--1100,.11 1111..33--1111..66 3322..44
119988 5511..88--55.33.,772277..00 110022..00 3344..44--3355..77 .33..33..00 1111..44--1111..88 1122..77--1133..11 2266..44
I122L155 4422..44--4422.9.94400..00 9999..11 .333.3.7.7--3344..11((3322..88») 3311..88 1100..77--1100..88 1122..99--1144..00 .3388..66
11221199 4411..44 .3355..77 9988..99 .3322..88 3355..33 1111..44 1122..44 4411..99
iImmppeerraattoorr 4477..11--5500..66 2299..11--33..33.1.1((.3311..00») 9988--111122 3311..22--3344..44((3333..00)) 2288..44--3399..77((3344..00)) 99..55--1122..88((1111..44))1111..44--1111..77 3355..11--3399..44
ppamrtihlwennooppce 4444..11--5500..44 2299..33--33.33..11((3311..33») 110022--111155 3311..99--3355..55((33.33..77)) 2266..22--3322..77((2299..44)) 88..66--1100..66((99..77)) 1111..77--1122..66 3333..33--3388..66
jJuulliiuuss 4499..22--5522..992299..88--3333..00((3311..33») 9944--110055 2299..77--3311..66((3300..66)) 3344..11--4411..99((3377..44)) I1LLII--1122..88 -
ie'pphhiippppiifgteerr 4422..77--4488..773300..33--3322..55((3311.8.8»)110011--111122 3322..77--3355..66((3344..11)) 2299..88--4422..22((3366.0.0))1100..55--1133..99<(1122.0.0))1133..22--1144..44 2266..66--3300..88
ppaappuuernnssiiss 4477..00--5511..882299..00--3300..88((3300..00») 9977--110055 3322..11--3355..44 2288..99--3355..88 1100..22--1111..99((1111..33))1122..66 3355..55--3366..00
greater N-Pt distance (13.8 mm), shouldbear a higher ratiothan itdoes(no less
than 70%), thus indicating heavier stretching of the basal half of the wing
membrane.No. 156(Fig. 2) possesses a very large N-Pt distanceofabout 16.1
mm(never seen inliving species); itsinadequately low p/a ratioofabout66.7%,
based on our N-Pt distance of ’’normal” length (12.6 mm), again points to
lengthening, (probably) ofthebasal halfofthe wing (the relatively long Ptmay
serve asadditionalreference). No. 1209Ashows bothashort N-Ptdistance (10.5
mm) and a low p/a ratio (63.0%); thus the demonstrated broadening of the
membrane is accompanied by some shortening ofitsapical half(Tajb. I).
In summary, the measurablefossil forewings express the following kinds of
deformation:No. 155(Fig. 1) — stretchingofthebasal half;No. 156 — extreme
152 G.Gentilini & G. Peters
stretching along the whole wing axis; No. 157A — lengthening mainly ofthe
basalpartofthemembrane;No. 160 — nearly normalconfiguration; No. 1209A
— broadening and some shortening oftheapical part.
Of the hindwing remnants three specimens allowed making measurements:
Nos 185, 198 (Figs 7, 9) and 1219(Figs 11, 14). Two others (Nos 197Aand
1215) allowed to determinewith certainty only their p/a indices. Inthe above
mentioned four Anax species, the relative breadthofthe hindwing varies from
29.0to33.1%,almostthewholerangeofvariationbeing met withinA. imperator
alone (Tab. I). Themean valuesare 31.0(imperator), 31.8 (ephippiger) and31.3
(julius, parthenope). In femalesthe hindwing usually grows somewhatbroader
than in males, but with extensive overlap of variation. None ofthe calculated
ratios oftherelative breadthofthethree fossil hindwings falls within the limits
ofvariationofextantAnax.No. 185(Fig.7), havinganearlynormalconfiguration,
with an index of34.3% comes nearest to the Anax condition.The index value
of35.7%forhindwingNo. 1219(Fig. 11)indicatesaverybroadwing membrane.
In contrast with the calculated values its basal wing cells are not distinctly
elongated backwards. Equal conditionsseem to prevail in the partly preserved
hindwing No. 1215(Fig. 10) with aneven higher relative breadth (about 40%).
Onthe otherhand,indicationsofcelldeformationincoaxialdirectionare clearly
seen in No. 198,accompanied by disruption ofthe wing tip (Fig. 9); its index
ofrelative breadth is correspondingly low (about27%).
With 96.1%, the value of its p/a index, hindwing No. 185 falls within the
range of variation in extant Anax specimens, with absolute N-Pt length 15-16
mm. Thus, a nearly normal configuration of the wing seems to be confirmed.
Hindwing No. 198,by itsindexvalueof 102%,documentsanextremely stretched
condition, completed by extraordinary length (52.0-53.7 mm) and a very long
N-Pt distance (18.5 mm). The wings Nos 1219and 1215 (the tip ofthe latter
beyond Ptis missing), incontrast with the high valuesofrelativebreadth (35.7
in No. 1219),bearhigher p/a ratiosthan the ’’nearly normal”No. 185(98.9 and
99.1%). Even ifsome broadening ofthese wings cannot be excluded, there are
no indications of their being shortened during or afterfossilisation. And more
than that: if the relatively high p/a index ofboth wings in face oftheir very
broad wing membrane cannot be interpreted as an indicationofstretching, the
index valuesremain tobe seenas a qualifying characterstate, supplemented by
the small length ofNo. 1219(41.4 mm only)!
Keeping in mindthe ascertained data on thepossible ranges of deformation
inthefossilAnax wings, somecomparisons can bemadeusing theirdimensions.
There are only two extant Anax species bearing wings shorter than 48 mm
(forewing) and46mm (hindwing): A.parthenope (only incentralEurope, while
it is larger in almostall other ranges ofits distribution) and A. ephippiger.The
fossilforewings Nos 160and 1209andthehindwingNo. 1219are with certainty
notlonger than 47 mm(47.0, 42.1 and41.4mmresp.). Specimen No. 1215may
UpperMioceneAeshnidaeofMonte Castellano 153
Fig.4. Female fore- andhindwingsof(a)AnaximperatorLeach; — (b)Aeshna isosceles (Muller).
well be addedto thethree fossils mentioned:calculatedfrom itsN-Pt distance,
the wing would have been no longer than 45.3 mm (most probably 42.4-42.9
mm; Tab. I) (the presumed strong lineardependence betweenN-Pt distance and
wing length has beentested previously in thecaseofA.imperator, parthenope,
ephippiger,juliusandjunius).Thereforethereisnoreasontospeculate onpossible
affinitiesofthesespecimens withA. imperator orany otherAnaxspecies, whose
154 G. Gontilini& G. Peters
wings usually measure more than 48 mm. It should be added that specimens
with wings shorter than 48 or 46mm are found in NorthAmerican populations
of A. junius. But this species possesses very long Pt (5.4 and more mm in
forewing, 5.8andmore mminhindwing), not seen inthe short-winged Anaxfos-
sils.
The remaining measurable fossil Anax wings (forewings Nos 155, 156and
157A,hindwings Nos 185, 197Aand 198),with theexception ofthe extremely
stretched forewing No. 156,fall within therangeoflength dimensionsoflarge-
-winged Anax,such asA. imperator,julius or speratus. Bearing inmindthevery
smallrelativebreadthofthe wing No. 156(18.5%) itshouldhavebeen distinctly
shorter ”in life”than the 62 mm measured inthe fossil.
Figs 5-6. Anax cryptus sp.n.. forewings ofthe paralypcs: (5) specimen No. 160; — (6) specimen
No. 1209A.
Afterhaving introduced dimensionsand some wing proportions into analysis
and having compared the fossil Anax wings with those of extant species, as a
preliminary resultofourinvestigation itshouldbestatedthattheMonteCastellaro
set containsat least two differentspecies, asmall-winged and alarger one. For
the former the question of closer affinity to one or the other ofliving species
has already pointed to either A. ephippiger or A. parthenope. Concerning the
long-winged specimens, no indicationcould sofarbepresented andthe question
arises whether these represent only two or perhaps more species.
Intending to solve the affinity problem, the fossilAnax wings and samples of
specimens ofseveralextantAnaxspecies (imperator, parthenope,Julius,ephippi-
UpperMiocene Acshnidae ofMonte Caslcllaro 155
ger, papuensis, speratus, junius, nigrofasciatus, guttatus,immaculifrons) were
analysed andcompared witheachotherby registration and weighting ofquantita-
tive and qualitative characters in fore- and hindwing. Beyond the onesalready
introduced, the following characters have been checked: breadth of discoidal
triangle, T (in % of its length), length ofT (in % of wing length and in % of
length ofsupratriangle, ST), height ofbulge ofR?! below Pt (expressed by the
minimaldistancebetween RP[ and RP2 in % ofthemaximal distancebetween
RP, and IR2), length (posterior extension)ofthe ’’analtriangle” compared with
the length of anal loop, numbersof antenodal(Ans) and postnodal crossveins
(Pns), of crossveins between primary Ans, in the supratriangle and anteriorto
the bridge vein betweenRPandIR2, numbersofcellsalong theRP3.4proximal
anddistal MA-bifurcation, inT, inthe longestoblique row between 1R2andRspl
as wellas betweenMAand Mspl, intheanalloop, paranal loop, ’’analtriangle”,
and the numberof paired cells (including tripled ones) between MP and CuA
(the four last listed countings only in hindwings).
In the small-winged Anax fossils the relative length of Pt has proved of
diagnostic value; there are nospecimens ofA.parthenope possessing aPt longer
than 37% oftheN-Pt distance inthe forewing and33% in the hindwing. In A.
ephippiger thesameratioreaches from 37to 55% andfrom 30to42%, respecti-
vely. Thus there is only a small overlap in the hindwing index values between
thetwo species. Thesmall fossilAnaxforewings No. 160(Pt/N-Pt index40.9%)
and No. 1209A(41.2%) fall within the variation range of relative Pt length
characterizing A. ephippiger. Hindwing No. 1215, by its index valueof31.8%,
is situated in the zoneofparthenope-ephippiger overlap (Tab.I). This transgres-
sion can be minimized by calculating the index Pt length in % of hindwing
length. The variationofthisratio reaches from8.6-10.6%inA. parthenope and
,
from 10.5-12.4%inA.ephippiger (both indicesofrelativeTtlength are not sex-
-dependent). In the hindwing No. 1219the Pt length measures 11.4% of wing
length; thecalculationsonthepartlypreserved shorthindwing No. 1215produced
values between 10.7-11.1%,thus equaling the ephippiger ratio variation.
Anotherkey condition is given by thealready debatedp/a ratio (its species-
-specific variationis not correlated with wing length). In the forewings ofA.
ephippiger theindex varies from62 to 69%; inforewings ofA.parthenope the
rangeextendsfrom76 to82%.The small forewings Nos 160and 1209Apossess
ratiosof63.4and63.0%!As tothehindwing p/aratios, abroadoverlap between
the two species ispresent, butthelower valuesare foundinA. ephippiger (101-
-112%; 102-115%inparthenope). Thetwo smallhindwings, withratiosofabout
99 and 100%,are nearer to ephippiger thanto parthenope.
In short: according to their small dimensionsthe fossil Anax wings Nos 160,
1209A, 1215 and 1219 cannot be compared with other extant species than A.
parthenope and ephippiger (small North American A. julius are excluded by
relatively very long Pt). Using the relatively clear-cut differences in Pt length
156 G. Gentilini & G. Peters
indices and forewing p/a ratios between A. parthenope and ephippiger as a
dividing criterion, we have to state that the individualratiosof the four small
fossil wings fall well within theranges of variationknown fromA. ephippiger
(Tab. I). Thus in these features thesefossilspossess moreaffinities tothe latter
than to any other extant Anax species.
Looking forconfirmationoftheephippiger affinity ofthesmallfossil wings,
only one additional indicationwas detected:there is a statistically proved dif-
ference betweenparthenope andephippiger inthe ratioN-Pt distanceto length
of forewing (meanvaluesinparthenope 28.1+0.17%, inephippiger25.6+0.22%).
The individualratiosofbothsmall forewings (Nos 160and 1209A), namely 25.7
and 25.0%,fall nearthe mean valueofephippiger andfaroutside thevariation
range ofthat index inparthenope (26.5-30.9%).
Figs 7-9. Hindwings:(7)Anaxcf.imperator,specimenNo. 185; — (8)A.cf.parthenope,specimen
No. 197A; - (9)A.cf.imperator,specimen No. 198.
Bearing in mindsome remarkable venational differencesbetween A. parthe-
nopeand ephippiger (relative length ofT, CuA branching, numberofpaired or
tripled cellsbetweenCuA andMP)we haveto notethatthe small-winged fossil
Anax do not meet the diagnostical peculiarities ofA. ephippiger. As will be