1984 - francke & sissom

8/7/2019 1984 - Francke & Sissom

1/20

Francke, O. F. and W. D. Sissom. 1984. Comparativereview of the methodsused to determine thenumberof molts to maturity in scorpions (Arachnida), with analysis of the post-b~th developmentof Vae/ovis coahuilaeWilliams(Vaejovidae). J. Arachnol., 12:1-20.

COMPARATIVEREVIEW OF THE METHODSUSED TODETERMINE THE NUMBEROF MOLTS TO MATURITY

IN SCORPIONS(ARACHNIDA), WITH ANALYSISTHE POST-BIRTH DEVELOPMENTOF VAEYOVIS

COAHUILAEWILLIAMS (VAEJOVIDAE)

Oscar F. Franckeand W. David Sissom1Department of Biological Sciences

Texas Tech UniversityLubbock, Texas 79409

ABSTRACTLife history studies on scorpions have taken various approaches. Thetheoretical methoduses aprogression factor of 1.26 in linear dimensionsat each molt to predict the numberof molts requiredby youngscorpions of knowninstar to reach adult size. The direct empirical approach consists ofraising scorpions to maturity in captivity. The indirect approachis based on morphometdcanalyses offield caught samples, and assumesthat discrete size classes can be recognized and interpreted asrepresenting the various instars. The mixedapproachuses extrapolation to predict maturity from theresults of a partial life history. Thereliability of the various approachesis evaluated(a) by analyzingthe life history of Vae]oviscoahuilaeWilliams,and (b) by reviewingthe results of all prior fife history

studies on scorpions.INTRODUCTION

Studies on scorpion life histories, in particular those that determinethe numberofmolts required to attain sexual maturity, follow one of two fundamentalmethods:theoretical or empirical. Furthermore,within the empirical methodthere are two ap-proaches:indirect and direct. Before proceeding any further, however, we are compelledto state explicitly our usage of various terms to avoid possible misunderstandings. Aninstar is the period or stage between molts, numberedto designate the various periods;e.g., the first instar is the stage betweenthe egg and the first molt (in the scorpion litera-ture often referred to as a larva or pullus, followed by the first nymphal imtar-which isactually the second instar). The stadium is the interval between molts, measured in somechronometric unit of time. An age class is a group of individuals born at the same time(day, month, season, whichever temporal parameter is chosen). Individuals belongingthe same age class can conceivably differ in size because of differential growth rates, orbelong to different life stages because of different developmental and molting rates. A

I Present address: Departmentof General Biology, Vanderbflt University, Station B, Box1812,Nashville, Tennessee37325.


2/20

2 THEJOURNALOFARACHNOLOGYsize class is a collection of individuals of the samesize, regardless of age or life stage;e.g., someindividual spiders mayactually decrease in size during a given molt, thus theirsize class would differ from other membersof their age class, and they would be in adifferent instar than the remainder of the population belonging to the same size class.

The theoretical method is based on the knowledge that certain measurements insuccessive stages of at least somearthropods increase in a regular geometrical progression(Dyar 1890, Przibram and Megushr 1912). A progression law was proposed by Przibramand Megusarbased on studies on the Egyptian Preying Mantis, SphodromantisbioculataBurmeister. They found that mass increased by a factor of two during intermolts andlength increased by a factor equal to the cube root of two (= 1.26) during molts. Thus,knowingthe weight or the length of some structures on newbornand adult arthropods,and assuming the progression law to be valid for that species, the numberof molts (orintervening stages in the geometrical progression) linking newbornand adult measure-ments can be calculated.The indirect approachis based on establishing size classes within population samples,and equating the resulting size classes with instals. With relatively small samples thespecimensare arrangedin a linear series of increasing (or decreasing)size, whichis visuallyinspected for gaps and clusters which define the various size classes. The gaps andclusters are respectively equated with molts and instars (e.g., Auber1959, Vachon1940,1948, 1951, 1952). With larger samples usually one or more structures on each specimenare measured, and the frequency distribution is plotted either in a univariate histogram(Fox 1975), or in a bivariate morphometfic plot (Smith 1966, Shorthouse 1971, Polisand Farley 1979). The plot is inspected to determine the clusters which presumablyrepresent the various instars.

The direct approach uses pregnant females either from the field or from matings in thelaboratory. Followingparturition in captivity the youngare raised to maturity. This is themost commonapproach (Table 6).The direct methodhas several shortcomings, including a considerable investment intime and energy by the investigators. Often only a partial life history is obtained becausescorpions frequently die before reaching sexual maturity. Francke (1976) proposedmixed morphometricmethodto predict, by extrapolation from the knownsize and ageclasses of a partial life history, the size classes of instars not observed. Sexually maturespecimens(e.g., the motherof the youngproviding the partial life history or a series offield collected animals) are then comparedto the predicted size classes, and hypothesesabout their instar(s) are formulated (Francke 1976, 1979, 1981, Lourenqo1979, Sissomand Francke 1983). This mixed method resembles the theoretical method, but usesempirically obtained progression factors rather than the theoretical progression factor of1.26 proposed by Przibram and Megusar(1912).The primary objectives of this study were three. First, to analyze the life history ofFae/ovis coahuilae Williams, using both theoretical and empirical (including direct andindirect approaches) methods. Second, to review and compareall previous life historystudies on scorpions, using as manymethods as the published data allows. Finally, tocommentfurther on the strengths and weaknesses of each method.

MATERIALS AND METHODSAll measurementswere obtained using a dissecting microscope fitted with an ocularmicrometer calibrated at 10x. The measurementsare accurate to 0.1 mm,and thus have


3/20

FRANCKEANDSISSOM-REVIEWOF NUMBEROF MOLTSIN SCORPIONS 3

twosignificant digits. All statistical computations(mean,standard deviation) were carriedout to four digits, as recommendedby Steel and Torre (1960). The results were roundedoff, and are presented, to three digits because this numberis meaningfulwith respect to1.26, the value of the progression factor hypothesized by Przibram and Megusar(1912).Theoretical method.-If growth proceeds in steps which follow a geometric progressionthen the size of a given structure in youngand adult arthropods is related as A= ypnwhereA is the dimensionof the adult structure, Y is the dimensionof the same structurein a youngspecimenof knowninstar, P is the progression factor (1.26 in the theoreticalmethod, or the observed value in the mixed method), and n is the number of moltsrequired by the young specimen to reach adult size. Transformed into logarithms theequation becomes: log A = log Y + n log P. By rearrangement the working formula todetermine the numberof molts can be obtained: n = (log A- log Y)/log P. Since log 1.26= 0.1, the equation is simplified to n = 10 (log A- log Y).Twofemales (designated A and B) of Vae/ovis coahuilae which gave birth in thelaboratory, and 10 each of their respective second instar young(litters designated A andB, respectively) were used for analysis. Uponbirth, first instar scorpions climb onto theirmothers back, wherethey remain until a few days after their molt to second instar. Thus,females carrying youngof a knowninstar are occasionally caught. First instars are poorlysclerotized and measurementsobtained from them are not considered reliable for pur-poses of this methodfor determining the numberof molts to maturity. Amongvaejovidscorpions second instars are morphologically indistinguishable from subsequent instars,except perhaps by size (see Indirect methodbelow). Therefore, it is almost impossibledeterminethe instar of a small field-collected specimen,unless it happensto be a secondinstar riding uponits mother, which is whywe suggest the use of second instars with thismethod. On each specimen three structures were measured: carapace length, pedipalpchela length, and metasomalsegmentV length. There were no significant differences inthe dimensionsof each of the three structures betweenthe two litters: carapace lengths A= 1.55 + 0.05 (mean+- standard deviation), = 1.54 + 0.05, P(A= B)= 0.67 (t- test, 18d.f.); pedipalp chela lengths A= 1.76 +- 0.05, B = 1.76 +- 0.05, P(A= B) = 1.00 (t-test, 18d.f.); metasomalsegmentV lengths A = 1.42 -- 0.04, B = 1.44 +- 0.07, P(A= B) = 0.45(t-test, 18 d.f.). Therefore, the average dimensions(n = 20 for each structure)were usedin the analyses. In addition, to facilitate comparisonwith the results from other methods,the upper and lower size observations for each structure on second instars were extrapo-lated by multiplying x 1.26 to producethe theoretical distribution of size ranges for eachinstar up to the observedadult dimensions.Indirect method.-A comparison of size ranges of a sample of 80 specimens from RioGrande Village, Big Bend National Park, Brewster Co., Texas, and 100 specimens fromvarious localities in Arizona, NewMexico, and Texas revealed no trends in geographicvariations in size. Therefore, a total of 180 field collected specimensfrom throughoutthespecies range (the ChihuahuanDesert of North America) were used. The total numbersamples (not specimens) per month was: January 1, February 0, March1, April 3, MayJune 32, July 10, August 6, September 13, October 2, Novemberand December0. Thus,we consider that our samples adequately represent the species phenology. The samethreestructures used above were measuredon each individual. The meristic data were plottedas follows: carapace length versus pedipalp chela length, and carapace length versus meta-soma1segment V length.

The authors and several colleagues visually inspected the plots for gaps and clusters, asdone by previous investigators. There was considerable disagreement concerning (a)how


4/20

4 THE JOURNALOF ARACHNOLOGYmanysize classes are represented in the plots, and (b) what the limits of each class shouldbe. Weare unawareof any objective procedures available to determine the actual numberof clusters that are present in data sets similar to ours. Hierarchical clustering (Johnson1967, Helwig and Council 1979) was used to circumvent the second problem, i.e., todetermine the limits of each class-depending on howmanyclasses one wishes to recog-nize.Direct method.-A female (A) caught at Kermit, Winkler Co., Texas, on 1 April1978, gave birth to 41 young on 21 August 1978. A second female (B), from Castolon,Big BendNational Park, Brewster Co., Texas, collected on 8 August 1979, gave birth to33 young on 16 June 1980. All scorpions were maintained by previously describedmethods (Francke 1979, 1981).Mixedmethod.-In addition to the data used in the indirect method(particularly onthe size of sexually mature specimens), and the data obtained with the direct method(particularly progression factors), the morphometricdata obtained from two femaleswhich molted to maturity in captivity were used. Both females form part of a samplecollected at Rio GrandeVillage, Big BendNational Park, Brewster Co., Texas, on 29 July1978. Each female molted once and attained sexual maturity, one on 12 January 1980,after 1.5 years in captivity; and the other on 23 June 1981, after 3 years in captivity.

SPECIFIC RESULTSAND DISCUSSIONTheoretical method.-The predictions of this method, under the assumption of aconstant 1.26 progression factor, appear in Table 1. The predicted numberof molts bysecond instar structures to attain the size of their mothersare: 5.2 and 5.5 for carapacelength, 5.8 and 6.0 for pedipalp chela length, and 6.1 and 6.2 for metasomalsegment Vlength. Duringecdysis all exoskeletal structures are shed simultaneously, thus it wouldbeabsurd to postulate five molts for the carapace and six molts for the metasomalsegmentV of second instars to reach the respective sizes on female A. Rather, these data indicate

aUometry,with the three structures used deviating more or less from each other, deviatingfrom a constant rate within one structure at different molts, and also deviating fromthe assumedtheoretical progression faetor of 1.26. Averaging the predicted numberofmolts from the three different structures yields 5.7 and 5.9, respectively, as the averagetheoretical numberof molts neededby second instars to attain the sizes of females A andTable1.-Theoretical morphometricpredictions on the numberof molts (n) by secondinstareoahuilaescorpions(Y) to attain the size of their mothers(A andB). Valuesderivedusingthe equa-tion n = (log A- log Y)/logP, whereP is PrzibramandMegusars(1912)progressionvalueof 1.26.measurementsare lengthsof the structuresin millimeters.

FemaleA FemaleB SecondlnstarsA log A n B log B n Y log Y

Carapaee 5.2 0.716 5.2 5.5 0.740 5.5 1.55 0.190Pedipaip 6.8 0.832 5.8 7..1 0.851 6.0 1.76 0.246chelaMetasomal 5.8 0.763 6.1 6.0 0.778 6.2 1.42 0.152segmentVff=5.7 li= 5.9


5/20


Table2.-Theoreticalsize rangesfor consecutiveinstars of V. coahu~ae,derivedfromthe observedsize rangesof secondinstars assumingthat a progressionfactor of 1.26 is in operation(PrzibramandMegus~r1912).Themeasurementsare lengthsof the structures in millimeters.INSTAR Carapace Pedipalp Metasomalchela segmentVSecond 1.5 - 1.6 1.7- 1.9 1.4- 1.5Third 1.9 - 2.0 2.1 - 2.4 1.8- 1.9Fourth 2.4- 2.6 2.7 - 3.0 2.2- 2.4Fifth 3.0- 3.2 3.4 - 3.8 2.8- 3.0Sixth 3.8 - 4.0 4.3 - 4.8 3.5- 3.8Seventh 4.8 - 5.1 5.4 - 6.0 4.4 - 4.8Eighth 6.0- 6.4 6.8 - 7.5 5.6- 6.0FemaleA 5.2 6.8 5.8FemaleB 5.5 7.1 6.0B (Table 1). Since there is no such thing as a fraction of a molt we must round-offto thenearest integer: six. After six molts the initial second instars have becomeeighth instars,whichis the predicted instar to whichthe mothers belong.The theoretical predicted size ranges of the three structures for consecutiveinstars ofVae/ovis coahuilae, obtained by extrapolation appear in Table 2. The carapace lengths ofthe two mothers are between the predicted size ranges for seventh and eighth instars,whereas the observed pedipalp chela and metasomal segment V lengths are within thepredicted size ranges of eighth instars. Thus, accordingto this variation of the theoreticalmethodadult females are also predicted to be eighth instars.

Indirect method.-The hierarchical clustering algorithm starts off recognizing 180clusters, each madeup of a single individual. Euclideandistances are calculated and thetwo nearest neighbors are clustered, and so on successively until a single cluster of 180individuals is left. Proceeding backwards, the solutions present are for two subequalclusters (separated by the line labeled as 1 in Figs. 1 and 2), for three clusters (separatedby lines 1 and 2, respectively), for four clusters (separated by lines 1, 2 and 3, respective-ly), and so on. Thus, we can objectively establish accurate limits for any numberofclusters up to the total numberof individuals present in the data set, or size classes wewish to recognize, although only a maximumof 13 clusters are identified in Figs. 1 and 2.Furthermore, depending on how manyclusters are recognized we can calculate averagedimensionsof each structure for each size class, and from those obtain progression factorestimates (Table 3). Doingthis, however,does not resolve the critical problemof deter-mining howmanysize classes actually are present in the sample! Additional evidencecan be used to reduce the numberof possible size-classes that might indeed represent trueinstars. For example, knowingthat carapace length in second instars averages 1.55 ram,and in their mothers it measured5.2 and 5.4 ram, then the schemesin Table 3 where onlythree or four size classes are recognized can be eliminated as being unrealistic. By refer-ence to the average progression factors in scorpions of 1.28 + 0.04 (Polis and Farley1979), 95%confidence limits ( + 2 S.D.) of 1.36 and 1.20 could be used to dismissthose schemesin Table 3 whichrecognize less than five or morethan eight size classes asalso being unrealistic. Nonetheless,there still exist three viable alternatives; five, six orseven size classes (from second through sixth, seventh, or eighth instar) with no objectivemeansof choosing amongthem, and a lingering doubt about the elimination of unrealisticschemes.


6/20

6 THE JOURNAL OF ARACHNOLOGY

Direct method.-Chronologicaldetails of the postembryorticdevelopmentof Vae/oviscoahuilae, in the laboratory, are summarizedin Table 4. The youngfromboth littersunderwenttheir first molt at 9-12 days of age. A second molt was successfully completedby 16 specimens, of whichsix molteda third time. Fourspecimensmolteda fourth time,andof these three moltedonce again, entering the sixth instar at ages of 594 (female),939 (male), and1037 (female) days. The sixth instar male provedto be sexually mature(dissection uponits death revealed fully developed hemispermatophores),whereasthetwo females provedto be still immature(upontheir deaths dissection revealed underde-velopedovariuteri withoutany matureovarianfollicles).Morphometrieanalyses of the postembryonicdevelopmentare presented in Fig. 3 andTable5. Figure 3 presents data similar to those in Figs. 1 and2: carapacelength versuspedipalpchela length (circles), and carapacelength versus metasomalsegmentV length(triangles). In addition to the youngreared in the laboratory, the two mothersare in-eluded (M). Becausethe two laboratory reared females died in the sixth instar before

8

7T

5T

2T

TX 1,26 .__>

%..%

o % ,. i2 a

,%9

ooe% ".. ~". e-: "-

~o

FEMALES MALES =1 =1 =2-3 =2-3 =4-5 "=4-5e=6-7 A=6-7

I I I I I I I I I I I t I I2 3 4 5 6 7 8I I I I I I I i /2 3 4 5 6 7 8 9 1oPEDIPALP CHELA LENGTH

Fig. 1.-Logarithm X logarithm plot of carapace length versus pedipalp chela length for a field-caught sample of 180 Vae/ovis eoahuilae Williams. Circles represent females and small immatures(toosmall to be sexed accurately), and triangles represent males. Diagonal lines numbered1 through 12indicate consecutive splits in the data set determinedby hierarchical clustering procedures. For exam-pie, if one wishes to recognize four size clusters, their limits are defined by lines 1, 2. and 3. The sizeranges numbered2 through 8 along the axes represent the theoretical (X 1.26) limits for each sizeclass (FromTable 2).


7/20

FRANCKEAND SISSOM-REVIEWOF NUMBEROF MOLTSIN SCORPIONS 7

reaching sexual maturity or attaining the size of the mothers, we conclude that femalessometimes mature after the sixth instalThe statistics pertaining to instar size and progression factors are presented in Table 5.The average progression factor between successive molts for carapace length was 1.24, forchela length 1.26, and for segment V length 1.29; the grand average for all structuresthrough all molts recorded in the laboratory was 1.26 + 0.04.

Mixed method.-Because two of the three specimens which died as sixth instarshad not reached sexual maturity, we must resort to this method to elucidate furtherdetails of the life history of V. coahuilae. First, because we know: (a) that at least twothe three specimens required at least one additional molt before attaining sexual matur-ity, i.e., at last a seventh instar is present; (b) that the small male which matured as a sixthinstar belongs in the same size class as some subadult males, and (c) numerous adult males(Figs. 1 and 2) are considerably larger than the known sixth instar male, we hypothesizethat in this species males cart mature at either the sixth or seventh instars.

Second, we known that at least some females must mature at some instar after thesixth. The carapace lengths of the two known sixth instar, immature females were 3.4mmand 3.8 mm. Amongall the females (n = 23) dissected to examine the condition ofthe re )roductive tract there are no mature individuals within that size class; the smallest8-

7-6- 8]- . :

.t:. "5- 7T -~ "i1~ I()", e".-%*- \.4-6I ..:.v--- ,"" i, .... b

", A "T .. ,~* " 2O 3- ". " 5Z ". =.." I0 4 .. -,--". 1

< .",, ". ~ FEMALES MALESn,< .. ." \ .=1 -=1. 12(0 2- 3]- "" . " " " \ ,=2-3 -=2-3" "4 =4-5 =4-5 ". e=6-7 A=6-72.~ . "~

TX 1.26 ~ I--I t----{ H H I--] I.----I I---I2 3 4 5 6 7 8I I I I I I I I2 3 4 5 6 7 8 9METASOMALSEGMENTV LENGTH

Fig. l-Logarithm X logarithm plot of carapace length versus metasomalsegmentV. See le~nd toftg. 1 for explanation.

IlO


8/20


,.2-2,: ~

z z z ~

..~ ~ ~

4~4,~

~.,~, .~,

~,i,4

~e...~e4:

~ z ~ z

~ .... ~4M ~4~NNNm ~ ~

M~.~


9/20


Table 4.-Chronologyof Vaejovis coahuilae life history in the laboratory, ages and durations indays (mean + one standard deviation). Three specimens attained the sixth instar: one a sexuallymature male, and the other two were subadult females.INSTAR n DURATION CUMULATIVEAGE

S.D. range ~ + S.D. rangeFirst 9 - 12 9 - 12Second 16 229 54 139 - 291 201 54 151 - 303Third 6 216 93 105 - 326 444 102 308 - 542Fourth 4 177 59 89 - 210 636 101 517 - 752Fifth 3 208 114 77 - 285 857 233 594 - 1037mature females examined had carapace lengths of 4.5 to 4.6 ram. Therefore, we hypothe-size that unlike males, no females mature at the sixth instar. In addition to the scorpionsborn and raised in captivity, two field collected females molted once (to maturity)captivity (Fig. 3, data points connected by dashed lines). The smaller of these is thesame size as the captive-reared sixth instals; thus, we assume that it was caught as a sixthinstar and that it molted in the laboratory into a sexually mature seventh instar female.The second was considerably larger than the sixth instars when brought into the labora-tory, being almost as large as the hypothesized seventh instar female (above). Further-more, because (a) the hypothesized seventh instar female is considerably smaller than thetwo females which gave birth in the laboratory and the largest field sampled females(Figs. 1 and 2), and Co) the field caught female in question molted in captivity intosexually mature female of the size class of the mothers, we hypothesize that those largefemales represent eighth instals. Therefore, according to the mixed method we hypothe-size that V. coahuilae males mature at the sixth and seventh instars, and females matureat the seventh and eighth instals.

Table 5.-Morphometdcsof laboratory reared Vaejovis coahuflae, indicating size and progressionfactors for three structures, showngraphically in Fig. 3. Measurementsare lengths of the structures inmillimeters (mean~ one standard deviation). PF = progression factor associated with a given molt.Pedipalp MetasomalInstar n Carapace chela segment V

Second 15 1.55 0.06 1.79 0.06PF 15 1.23 0.04 1.25 ~ 0.05Third 14 1.90 0.06 2.23 0.09PF 5 1.22 0.04 1.26 0.09Fourth 5 2.34 0.09 2.86 0.11 ck~PF lc~+ 3 ~? 1.27 + 0.02 1.25 0.04 1.28Fifth 4 2.98 0.13 3.55 + 0.10 3.2PF lt5+ 2~9 1.26 0.04 1.32 0.08 1.44Sixth 3 3.73 0.30 4.67 0.42 4.6

1.41 0.031.30 0.061.84 0.101.29 0.082.42 0.13 ~91.272.9 - 3.11.24, 1.333.6, 4.OAverage PF 27 1.24 0.04 1.26 + 0.06 1.29 0.06Mothers A 5.2 6.8 5.8B 5.5 7.1 6.0


10/20

10 THEJOURNALOFARACHNOLOGY

Discussion.-Theresults obtainedamongthe various methodsused to analyze the lifehistory of F. coahuilaeare considerednext. First however,it is importantto note thatmaturityoften occurs at morethan one instar in scorpions (Table 6). Maturityat differ-ent instars is recognizedif the sampleincludes one size class whichcontains both im-matureand sexually matureindividuals. Thus, in the sampleused for the mixedmethodin this study the smallest sexually maturemales (confirmedby the presence of hemi-spermatophores) measured: 3.7 and 3.9 mmin carapace length, 4.8 and 5.0 mminpedipalp chela length, and 4.1 and 4.4 mmin metasomalsegmentV length, respectively,whereasthe largest subadult male (no hemispermatophores,nor fully developedparaxialorgans) measured4.1 mmin carapacelength, 5.2 mmin pedipalp chela length, and 4.4mmin metasomalsegmentV length. Therefore we assumethat those specimens, repre-senting the samesize class, also representthe sameinstar, whichin turn indicates that atleast in malessexual maturityis attainedat two different instars.

8-

M

71- ""... s"~".~"~ " t "\6"TI \ 7 t N*, ~. ,

5 ".,-.

4 "41-3?-2T

~,.. ,..,.

3..

2~.

=1 -=1 =2-3 "=2~ =4-5 "=4-5o=@-7 A=6-7

2 3 4 5 6 7 8] I I I I I I I I I "{ I I I "X 1.26 __.> ~ H I---I H ~ I----t I.-.--.I 2 3 4 5 6 7 8I

i i I i i i i i i2 3 4 5 6 7 8 9 10PEDIPALP CHELA LENGTH

METASOMALSEGMENTx/" LENGTH Fig. 3.-Logarithm X logarithm plot of carapace length versus both pedipalp chela length (circles)and metasomalsegment V length (triangles) for Vae]ovis coahuilae Williams raised in captivity. The

two sets of points linked by dashed lines represent field-caught females which molted once to attainsexual maturity in captivity. M= mothersof the two litters born and raised in captivity. Diagonal linesalong scatter diagram indicate size class limits observed. The size ranges numbered2 through 8 alongthe axes represent the theoretical (X 1.26) limits for each size class (from Table2).


11/20

FRANCKEANDSISSOM-REVIEWOF NUMBEROF MOLTSIN SCORPIONS 1|The theoretical methodpredicts successive size classes regardless of whether sexual

maturity is attained. Thus, the two female V. coahuilae whichgave birth in captivity werechosen as the size class criteria for determining sexual maturity, and the prediction thatthey represent eighth instars was corroborated by the mixedmethod. The applicability ofPrzibram and Megusars (1912) progression law (PF = 1.26) was tested against the em-pirical results (see Table 5) using Students t-tests (Steel and Torrie 1960). In all cases,whether each structure was considered at each molt, or whether one structure was con-sidered throughall molts, the results were the same: The empirical results are not signifi-cantly different (P > 0.05) from 1.26.The indirect methodyields inconclusive results with respect to the numberofinstarsto maturity in V. coahuilae. Althoughdiscrete clusters can be recognized, there are noobjective procedures to unmistakablyequate the presumedsize classes with actual instars.

GENERALDISCUSSIONThe information available on the stadia of adult scorpions is presented in Table 6.

Second instar and adult measurements of one or more structures, and or sexes, areavailable for the majority of the taxa included in that table. Those measurementswereused to predict the numberof molts betweenseeond instars and adults using the theoret-ical method(Table 7). Thus, for mosttaxa the results of at least two different methodsdetermining the numberof molts to maturity can be compared.Using the indirect methodof sorting specimens by size, Vachon(1948, 1951, 1952)postulated that Androctonusattstralis hector Koch,from Chellala, Algeria, maturedat theseventh instar. Vachon(1952) indicated that A. a. hector is the only subspecies presentin North Africa. Auber-Thomay(1974) reared the progeny of a female A. tmstralis L.,from the island of I)jerba, Tunisia (from where Vaehon1952 reported A. a. hector), andfound that both sexes mature at the eighth instar. Although we cant be sure of thesource of the discrepancy in this case, careful comparison of Auber-Thomaysdata(1974:47, fig. 1) and Vachonsillustrations (1952:162-163,figs. 208-213) suggestthe gap betweenthe presumedthird and fourth instars is unusually large in the latter. Thetheoretical method, based on measurements from Auber-Thomay(1974)and from esti-mates derived from Vachonsillustrations (1952:162-163, figs. 208-213), indicates thatsix molts wouldbe required by second instars to reach adult size (Table 7), reinforcingour suspicions that the indirect methodemployedby Vachonis faulty.Auber (1959) raised Belisarius xambeuiSimon, from the Pyrenees of France, to thethird instar, and by comparisonwith other presumedsize classes she recognized at leastten, and possibly eleven instars for sexually mature specimens. Francke (1976)calculatedthe progression factors separating Aubers size classes in B. xambeui,and the overallaverage of 1.19 was considered to be too low in comparison with those of 1.25 to 1.30actually observed in other scorpions. Thus, by extrapolation from the knownsize ofsecond instars and using the mixed method, Francke revised Aubers estimate downto 6-7 molts to maturity for that species. The theoretical methodusing Auberspublisheddimensions for an adult male of/~ xambeuipredicts 10 instars, whereas estimates basedon an adult male (hernispermatophore present) we examined are of only eight instars(Table 7). All of these hypothesesawait testing by the direct method.A more interesting problem is presented by Buthus oceitanus Amoreux. Vachon(1940) using the indirect methodpostulated that adults represent the seventh instar; thiswas confirmed by Auber (1963) who raised six males and six females of this species


12/20


Table 6.-List of species for whichthe instar(s) of sexually mature individuals were determined.= direct method, reared to maturity in captivity; I = indirect method, numberof stadia determinedvisually or motphomotrically from specimen samples; M= mixed method, with partial rearing incaptivity and morphometricextrapolation to account for specimensamples.

Taxon Sexually Method Sourcematureinstar(s)dd 99BUTHIDAEAndroctonusaustralis (L.) 8 8Androctom~australis hector Koch 7 7Buthotusalticola (Pocoek) 6Buthotu~minax occidentalis 6+7 6+7+8Vachon and StockmannButhus occitanus Amoreux

Buthus occitam~sparis (Koch)Centn~roides aguayoi MomnoCentruroides anchorellus AxmasCentruroides armadaiArmasCet, truroidesgracilis(Latrieile)Centruroides guanensis cubenMsMorenoIsometrus maculams(DeGeer)Orthochirusinnesi Simon7~tyus bahiensis (Perty)2Ttyus mattogrossensisBorelli7Ttyus serrulatus Lutz and Mello(parthenogenic)7TOussagmurus(Thomll)Ti~. us faseiolamsPessSaCHACTIDAEBelisarius xambeuiSimonEuscorpiusitalicus (Herbst)Megacormusgertsehi DiazDIPLOCENTRIDAEDiplocentrusspitzeri StahnkeNebohierichonticus (Simon)SCORPIONIDAEHeterometrus longimanus(Herbst)Pandinusgambiensi8 PocockUrodacusmanicatus(Thorell)Urodacusyaschenkoi (Birula)VAEJOVIDAEParuroctonusbaergi WilliamsandHadleyParuroctonu~mesaensis Stahnke

Uroctonus mordaxThorellVae/ovis bilineatus PocockVae]oviscoahuilaeWilliams

77+87 765+6 65+6 5+6

5+6 67 76+7 75+6 67 6+75 65 5+6

10or 117 or8

5+6

6 6+7

67or8

7+866

D Auber-Thomay1974I Vachon1948, 1952I Vachon 1951D Stockmann 1979I Vachon 1940, 1951D Auber 1963Mthis studyI Vachon 1951D Armas 1981D Axmas 1981D Armas 1981D Armas 1981D Ftancke and Jones 1982D Armas 1981D Probst 1972MShulov and Amitai 1960D Matthiesen 1970M Louren~o 1979D Matthiesen 1962D San Martin and Gambardella1966D Matthiesen 1971D Louren~o 1978I Auber 1959M Francke 1976D Angerman1957M Francke 1979M Francke1981I RosinandShulov1963D Schultze 19277+8 D Vachonet al. 19706 I Smith 19666 I Shorthouse 1971

7+8 7+8 I Fox 19757+87866+7

7+8 I Fox 19757 I Polis and Farley 19798 D Francke,in pressM Francke 19766 M Sissom and Francke 19837+8 M thisstudy


13/20


to maturity. However, Auber (1963:282-283, fig. 8) recognized "lexistence dunecategorie dindividus de grande taille" amongpreserved specimens, and to us thoselarge specimens represent an eighth instar. The theoretical method, using the averagepedipalp chela length of females reported by Auber(1963), predicts eight instars inoccitanus (Table 7). Thusit is possible that in Aubersstudy, as in the present one,large specimens were reared in captivity although they occur and can be recognized byresorting to other methods. Vachon(1951) postulated that adults represent the sixthinstar in Buthus occitanus paris (Koch), from Morocco. The reason for the differencebetween the nominate subspecies, found in Europe and northern Africa, and the sub-species paris is not known.However,Vaehons(1951) estimates are based on 11 speci-mens, six adults and five juveniles; and thus the sample size appears inadequate. Eitherthe second instar could be missing from the sample, or an oversize gap could appear inthe sequence due to the absence of one of the clusters used by this indirect method.Armasand Hemandez(1981) raised Centruroides anchorellus Armasin captivity andobtained sexually mature males and females at both fifth and sixth instars. The theoret-ical method, based on measurementsof one female and her litter, predicted that femaleto be a sixth instar (Table7).Armas and Hemandez(1981), and Francke and Jones (1982) raised Centruroidesgracilis (Latrielle) in captivity and obtained sexually mature seventh imtar females.However,based on specimensactually raised in captivity, the theoretical methodpredictseight instars for those seventh instar females! The observedprogression factors for cara-pace length for females of this species average slightly over 1.31 (Franeke and Jones1982), which fully accounts for the discrepancy in the theoretical predictions (11.267 ~ I x 1.316 ~ 5.05).Francke (1981) using the mixed approach hypothesized that l~ploeentrus spitzeriStahnkematures by the sixth instar. The predictions of the theoretical methodare for 4.5molts between second instars and adults, which would thus represent either the sixthor the seventh instal Althoughit is possible that different individuals of D. spitzeri canattain sexual maturity at two different instars, the ambiguousresults of the theoreticalmethod applied to an individual female are indicative of the problems occasionallyencountered with this method.Angerrnan(1957) raised Euscorpiusitalicus (Herbst) in captivity and found that malesand most females mature as sixth imtars, which is what the theoretical methodpredicts(Table 7), and a few females molt once moreto mature at the seventh instar.The theoretical methodpredicts seven imtars for Isometrus maculatus(DeGeer)(Table7), which is indeed what Probst (1972) obtained for males and most females usingdirect method. Approximately 10%of the females of this species, however,matureas sixth instars.Using the mixed method Francke (1979) hypothesized that Megaconnusgertschi I)iazmatures at the eighth instar. Theoretical considerations, however,predict nine imtars foradult females (Table 7). Whetherfemales actually mature at the eighth, the ninth, or boththe eighth and ninth instars in this species is not known,and thus it is not possible todetermine which methodis more reliable in this case. The advantage of the mixedmethodis that it is based on empirical progressionfactors.Rosin and Shulov (1963) estimated indirectly that Nebo hietichonticus (Simon)matures at either the seventh or the eighth instar. Theoretical predictions based onmeasurementsby Francke (1981) indicate that sexually mature females represent theeighth instar (Table 7).


14/20


Table 7.-Theoretical predictions of the numberof instats (Ni) to sexual maturity using secondinstar (II) and adult (A) measurements(in millimeters), and a progressionfactor of 1.26 (seefor details of calculation method).

Taxon Structure II A n Ni SourceA. australis carapace L 4.1 16.5 6.0 8 Auber-Thomay1974A. australis hector carapaceL (units) 2.3 9.3 6.0 8 Vachon 1952B. xambeui chela L 1.7 11.2 8.0 10 Auber 1959chela L 7.6 6.3 8 pets, obs.B. occitanus chela L 3.1 11.5 5.7 8 Auber 1963B. occitanus paris movablefinger L 2.2 9.0 6.1 8 Vachon 1951C. anchorellus carapace L 1.5 4.0 4.3 6 Atmas, pets. comm.C. gracilis carapace L 2.2 8.6 5.8 8 Armas, pers. comm.carapace L 2.1 8.1 5.8 8 pets. obs.chela L 3.4 13.8 6.1 8 pets. obs.segment V L 2.1 9.3 6.4 8 pets. obs.D. spitzeri carapace L 2.2 6.0 4.4 6 Francke 1981chela L 3.3 9.6 4.6 7 Francke 1981segment V L 1.6 4.6 4.5 6-7 Francke 1981E. italieus carapace L 1.5 3.9 4.1 6 Angerman1957I. maculatus caxapac L 1.6 4.6 4.5 6-7 Probst 1972chela L 2.6 8.5 5.2 7 Probst 1972

segment V L 1.6 5.4 5.1 7 Probst 1972211.gertschi carapace L 1.6 7.0 6.5 8-9 Franeke 1979chela L 2.5 12.1 6.8 9 Franeke 1979segment V L 1.1 6.1 6.9 9 Francke 1979N. hie~chonticus carapace L 2.7 10.8 6.0 8 Francke 1981O. innesi total L d 12 28 3.7 6 Shulov &Amitai 19609 12 32 4.2 6 Shulov & Amitai 1960P. gambiensis movablefinger L 4.5 18 6.0 8 Vachonet al. 1970P. baergi carapace L (small A) 1.9 5.8 4.8 7 Fox 19750arge A) 6.6 5.4 7 Fox 1975P. mesaensis carapace L (small A) 2.2 7.2 5.1 7 Fox 1975

(large A) 8.9 6.0 8 Fox 1975carapace L 1.8 6.9 5.9 8 pets. obs.T. bahiensis movablefinger L 3.3 7.5 3.6 6 Matthiesen 1970T. mattogrossensis carapace L 1.8 3.7 3.1 5 Lourengo 1979movablefinger L 2.1 4.4 3.2 5 Lourengo 1979segment V L 1.8 5.0 4.4 6 Loutengo 1979T. fasciolatus carapace L (small c0 2.5 4.6 2.6 5 Louren~o 1978(reed. ~) 6.1 3.9 6 Lourengo 1978(large d) 8.5 5.3 7 Lourengo 1978(small 9) 5.1 3.1 5 Lourengo 1978(reed. 9) 6.0 3.8 6 Loureno 1978(largec2) 7.0 4.4 6 Louren~o 1978U. mordax carapace L 2.0 7.3 5.6 8 Francke 1976chela L 3.3 12.9 5.9 8 Ftancke 1976segment V L 1.6 8.6 7.1 9 Franeke 1976U. manicatus log carapace L 0.3 0.8 4.8 7 Smith 1966log tail L 0.8 1.3 5.1 7 Smith 1966[ yaschenkoi carapace+ tail L 1.5 4.0 4.2 6 Shotthouse 1971chela L 0.8 1.7 3.3 5 Shorthouse 1971V. bilineatus carapace L (small A) 1.4 3.5 3.9 6 Sissom &Francke 1983(large A) 4.2 4.6 7 Sissom & Francke 1983chela L (small A) 1.6 4.2 4.1 6 Sissom & Francke 1983(large A) 5.0 4.9 7 Sissom & Francke 1983segmentV L (small A) 1.3 3.8 4.5 6-7 Sissom & Ftancke 1983(large A) 4.5 5.2 7 Sissom & Francke 1983


15/20


Shulov and Amitai (1960) using laboratory observations on early instars of Ortho-ehfrus innesi Simon, supplementedwith indirect techniques, proposed that males matureas fifth instars, whereasfemalesmatureas sixth instars (Table 6). Basedon their measure-ments of total length (which in someinstances can be influenced by the nutritionalcondition of the animal) the theoretical methodpredicts that both males and femalesrepresent sixth instars (Table 7). According to Shulov and Amitai fourth instar malesmeasure 15-25 ram, and females measure 15-24 ram, whereas adult measure 26-30.5 mmand 28-35 ram, respectively. Basedon the broad overlap in size amongadults it is difficultto believe that a different numberof molts would be required after the fourth instar.Thus, we consider that adults of both sexes should be regarded as sixth instars untilstronger evidence to support the presumedsexual differences is presented.Males and females of Pandinus gambiensis Pocock raised in captivity matured asseventh and eighth instars (Vachonet al. 1970). There is a paucity of measurementsgivenin that study, but based on an estimated average adult pedipalp chela movablefingerlength of 18 ram, the theoretical prediction calls for eight instals in this species (Table7).Fox (1975) used indirect, univariate techniques to postulate that in aruroctonusbaergi Williams and Hadley, both males and females attain maturity at the seventh andthe eighth instars. Based on Foxs measurementsthe theoretical methodpredicts thateven the largest specimensare seventh instars (Table 7). Unless this species has a rathersmall progressionfactor (1 x 1.22s ~- 1 x 1.27v ~- 5.0) it is difficult to justify the recogni-tion of the hypothesizedeighth instar in this species. Francke(in press) analyzed the in-direct methodused by Fox and failed to Fred objective criteria by which Foxs resultscould be repeated-in our opinion the strongest argument against the indirect method.

Paruroctonus mesaensis Stahnke has received more attention than any other scorpionwith respect to its life history. Fox (1975)used the same univariate technique mentionedabove to postulate that both males and females mature at the seventh and eighth instarsas well. Polis and Farley (1979) used an indirect, bivariate methodto arrive at the conclu-sion that both males and females mature at the seventh instar, and categorically deniedthe existence of an eighth instar in . me~aensis.Applyingthe theoretical methodto theirdata we predict eight instals (Table 7). Likewise, applying the theoretical methodtofemale and her captive-born young, we predict eight instars (Table 7). Francke (in press)raised one specimen in captivity from second instar to sexual maturity at the eighthinstar; the indirect methods of Fox (1975) and Polis and Farley (1979) are criticallyexaminedin that contribution and found to lack objectivity.Matthiesen(1970) raised l~tyus bahiensis (Perry) in captivity. His results indicate thatmales mature as fifth instars, whereas females mature as fifth and sixth instars. Thetheoretical method, using a second instar measurementfrom Matthiesen (1970) andadult female measurementfrom Mello-Leit~io (1945), predicts that the latter is a sixthinstar-in conformitywith the empirical results.Louren~o(1979) used the mixed method to hypothesize that ~tyus mattogrossensisBorelll attains sexual maturity at the sixth instar. Theoretical calculations predict thepresence of only five instars based on two structures, and six instars based on a thirdstructure (Table 7). The actual average progression factors (over two molts)reportedLouren~oare 1.20 for carapace length, 1.25 for pedipalp chela movablefinger length,1.29 for metasomal segment V length. Thus, the low progression factor for carapacelength in this species might account for someof the discrepancies noted between themixed and theoretical methods. Examination of the data pooled by Louren~oto obtainaverage adult dimensionssuggests an alternative explanation though. Onefemale and five


16/20

16 THE JOURNALOF ARACHNOLOGYmales have measurementsin the following ranges: carapace length 3.4-3.6 mm,movableFinger length 4.1-4.4 mm,and metasomalsegment V length of 4.2 mmin the female and4.7-5.4 mmin the males. Another three females and one male measure 3.8-4.4 mmforcarapace length, 4.6-5.2 mmfor movablefinger length, and 4.8-5.5 mmand 5.8 mmformetasomalsegment V length, respectively. If those two size classes indeed representdifferent instars, then the theoretical methodpredicts that the formerare fifth instars andthe latter are sixth instars. If that is the case, the predictions of the theoretical methodconformwith those of the mixedmethod, and the differences noted above are an artifactdue to the combination of measurementsfrom two distinct size classes by Lourenfo.

Lourenfo(1978) raised Tityus fasciolatus Pess6a in captivity and succeededin obtain-ing nine sexually mature specimens:five sixth instar females, two fifth instar males, andtwo sixth instar males. Furthermore, using the mixedmethodLouren~opostulated that avery large field caught male represented the seventh instar. The theoretical methodpredicts that field caught adults of average dimensions represent sixth instars in bothsexes, the smallest adults wouldrepresent fifth instar adults in both sexes, and very largemales wouldrepresent the seventh instar (Table 7). Thus, in general the theoreticalmethodis in full agreementwith the empirical observations, differing only in the predic-tion that somefemalesmature as fifth instars. That prediction awaits testing by the directmethod.

Francke (1976) raised one specimen of Uroctonus mordaxThoreUto the fifth instar,and predicted by extrapolation that adults represent seventh instars. The theoreticalmethodpredicts that, based on carapace and pedipalp chela length, adults should be inthe eighth instar, whereas based on metasomalsegment V length adults should be ninthinstars! The problemsof allometry and of progression factors considerably greater than1.26 (1.31, 1.30 and 1.41 for carapace length, pedipalp chela length and segmentlength, respectively) point clearly to someof the shortcomingsoccasionally encounteredby the theoretical approach.Smith (1966) used the indirect methodto analyze the life history of Urodacusmani-catus (Thorell). He indicated, based on only 23 specimens, that males have six distinctsize classes and instars. Amongfemales, however, based on 21 specimens, he only recog-nized five distinct size classes, but postulated six instars anyway!Thus, the precarious andsubjective basis of this approach becomesapparent even with small sample sizes. Smithprovided no measurementsor progression factor estimates associated with those putativeinstars. Averagedata (log length of prosomaand log length of tail for second instars andadults [adult carapace length corrected slightly using Kochs 1977 data]) presented inSmiths Fig. 1, yield predictions that adults represent the seventh instar (Table 7). Addi-tional data are neededbefore either the indirect or the theoretical methodare consideredinadequatein this case.Shorthouse (1971) used both the indirect and mixed methods to analyze the lifehistory of Urodacusyaschenkoi (Birula). Using a sample of 210 specimens he obtainedfive distinct size classes, and since first instars wereexcluded,postulatedthat there are sixinstars in both males and females. In addition, he reported 79 cases where markedspeci-mensmolted during that study, and the progression factors observed agreed closely withthose derived by the indirect method. Finally, he excavated five burrows and obtainedboth the scorpion inhabiting it and its exuviumfrom the preceding instar. These observa-tions support the progression factors estimated from the morphometricanalysis and fromthe measurementsof the 79 specimens which presumably molted during the study. Thetheoretical method, using the combined carapace + metasomal segments I-V length


17/20

FRANCKEANDSISSOM-REVIEWOF NUMBEROF MOLTSIN SCORPIONS 17predicts that adults are indeed in the sixth imtar (Table 7). However,predictions basedon pedipalp chela length are that adults are only in the fifth instar (Table 7). Theprogres-sion factors reported by Shorthousefor chela length are 1.26 + 0.01 (n = 43), 1.22 -+ 0.01(n = 23), and 1.19 +- 0.01 (n = 13) for the molts from second to third, third to fourth,and fourth to fifth instars, respectively, whereasthey are 1.31 _+ 0.01, 1.28 -+ 0.01 and1.27 + 0.02 for carapace + taft segments. Koch(1977:188) describes the chela on thisspecies as "short and squat" and the "fingers moderately short to short." Thus, thedifferential rate of growth for chela length in successive instars (as the hand becomesprogressively wider) accounts for the incongruence between the theoretical predictionsand the observations by Shorthouse.Sissom and Francke (1983) obtained a partial life history for Vae]ovis bilineatusPocockin captivity, and used the mixedmethodto hypothesize that adult females are inthe sixth instar. Three large field caught females are within the size range predicted forseventh instars; however, because of possible variability in scorpion size at birth theydesignated those females as large sixth instars. The theoretical methodusing observedsizeranges for adult females predicts that they represent both the sixth and the seventh instar(Table 7).The results of the study on V. coahuilae showthat the theoretical and mixedmethodsyield congruent results. However,the indirect, morphometricapproach produced incon-clusive results and suffers fromlack of objectivity.

SUMMARYAND CONCLUSIONSThere have been 19 scorpion life histories determined using the direct method, andtwo of those actually represent corroborations of previous direct methodresults (T.

serrula~s and C. gracilis]. Measurementsof second instar and adult structures wereprovided for 10 of the 17 taxa studied, enabling us to evaluate the theoretical method.Ineight taxa: A. australis, C. anchorellus, E. italicus, I. mac~datus,P. gambiensis,P. mesaen-sis, T. bahiensis, and T. fasciolatus, the predictions of the theoretical methodagree withthe empirical observations. However,in two taxa R occitanus and C. gracilis the twomethods differ. Only seven instars of B. occitanus were reared in the laboratory, andlarger than average adults were reported from the field. The theoretical methodpredictedeight instars whichis not necessarily incorrect but merely remains untested. Centruroidesgraeilis exhibited average progression factors larger than most other scorpions, accountingfor the erroneous predictions by the theoretical method.There have been eight studies based on partial life histories which used the mixedmethodto predict the instar(s) of maturity. On one of these, B. xambeui, the mixedmethodwas used to propose an amendmentto results obtained by the indirect method,and predictions derived using the theoretical methodagree with those of the mixedratherthan the indirect approaches. Ontwo others, R occitanus and V. coahuilae, the mixedmethodwas used to supplementempirical results in explaining the presence of larger thanlaboratory reared males. The theoretical methodpredicts eight instars for large B. occi-tanus and V. coahuilae. Thesame situation probably applies to studies on V. bilineatus.In T. mattogrossensis the mixed and theoretical methods differ by one instar in theirpredictions, and the discrepancy is probably due to an artifact in the characterization ofadults. Finally, in three taxa, D. spitzeri, M. gertschi, and U. mordax,the theoreticalmethodpredicts one more instar than the mixedmethod. The growth rates (= progressionfactors) in those three taxa are consistently larger than 1.26, whichis the reason extrapo-lation using the mixedmethodwas originally proposed.


18/20

18 THE JOURNALOF ARACHNOLOGYThere have been 12 life history studies using the indirect method. On three of thetaxa, A. australis hector, B. occitanus paris, and P. mesaensis(two indirect studies) the

results obtained by indirect methodshave been contradicted in part by empirical resultsobtained by rearing the species in question. For B. xambeui the mixed methodand thetheoretical methodraise serious doubts about the results obtained by indirect methods.In the case of B. alticola lack of moristic data makeit impossible to analyze furtherdetails by resorting to the theoretical method. For P. baergi the indirect methodpre-dicted maturity at the seventh and eighth instars, and the theoretical methodindicatesthat only seven instars are necessary to account for even the largest specimens. For B.occitm~us, IV. hierichonticus, U. manicatus,and U. yaschenkoi the predictions from theindirect and theoretical methodsare similar, but must be tested empirically before theircorrectness is ascertained. Finally, our attempts to determine the life history of Kcoahuilae using the indirect methodwere inconclusive.The most significant difference betweenthe various methodslies in the verifiability ofthe results. The direct, empirical approachproduceshard data, subject to testing by thecriterion of repeatability. The theoretical and mixed methodsare rigorous enough tosatisfy the criterion of repeatability and yield hypothesessubject to testing by the ac-quisition of hard data, i.e., by resorting to the direct method. The theoretical methodgives misleading results whenallometric growth is experiencedby certain structures, andknowledgeof which structures are affected can lead to improvedpredictions. However,the advantage of using the mixedmethodis that the observed progression factors providea reliable measureof allometry, rather than having to estimate it. The indirect methodsometimesyields inconclusive results and suffers primarily from a lack of objectivitywhichprevents repeatability.

ACKNOWLEDGMENTSJames C. Cokendolpher and William M. Rogers helped with the laboratory chores.

MontA. Cazier, Linda Draper, and James V. Moodyassisted in collecting scorpions in theChihuahuanDesert between 1970 and 1980. James C. Cokendolpher, Gary A. Polis, GregSpicer, and Eric C. Toolson constructively criticized various drafts of the manuscript.Lorie A. Prien processed endless drafts of the typescript. Our warmestappreciation goesto all of them. The work was supported in part by the Institute for MuseumResearch atTexas Tech University.LITERATURECITED

Angcnnan,H. 1957. UbetVerhalten, Spermatophorenbildungund SinnesphysiologieyonEuscorpiusitalic-us Herbstuudve~vandtenArten.Z. Tierpsycbol.,14:276-302.Axmas,L. F. de, and N. Hernandez.1981. Gestaci6ny desarrollo postembrionarioen algunosCen-rruroides(Scorpiones:Buthidae)de Cuba.Poeyana,No. 217, 10 pp.Auber,M.1959.Observationssur le biotopeet la biologie du scorpionaveugle:BelisariusxambeuiSimon.Vieet Milieu,ser. C. 10:160-167.Auber,M.1963.Reproductionet croissancede ButhusoccitanusAmoreux.Ann.Sci. Nat. (Zoo1.),set., 5:273-286.Auber-Thomay,M.1974. Croissanceet reproductiond Androctonusaustralis (L.). Ann.Sci. Nat.(Zoo1.),12set., 16:45-54.


19/20


Dyar, H. 1890. The numberof molts in lepidopterous larvae. Psyche, 5:420-422.Fox, W. K. 1975. Bionomicsof two sympatric scorpion populations. Ph.D. Thesis, Arizona StateUniversity, Tempe,Ariz., 85 pp.Ftancke, O. F. 1976. Observations on the life history of UroctonusmordaxThorell. Bull. BritishArachnol. Soc., 3:254-260.Francke, O. F. 1979. Observations on the reproductive biology and life history of Megacormus

gertschi Dlaz. J. Araehnol., 7:223-230.Francke, O. F. 1981. Birth behaviorand life history ofDiplocentrusspitzeri Stalmke.Southwest.Nat.,25:517-523.Francke, O. F. in press. The life history ofParuroctonusmesaensisStahnke(Scorpiones, Vaejovidae).J. Arachnol.Francke,O. F. and S. K. Jones. 1982. Thelife history ofCentn4roidesgracilis(Latrielle). J. Arachnol.,10:223-239.Helwig, J. T. and K. A. Council. 1979. SASUsers Guide, 1979Edition. SASInstitute Inc., Cary,North Carolina, 494 pp.Johnson, S. C. 1967. Hierarchical clustering schemes.Psychometrika,32:241-254.Koch, L. E. 1977. The taxonomy,geographic distribution and evolutionary radiation of Australo-Papuanscorpions. Rec. West. Austr,lia~ Mns., S :83-367.Louren~o, W. R. 1978. Etude sur les scorpions appartenant au "complexe"Tftyus tHvittatus Krae-pelin, 1898et, en particuliet, de la sous-especeTftyus trivittatus fasciolatus Pess6a1935.These3ecycle Univ.Pierre et MarieCurie, Paris, 128pp., 55 pis.

Louren~o, W. R. 1979. Le scorpion Buthidae: 7iO,us mattogrossensis Borelli, 1901 (Morphologie,ecologie, biologic et developpementpostembryonnaire).Bull. Mus.Nat. Hist. Nat., Paris, 4e set.,1:95-117.Matthiesen, F. A. 1962. Parthenogenesisin scorpions. Evolution, 16:255-256.Matthiesen, F. A. 1970. Le developpementpost-embryonnairedu scorpion Buthidae: 7~tyus bahiensfs(Petty, 1834). Bull. Mus.Nat. Hist. Nat., Paris, 2e set., 41:1367-1370.

Matthiesen,F. A. 1971. Observationson four species of Brazilian scorpionsin captivity. Rev. BrasileiraPcsq. Mcd.Biol, 4:301-302.Mello-Leit~o, C. de. 1945. Escorpi6es Sul-Americanos. Arq. Mus. Nac., Rio de Janeiro, 40:1-468.Polls, G. A. and R. D. Farley. 1979. Characteristics and environmental determinants of natality,growth and maturity in a natural population of the desert scorpion, Paruroetonusmesaensis. J.Zool., London, 187:517-542.Probst, P. J. 1972. Zur Fortpflanzungsbiologie und zur Entwicklungder Giftdrussen bein SkorpionIsometrus maculatus(DeGeer, 1778). Acta Tropica, 29:1-87.Przibram, H. and F. Megnsfir. 1912. Wachstummessungenan Sphodromantisbioeulata Burm. 1. Langeund Masse. Arch. EntwMech.Org., 34:680-741.Rosin, R. and A. Shulov. 1963. Studies on the scorpion Nebohferichonticus. Ptoc. Zool. Soc. London,140:547-575.San Martin, P. R. and L. A. de Gambardella. 1966. Nueva comprobaci6nde la partenogenesis enI~.us serrulatus Lutz y Mello Campos1922. Rev. Soc. Entomol.Argentina, 28:79-84.Schultze, W. 1927. Biology of the large Philippine forest scorpion. Philippine J. Sci., 33:375-389.Shorthouse, D. J. 1971. Studies on the biology and energetics of the scorpion UrodacusyaschenkoLPh.D. Thesis, Australian National University, 163 pp.Shulov, A. and P. Amitai. 1960. Observationssur les scorpions Orthochin~sinnesi E. Sim., 1910ssp.negebensisnov. Arch. Inst. Pasteur Algiers, 38:117-129.Sissom, W. D. and O. F. Francke. 1983. Post-birth developmentof Vae/ot#s bilineatus Pocock. J.Arachnol. 11:69-75.Smith, G. T. 1966. Observationsof the life history of the scorpion UrodacusabruptusPocock(Scor-pionidae), and an analysis of its homesites. Australian J. Zool., 14:383-398.Steel, R. G. D. andJ. H. Torrie. 1960.Principles and Proceduresof Statistics with Special Referencetothe Biological Sciences. McGraw-HillBookCo., Inc., NewYork, Toronto, London.481 pp.Stockmann, R. 1979. Developpementpostembryonnaire et cycle dinternme chez un scorpion Buthi-dae: ~thotus minaxoccfdentalis (Vachonet Stockmann).Bull. Mus.Nat. Hist. Nat., Paris, 4e set.,1:405-420.


20/20


Vachon,M. 1940. Sur la systemarique des scorpions. Mere. Mus.Nat. Hist. Nat., Paris, 13:241-260.Vachon,M. 1948. Etudes sur les Scorpions. Arch. Inst. Pasteur Algerie, 26:441-481.Vachon,M. 1951. Scorpions collectes au Marocpar MM.P. Strinati et V. AeUen(MissionScientifiqueSuisse au Maxoc,Aout-Septembm1950). Bull. Mus. Nat. Hist. Nat., Paris, 2 sen, 23:621-623.Vachon,M. 1952. Etudessur les Scorpions. Inst. Pasteur Algerie, Alger. 482 pp.Vachon, M.. R. Roy and M. Condamin. 1970. Le developpement postembryonnaire du scorpionPandinusgcmbiensisPocock.Bull. Inst. ft. Afriquenoim, set. A., 32:412-432.

Manuscriptreceived July 1982, revised April 1983.

1984 - francke & sissom

Documents