When Is the Sex Ratio Biased in Social Spiders?: Chromosome Studies of Embryos and Male Meiosis in Anelosimus Species (Araneae, Theridiidae)

10 Pages·1991·5.2 MB·English

by  Leticia Avilés

1991. TheJournal ofArachnology 19:126-135
WHEN IS THE SEX RATIO BIASED IN SOCIAL SPIDERS?:
CHROMOSOME STUDIES OF EMBRYOS AND MALE MEIOSIS
IN ANELOSIMUS SPECIES (ARANEAE, THERIDIIDAE)
Leticia Aviles': Museum of Comparative Zoology, Harvard University, Cambridge,
Massachusetts 02138 USA and Department ofIntegrative Biology, University ofCal-
ifornia, Berkeley California 94720 USA
Wayne Maddison': Department ofIntegrative Biology, University ofCalifornia, Berke-
ley, California 94720 USA
Abstract. Embryochromosomepreparationsoffourspeciesofsocialspidersofthegenus
Anelosimusshow thatthe two species known orsuspectedto form permanent, multigener-
ationalcolonies,A. eximiusandA. domingo, haveahighlyfemale-biasedprimarysexratio.
Anelosimusjucundus and A. studiosus, on the other hand, are shown to produce an even
numberofmalesand females. The magnitudeofthe bias ofA. eximiusembryos is similar
to that reported for young preadult spiders ofthis species, therefore ruling out differential
mortality ofjuveniles as the cause ofthis species’ sex ratio bias. Chromosome counts of
nuclei in second division ofA. eximius male meiosis indicate that nuclei destined to yield
sonsand daughters are produced in equal numbers. Therefore, the sex ratio biasing mech-
anism in this species must act after male meiosis and before egg laying. The question of
how early the sex ratio bias arises still needs to be resolved in other social spiders. We
discuss some methodological and theoretical complicationsassociated with measuring sex
ratiosatdifferent stagesofthe lifecycle and present a fast andreliable techniqueto obtain
embryo chromosome preparations.
Theoccurrenceofhighlyfemale-biasedsexra- ciplehavebeenpointedoutbyHamilton(1967),
tiosamongadultsofseveralspeciesofsocialspi- who first notedthat biased sex ratios can evolve
ders has been known since the 1960’s (Buskirk when the assumption of panmixia, implicit in
1981), but there has been little study ofexactly Fisher’s argument, is violated. Parasitic and fig
when in the spiders’ life cycle the sex ratio bias wasps (e.g., Werren 1980; Waage 1982; Herre
arises. Knowledge ofthe timing ofthe sex ratio 1985) and hummingbird-flower mites (Wilson
bias is important on at least two accounts: first, andColwell 1981)arenotableexamples. Asfirst
fromanevolutionarypointofview,itwouldhelp noted by Aviles (1983, 1986), social spiders ap-
us determine whether differential parental in- pearto represent another case in which Fisher’s
vestmentisinvolvedinbiasingthesexratio;and, principle has been violated. In most social spi-
second, from a physiological point of view, it ders, however, this violation has not been con-
would bring us closer to identifying the mecha- firmedbecausetheirsexratiohasbeenmeasured
nismbywhichthesexratiobiasisaccomplished. late enough in the spider’s life cycle (usually
Fisher’s principle (Fisher 1930) states that, at amongadults) thathighermale mortalityduring
equilibrium,thetotalparentalinvestmentinoff- the preadult oradult instarscannot be ruled out
spring ofeach sex should be equal. Departures as the cause oftheir sex ratio bias. As noted by
from this equilibrium should bring about selec- several authors (Leigh 1970; Chamov 1982;
tion to restore an even sex ratio because indi- Trivers 1985), ifbiased sex ratios are due solely
vidualsoftheraresexwouldhaveareproductive tomortalityoccurringaftertheendoftheperiod
advantage. Exceptionsto Fisher’ssex ratioprin- ofparental investment then Fisher’s principle is
not violated. In three species, Anelosimus exi-
‘Present address: Dept, ofEcology and Evolutionary miusKeyserling,AchaearaneawauLeviandSte-
Biology,UniversityofArizona,Tucson,Arizona85721 godyphus dumicola Pocock, there is indirect ev-
USA. idence that more females are actually being
126
AVILES & MADDISON-SOCIALSPIDER SEX RATIO 127
produced. This evidence comes either from MATERIALS AND METHODS
young-preadult sex ratios estimated in isolated
natural colonies (Aviles 1983, 1986) or from Chromosomepreparationswereobtainedfrom
preadult/adult sex ratios measured among in- embryos and males collected in Ecuador from
dividualsraised undercontrolledconditions, ei- naturally occurring colonies of the four Aneio-
ther from egg sacs (Vollrath 1986b; Lubin and simusspecies. The sexofanembryo, orwhether
Crozier 1985; Lubin in press) or from a colony a nucleus in a spider testis is going to become a
maintained in the laboratory(SeibtandWickler male- or a female-producing spermatozoid, can
1988). Thesemeasurementsprovideevidenceof be determinedcytologicallythanks to thediffer-
an early bias, but may still be affected by mor- encein chromosome numberbetween male and
tality during thejuvenile instars. female spiders. The most common mechanism
Inthispaperwe presentacytogeneticmethod ofsexdeterminationinspidersinvolvestwopairs
that makes it possible to directly determine the of X chromosomes, with the two members of
sex of a developing embryo and, therefore, to eachpairpresentinfemales(X,X,X X andonly
2 2)
measure the sex ratio before mortality becomes one in males (X,X20) (White 1973). The sex of
a factor. We apply this method to four species an individual egg or ofa developing spermato-
ofthegenusAneiosimusSimon(Levi 1956, 1963) zoid can therefore be simply determined by ob-
presentinEcuador.Twoofthese,A. eximiusand taining its chromosome count.
A. domingo Levi, are among the most social in Aneiosimuseximiuseggsacswere collected in
the genus, with spiders that cooperate in prey July 1988 from two colonies, approximately 1.5
capture and brood care and share a permanent km apart, near the Recinto A. Perez Intriago,
communal nest generation after generation (A. ICm 113,Quito-Pto.Quitoroad(0°6'N:79°5'W).
eximiusreferences:Brach 1975;Tapia&DeVries AneiosimuseggsacswerealsocollectednearPer-
1980; Christenson 1984;A. domingoreferences: ez Intriago in June 1989 from one colony found
Levi&Smith 1982; Rypstra&Tirey 1989). The in a forest pocket near the Silanche river. Ane-
othertwo,A.jucundusO. P.- Cambridge andA. iosimusjucundus sacs were collected from two
studiosus Hentz, on the other hand, are known coloniesinCrucita,Manabi(0°52'S:80°33'W),in
toexhibitalessadvancedformofsocialitywhere August 1988 andA. studiosusnearCalderon, Pi-
the offspring ofa single female remain together chincha (0°6'S:78°27'W), in July 1989. Males of
for the early part oftheir life cycle but disperse the four species were collected from the same
beforereachingadulthood(Brach 1977;Nentwig sites as the egg sacs, except that in addition one
& Christenson 1986). UnlikeA. eximius, A.ju- maleofA. studiosuswascollectedfromElTingo,
cundus and A. studiosus have been reported to Pichincha(0°17'S:78°27'W). Egg sacs and males
have 1:1 preadult or adult sex ratios (Fowler & were brought alive to the laboratory where the
Levi 1979; Nentwig & Christenson 1986; Voll- preparations were made.
rath 1986brearedoffspringfromtwoA.jucundus The technique we developed to obtain chro-
sacsandobtainedanevensexratio).Aneiosimus mosomepreparationsfromindividualeggsisde-
domingosexratios have notbeen previouslyre- scribed in the Appendix. Basically, it is much
ported. liketypicalaceticsquash methods(e.g., Darling-
Should it be confirmed that the embryo sex ton & La Cour 1975) which stain then squash,
ratios are biased, the next question to be an- exceptthatitstainsaftersquashingsoastoyield
swered is what is the mechanism by which the much superior squashing. The stage of devel-
sex ratio bias arises. This question is ofspecial opment ofthe eggs atthe moment the sacs were
interest in spiders because, unlike haplodiploid collected was not known. However, we found
organismsthatconstitutemostotherknowncases that good preparations can be obtained from a
ofextremesexratiobiases(Hamilton 1967),spi- wide range ofstages, from very young embryos
ders are diploid organisms with chromosomal whose limb budsarejustbeginningto appearto
sex determination and therefore lack the oppor- older ones with well formed buds prior to the
tunity to bias the sex ratio by choosingwhether development ofa cuticular covering. Once the
or not to fertilize the egg. In this paper we ex- cuticlehasbeenformed,squashingisnotasgood
amine the possibility that the earliest acting andthereare fewerdividingcells. Alleggs in the
mechanism,abiasinmalemeiosisleadingtothe A. domingo and A. eximius sacs that appeared
excess production of sperm destined to yield to be developing normally were squashed (one
daughters, may occur in A. eximius. to three eggs in the first three A. eximius sacs
128 THEJOURNALOFARACHNOLOGY
on the microscope slide until at leastthree were
foundwiththesamechromosomecountofeither
1
22(themalediploidcomplement)or24(female);
ifthreenucleiwith 22 or24chromosomescould
not be found, the egg was deemed unscorable.
With few exceptions (see Table 1), around 90%
ofthepreparationsforagivensaccouldbescored.
Since it is reasonable to suppose that the eggs
scoredwerearandomsampleofthee^ssquashed
andsince,withtheonlyexceptionofA.jucundus,
the eggs squashed were all or a random sample
ofthose in a sac, the sex ratiosobtainedprovide
Figure 1.—Metaphase I nucleus inA. eximiusmale a direct estimation ofthe primary sex ratios of
meiosis.Note 10acrocentricbivalentsandtwoXchro- the species studied. Fewer than 90% percent of
mosomes. Scale bar = 5 /irm. the preparations in theA. domingo sacs 1 and 4
couldbescoredbecause,whenthesacswerefirst
opened, the eggs were still too young to yield
were lostdue to mishandling). InA. studiosus, a enough cells for scoring. After the first one or
random sample ofaround 30 eggs per sac was two eggs, these sacs were closed and the prepa-
chosen. InA.jucundus,asimilarsamplesizewas rations continued at a later date. In the case of
chosen, though with the two egg sacs unevenly A. jucundus, because of the asynchrony in the
represented due to differences in their develop- development ofthe eggs in a sac, some were too
mental stage. SamplinginA.jucunduscould not young or too old to yield good preparations.
beentirelyrandombecausetheeggsinasacwere Chromosome preparations from testes were
found to be widely asynchronous in theirdevel- obtained by using one of three techniques: (a)
opmentandtheonesthatwereobviouslytooold Feulgen, as done by Maddison (1982), (b)
to yield good preparations were not squashed. squashingafterstainingwith aceto-orcein, or(c)
Chromosomeswerecountedunder 1OOOX us- thesametechniqueasthatdescribedfortheeggs.
ing oil immersion. In the embryo preparations, OneA. domingo, nineA. eximius, twoA.jucun-
nucleiwithcountablechromosomesweresought dusand threeA. studiosusmales were examined
'4
2 3
/
Ah.
-
^
M
•
• ^
r
\1
%
• •
6
* /)>
1
Figures 2-4.—A. eximius embryo chromosomes: 2, male embryo (22 chromosomes); 3, 4, female embryos
(24 chromosomes). Scale bar = 10 moi.
AVILES & MADDISON-SOCIALSPIDERSEX RATIO 129
Figures 5, 6.—Second division nuclei of/I. eximius male meiosis: 5, two Metaphase II nuclei; top has 12
chromosomes, bottom, 10; 6, two pairsofearlyTelophaseII nuclei, tophas 12 chromosomeseach memberof
thepair, bottom, 10. Scale bar= 10 fim.
toconfirmthechromosomecomplementsofthe able nuclei found, scored. Counts were made of
four species. The nine A. eximius males were eithereachsecond metaphasenucleus(Fig. 5) or
furtherexaminedtoinvestigatethepossibilityof of each pair of early second telophase nuclei,
a bias in male meiosis leading to the overpro- which nearly always occurred together (Fig. 6).
duction of XX-bearing spermatids. A. eximius Confidencelimitsfortheproportionsatthe 95%
slides were scanned systematically and all scor- level were obtained from binomial confidence
Table 1.—Number offemale and male embryos present in individual egg sacs offour species ofthe genus
Anelosimus, as determined by their chromosome count: females, 24 chromosomes, and males, 22. The total
numberofeggsinasac,thenumbersquashedforchromosomesand,ofthose,thepercentthatyieldedpreparations
whose chromosomecountcould be scoredaregiven in columns 3-5.
Eggs
%
Species Sac no. Total Squashed scorable Females Males
A. domingo 1 16 16 81 12 1
2 13 13 92 11 1
3 14 14 93 12 1
4 15 15 73 10 1
A. eximius 1 51 43 91 35 4
2 45 44 89 37 2
3 53 50 88 39 5
4 51 51 92 42 5
A.jucundus 1 73 16 75 6 6
2 71 43 63 14 13
A. studiosus 1 39 31 94 15 14
2 47 30 83 13 12
130 THEJOURNALOFARACHNOLOGY
Table2.—PrimarysexratiooffourspeciesofthegenusAnelosimusreportedastheproportionofmaleembryos
contained in 2-4 egg sacs perspecies.
Proportion
Species No. sacs No. eggs Males Females males 95%c.i.
A. domingo 4 50 4 45 0.08 0.02-0.19
A. eximius 4 172 16 153 0.09 0.05-0.14
A.jucundus 2 39 19 20 0.49 0.33-0.66
A. studiosus 2 54 26 28 0.48 0.34-0.62
interval graphs, as presented by Remington and (ratio 0.49, A = 105 pairs, 95% confidence in-
Schork(1985). terval = 0.34-0.59), showing that male meiosis
is not biased. In the earlier stage ofmetaphase
RESULTS
II (Fig. 5) we obtained a slightly biased ratio
ChromosomecomplementofAnelosimusspp.— (ratio 0.39, N = 57, 95% confidence interval =
Chromosomecountsinmalemeiosisshowedthat 0.270.53) which however was not significantly
the male diploid complement in all four species different from 1:1 (p > 0.11, by an exact two-
ofAnelosimusexaminedis20autosomes + XXO tailed test based on the binomial probabilities).
(Fig. 1), the typical complement for the family TheslightbiasobservedinmetaphaseIIisprob-
Theridiidae(Suzuki 1954).Males,therefore,have ablyduetosamplingerrorgiventhatthenumber
22 chromosomes, and females should have 24 ofnuclei scored at this phase is smaller (57 vs.
chromosomes.Asexpected,developingeggswere 105) and that at the later telophase II stage the
found to have either 22 or 24 chromosomes in two types ofnuclei occur in even numbers.
all four species examined (Figs. 2-4). The 20
autosomes, as well as the two X chromosomes, DISCUSSION
areacrocentrics.OnemaleofA. studiosusshowed This studyshowsthatthe sexratioamongde-
an extra chromosome, possibly a supernumer- velopingembryos oftwo ofthe most social spe-
ary. cies ofthe genusAnelosimus, A. eximius andA.
Sex ratio of Anelosimusjucundus and domingo, is highly femalebiased. The bias inA.
A. studiosuswerefoundtohaveanevenprimary eximius is of the same magnitude as that pre-
sex ratio (Table 2), while the other two species, viouslyreportedfrompreadultindividualsofthis
A. domingo andA. eximius, were found to have species (Aviles 1986 and unpublished data) and
highly female biased primary sex ratios. Anelo- from individuals raised from eggs (Vollrath
simus domingo egg sacs contained a single male 1986b). This showsthatdifferential mortalityof
outof1 1 to 13eggsandA.eximiussacscontained the sexes during the juvenile instars is not re-
from 2 to 5 males out of39 to 47 eggs (Table 1). sponsible forthe sex ratio bias and that the bias
The proportion ofmales found among embryos resultsfromanoverproductionoffemalesbythe
from the foureggsacs ofA. domingois 0.08 and time the eggs are laid. The sex ratio is therefore
ofA. eximius, 0.09 (Table 2). biasedthroughoutthe lifecycle,andparental in-
AlltheeggsintheA. domingosacsandinthree vestment in A. eximius, as in A. domingo, is
oftheA. eximiussacswerefound to bedevelop- heavily skewed towards females. This removes
ing normally. One of the A. eximius sacs (#1) any doubts that this bias represents a violation
contained 1 dried up egg and 6 egg shells, most ofFisher’sprinciple,ontheonehand,andpush-
likely the remains ofeggs eaten up by two hy- esbackthemomentatwhichthesexratiobiasing
menopteran parasitic larvae found in the sac. mechanism must act to the period previous to
Is the sex ratio biased by male meiosis?.—By egg laying. In the othertwo species,A.jucundus
the second division ofmale meiosis, nuclei des- and A. studiosus, even primary sex ratios were
tinedto become male-producingsperm have 10 found.
chromosomes,thosedestinedtobecomefemale- ThedifferenceinsexratiobetweenA.jucundus
producing sperm have 12. At telophase II (Fig. and4.studiosus,ontheonehand,and4.eximius
6), the ratio of male-producing to female-pro- andA. domingo, on the other, is consistentwith
ducing nuclei was found to be very close to 1;1 whatisknownaboutthematingsystemandpop-
AVILES & MADDISON-SOCIALSPIDER SEX RATIO 131
ulationstructureofthesespecies.Anelosimusju- transfer of sperm to the female spermatheca,
cundus and A. studiosus form colonies that dis- spermactivationinthespermathecapreviousto
integratebeforeitsmembers(usuallytheprogeny the fertilization ofthe eggs, and fertilization it-
ofasinglefemale)reachadulthood(Brach 1977; self. Once the eggs have been fertilized, the sex
Nentwig & Christenson 1986; Aviles unpub- ratiomustbealreadydetermined, since,asVoll-
lished). Therefore, mating in these species takes rath (1986b) points out, reabsorption of male
place among individuals from the population at zygotesisnotlikelygiventhatthespermisadded
large and, as observed, an even sex ratio is ex- to the eggs as they are being laid. Some form of
pected. The coloniesofA. eximius, on the other sperm selection, involving either differential
hand, as a consequence ofpermanent sociality, death ofsperm, differential activation or sperm
constitute isolated lineages whose members re- competition, appears the most likely mecha-
produce by inbreeding generation after genera- nism.
tion (Overal & Ferreira 1982; Vollrath 1982; The question ofwhen the sex ratio is biased
Smith 1982; Aviles 1983, 1986). Accordingto a still needs to be resolved in othersocial spiders.
model proposed (Aviles 1983, 1986, in prep.), OutsideAnelosimus,withthealreadymentioned
the isolated descent ofmany small lineages and exceptionsofAchaearaneawauandStegodyphus
their rapid turnover rate would bring about the dumicola for which rearing experiments have
conditions under which selection at the colony beenconducted, sexratiosinothersocialspiders
level can override fisherian selection within col- have only been measured in adults (Jackson &
onies, making female-biased sex ratios evolu- Smith 1978; Riechertetal. 1986), in some com-
tionarily stable (see Frank 1987 for a different bination of adults and subadults (Pain 1964;
model). The population structure ofthe fourth Kullman et al. 1971) or in some unspecified in-
species,A. domingo, isnotyetknown. However, star, presumably mature individuals (Darchen
cooperation in this species extends through 1967; Main 1988; Jacson & Joseph 1973; Seibt
adulthood (Rypstra & Tirey 1989) and multiple & Wickler 1988 for 5. mimosarum Pavesi). As
egg-layingfemalesandspidersofallinstarsoccur already mentioned, adult sex ratios are not suf-
in the colonies (Aviles unpublished), suggesting ficient evidence that parental investment is bi-
thatsocialityispermanentandthatmatingtakes ased, and, therefore, that Fisher’s principle has
place within the parental colony, as occurs inA. been violated. When compared with data taken
eximius. The strongly female biased sex ratios at earlier stages, adult sex ratio data can never-
here reported lead us to predict that the popu- theless be useful as evidence that sex specific
lation structureofthis specieswill also be found mortalityormigrationoccurs.However,because
to be highly subdivided and that the conditions in spiders males often mature at least one molt
that favorthe evolution offemale biased sex ra- earlierthanfemales, measuringadultsexratiois
tiosinA. eximiusarealsopresentinA. domingo. much moreinvolvedthanhasbeengenerallyre-
One ofthe questions opened by the present garded. In social spiders, as in any other species
study has to do with the mechanism by which in which generations are discrete, either due to
suchalargeoverproductionoffemalesisaccom- seasonalityortorecentestablishmentofthepop-
plished. This study rules out early death ofem- ulationorcolonybyjustafewfounders(Bradoo
bryos as a possible mechanism since, with the 1972; Darchen 1978; Lubin & Robinson 1982;
only exception ofone sac, all eggs in the A. ex- Aviles 1986; Main 1988; Seibt&Wickler 1988),
imius and A. domingo sacs examined were de- the difference in the numberofmolts makes the
veloping normally and almost all were scored. proportion ofadult males to adult females de-
The biasing mechanism must therefore act dur- pendent on the point in the colony life cycle at
ing the period previous to the deposition ofthe which the sample is collected (e.g., see fig. 3 of
eggs. Our results also show that a bias in male Aviles 1986). This might explain why some au-
meiosis, the earliest actingpossible mechanism, thors have obtained some instances ofmale bi-
does not occur in A. eximius: by telophase II, ased sex ratios (e.g., Vollrath 1986a,b; Riechert
nucleidestinedtogive riseto female- and male- etal. 1986, pp. 185, 186; Main 1988, p. 66), the
producing sperm occur in equal numbers. This large variability found by most authors (Pain
leaves the following stages as possible targets 1964; Bradoo 1975; Jacson & Joseph 1973;
duringwich the biasing mechanism can act: the Riechert et al. 1986; Vollrath 1986a,b; Seibt &
finalstagesofspermatogenesis,sperminduction. Wickler 1988), andthedifferences amongthem.
132 THEJOURNALOFARACHNOLOGY
To solve this problem one might measure adult offspringand the per capita investment in them
sex ratioastheproportionofadultmalesvs. the over the period ofparental care. Ifthe sex ratio
chronologically equivalent preadult female in- values are the same, and no size difference be-
star. However, becauseofthedifferenttimesthe tween the sexes is obvious, as hasbeen foundto
sexes remain in those instars, this estimate is be the case in A. eximius then the investment
,
biased against females (female spiders are only ratio can be estimated from the numerical ratio
temporarily in the preadult instar until molting either among embryos or among young pread-
to adulthood while males ofseveral cohorts ac- ults. In studies in which the sex ratio ofa large
cumulateintheadultinstar). Ingeneral, theper- number ofcolonies needs to be estimated, mea-
sistencetimeinaparticularinstarshouldalways suring the preadult sex ratio may be the only
betaken intoconsiderationwhenanytwo stages feasiblealternative. Thepreadultsexratio, how-
inalifecyclearecomparedbyverticalsampling. ever, isprobablymoresubjecttoempiricalerror
The measurement that would more fairly com- because, through time, random mortality or
pare all males and females ofthe same cohort asynchronyinthedevelopmenttimesofthesex-
would have to count adult males vs. females of eswouldtendto increase thevarianceofthe sex
thesameandallolderinstarstowhichtheymolt ratio estimate. Aside from being perhaps more
while malesare still around. However, even this accurate, embryo sex ratios have the additional
estimate would vary depending upon when in advantage ofallowingan assessmentofwhether
the colony life cycle the sample was collected, if there is variation for the sex ratio among the
males migrate or die earlier than females. For progeny ofdifferent females (given that eggs of
these reasons, at the moment we do not have a clutch are laid together in a sac).
good evidence ofwhat the magnitude ofthe sex The importance ofknowing the primary sex
ratio bias is in other social spiders or whetherit ratio is certainly not limited to social spiders
represents a theoretically interesting bias. sinceissuesofsexratioandpopulationstructure,
Given ourcurrent knowledge about the social sexspecificdemographicphenomenaandsexra-
behaviorandpopulationstructureofothersocial tio variation are ofgeneral interest in spiderbi-
spiders,however,ourpredictionisthattheiradult ology. Thecytogenetictechniquethatwepresent
sex ratio bias will also be found to result from here greatly simplifies the estimation ofthe pri-
unevenparentalinvestment.Thispredictioncan marysexratioinspiders. Itnotonlyhasobvious
now be easily tested using the cytogenetic tech- advantages over time consuming egg-rearing
nique that we present here. It should be noted, techniques which risk producing a biased esti-
however, that in species in which parental care mateifthereismortality(Fiala 1980),butitalso
extendsbeyondconception, measuring sex ratio has several advantages over a previously de-
inembryoswillnotnecessarilytellusallwewant scribed technique for obtaining spider embryo
to know about parental investment. Whetheror chromosomes(Matsumoto 1977; Tugmonet al.
notbiasedatconception,the proportionofmale 1990):itallowsreliablepreparationofindividual
to female offspring, or their relative sizes, may eggsanditisfastenoughthatpopulationstudies
change during the period of parental care as a becomefeasible.Thistechniquehasalreadybeen
result ofdifferential mortality or differential al- successfully used in mites (M. Kaliszewski pers.
locationofresources. Todeterminewhetherthis comm.) and it can probably be used with equal
is the case, the sex ratio at the end ofthe period successinotherarthropods(seeCrozier 1968 for
ofparentalinvestmentshouldalsobeestimated. a more laborious technique used for insect pu-
Insocialspiders,parentalinvestmentcanbepre- pae).Widespreaduseofthistechniquewillmake
sumed to end at the instar at which the spider- available quantitative sex ratio estimates of a
lings start to participate in the activities ofthe phylogenetically diverse set of species, so that
colony and are therefore less dependent on the comparative studies to test specific predictions
parentalgeneration(inA. eximius, thisoccursat of sex ratio and population structure become
about the same time when males are first rec- possible. Shouldprimarysexratiobiasesbecon-
ognizable due to their enlarged palpi, Aviles firmedinthesocialspeciesforwhichbiasedadult
1986). Ifthis young preadult sex ratio is found sex ratioshave been reported, we wouldthen be
to be different from the embryo sex ratio, then facedwiththeinterestingquestionofhowspecies
parentalinvestmentwould need to beestimated in five different spider families (Agelenidae,
by integrating the numbers ofmale and female Dyctinidae, Eresidae, Theridiidae and Thomis-
AVILES & MADDISON-SOCIALSPIDER SEX RATIO 133
idae) have solved the common problem ofde- Crozier, R. H. 1968. Anaceticaciddissociation,air-
vising a mechanism by which to beat the odds dryingtechniqueforinsectchromosomes,withace-
imposedbythemeioticprocess(Williams 1979). to-lacticorceinstaining. StainTechnology,43:171-
173.
ACKNOWLEDGMENTS Darlington, C. D. & L. F. LaCour. 1975. The Han-
dlingofChromosomes, Sixth Ed., Wiley and Sons,
We are very grateful for the support and the NewYork.
laboratoryfacilitiesprovidedbytheDept, ofBi- Darchen, R. 1967. Unenouvellearaignee socialedu
ology, Pontificia Universidad Catolica del Ec- GabonAgelenarepublicanaDarchen. Biol. Gabon-
uador and the Museo Ecuatoriano de Ciencias ica, 3:31-42.
Naturales,wheremostofthechromosomeslides Darchen, R. 1978. Les essaimages de I’araignee so-
waenrdeinprtehpearleadb.orGa.torEys,teavnedzJa.ssCiasmtepdanina,thienfpiaerltd cbioanlaei,sAe.geCl.enRa. Hceobnsdo.ciSaetaancDeesniAsc,add.anScsi.laSefr.orDetSgcai-.
Nat. 287:1035-1037.
ofthechromosomescoring;wethankthemboth Fiala, K. 1980. On estimatingthe primary sex ratio
tfohretMheuisrediulmigeonfceVearntdebernattheusZioaoslmo.gJy. Pgaetnteornouasnldy Fisfherro,mRi.nAc.omp1l9e3t0e.daTthae.GAemn.etiNcata.l,Th1e1o5:r4y4o2f-^N4a4t.ural
allowed us to use theircytogenetics lab at U. C. Selection. Dover, NewYork.
Berkeley to carry out most of the scoring and Fowler,H.G.&H.W.Levi. 1979. Anewquasisocial
photographicwork. M. Slatkinprovidedusshel- Anelosimusspider(Araneae:Theridiidae)fromPar-
ter. A. Rypstra, F. Vollrath and O. Krauss re- aguay. Psyche, 86:11-17.
viewedapreviousversionofthemanuscriptand Frank, S. 1987. Demographyand sex ratio in social
gave valuable comments for its improvement. spiders. Evolution, 41:1267-1281.
ThisworkwassupportedbygrantsfromtheNa- Hamilton,W.D. 1967. Extraordinarysexratios.Sci-
ence, 156:477-488.
tainodnaHlaGrevoagrrdapUhniicveSrosciiteyty((O#E3B81g2r-a8d8u),atSeigstmuadeXni,t Herre,E.A. 1985. Sexratioadjustmentin figwasps.
Science, 228:896-898.
grants) to L. Aviles, and by an NSERC Canada Jackson, R. R. &S. E. Smith. 1978. Aggregationsof
postdoctoral fellowship to W. Maddison. Mallos and Dictyna (Araneae: Dictynidae): popu-
lation characteristics. Psyche, 85:65-80.
LITERATURECITED
Jacson, C. C. & K. J. Joseph. 1973. Life history,
Aviles, L. 1983. Proporcion de sexos en la arana bionomics and behaviour ofthe social spider Ste-
social Anelosimus eximius. Licenciatura thesis, godyphus sarasinorum. Insectes Sociaux, 20:189-
PontificiaUniversidadCatolicadelEcuador,Quito. 203.
Aviles, L. 1986. Sex ratio bias and possible group Kullman, E., St. Nawabi & W. Zimmermann. 1971.
selection in the social spider Anelosimus eximius. Neue Ergebnisse zur Brutbiologie cribellater Spin-
Am. Nat., 128:1-12. nen aus Afghanistan und der Serengeti. Zeitschr.
Brach,V. 1975. ThebiologyofthesocialspiderAne- KolnerZoo, 14:87-108.
losimuseximius(Araneae:Theridiidae).Bull. South. Leigh, E. 1970. Sex ratio and differential mortality
Calif. Acad. Sci., 74:37^1. between the sexes. Am. Nat., 104:205-206.
Brach, V. 1977. Anelosimus studiosus (Araneae: Levi, H. W. 1956. The spidergenera Neottiura and
Theridiidae) and the evolution ofquasisociality in Anelosimus in America (Araneae: Theridiidae).
Theridiid spiders. Evolution, 31:154-161. Trans. Amer. Micros. Soc., 75:407^22.
Bradoo, B. L. 1972. Some observationson theecol- Levi,H.W. 1963. TheAmericanspidersofthegenus
ogyofsocialspiderStegodyphussarasinorumKarsch Anelosimus (Araneae, Theridiidae). Trans. Amer.
(Araneae: Eresidae)from India. OrientalInsects, 6: Micros. Soc., 82:30-48.
193-204. Levi, H. W. & D. R. Smith. 1982. A new colonial
Bradoo, B. L. 1975. The sexual biology and mor- Anelosimus spiderfrom Suriname (Araneae: Ther-
phologyofthe reproductiveorgans ofStegodyphus idiidae). Psyche, 89:3-4.
sarasinorum Karsch (Araneae: Eresidae). Entomol. Lubin, Y. D. inpress. Patternsofvariation in female-
MonthlyMag., 111:239-247. biasedsexratiosin asocial spider. Biol.J. Linnean
Buskirk, R. 1981. Sociality in the Arachnida. Pp. Soc.
281-359, In Social Insects (H. R. Hermann, ed.) Lubin,Y. D. &R. H. Crozier. 1985. Electrophoretic
vol. II. Academic Press, NewYork. evidence for population differentiation in a social
Chamov, E. 1982. The Theory of Sex Allocation. spiderAchaearaneawau(Theridiidae). InsectesSo-
Princeton UniversityPress, Princeton. ciaux, 32:297-304.
Christenson,T. 1984. Behaviourofcolonialandsol- Lubin,Y. D. &M. H. Robinson. 1982. Dispersal by
itaryspidersofthetheridiidspeciesAnelosimusex- swarmingina social spider. Science, 216:319-321.
imius. Anim. Behav., 32:725-734. Main, B. Y. 1988. The biology ofa social thomisid
134 THE JOURNALOFARACHNOLOGY
spider. Australian Arachnology [Austral. Ent. Soc. Vollrath, F. 1986a. Environment, reproduction and
Misc. PubL, Brisbane], 5:55-73. thesexratioofthesocialspiderAnelosimuseximius
Maddison, W. 1982. XXXY sex chromosomes in (Araneae, Theridiidae). J. Arachnol., 14:266-281.
males ofthe jumping spider genus Pellenes (Ara- Vollrath,F. 1986b. Eusocialityandextraordinarysex
neae: Salticidae). Chromosoma (Berl.), 85:23-37. ratios in the spiderAnelosimus eximius (Araneae:
Matsumoto, S. 1977. An observation of somatic Theridiidae). Behav. Ecol. SociobioL, 18:283-287.
chromosomes from spider embryo-cells. Acta Ar- Waage,J.K. 1982. Sib-matingandsexratiostrategies
achnol., 27:167-172. in scelionidwasps. Ecol. Entomol., 7:103-112.
Nentwig,W.&T.Christenson. 1986. Naturalhistory Werren, J. H. 1980. Sex ratio adaptations to local
ofthenon-solitarysheetweavingspidzrAnelosimus matecompetitioninaparasiticwasp. Science, 208:
jucundus (Araneae: Theridiidae). Zool. J. Linnean 1157-1159.
Soc., 87:27-35. White, M. J. 1973. Animal cytology and evolution,
Overal, W. L. & P. R. Ferreirada Silva. 1982. Pop- 3rd. ed. Cambridge UniversityPress, Cambridge.
ulation dynamics ofthe quasisocial spider Anelo- Williams, G. C. 1979. The question ofadaptive sex
simuseximius(Araneae:Theridiidae).Pp. 181-182, ratioinoutcrossedvertebrates. Proc. R. Soc. Lond.
InThebiologyofSocialInsects(M. E. Breed,C. D. B, 205:567-580.
Michener&H.E.Evans,eds).Proc.oftheIXCongr. Wilson, D. S. & R. K. Colwell. 1981. Evolution of
Int. Union forthe StudyofSocial Insects, Boulder, sex ratio in structured demes. Evolution, 35:882-
Colorado. 897.
Pain,J. 1964. Premieresobservationssuruneespece
nouvelle d’araignees sociales, Agelena consociata. Manuscript received August 1990, revised February
Biol. Gabonica, 1:47-58. 1991.
Remington, R. D. & M. A. Schork. 1985. Statistics
withApplicationsto the Biologicaland HealthSci-
ences. Prentice Hall, NewJersey.
Riechert, S., R. Roeloffs & A. C. Echtemacht. 1986.
TheecologyofthecooperativespiderAgelenacon-
sociata in Equatorial Africa (Araneae, Agelenidae). APPENDIX
J. Arachnol., 14:175-191. PROTOCOLTO OBTAIN SPIDER EMBRYO
Rypstra, A. L. & R. S. Tirey. 1989. Observationson CHROMOSOME PREPARATIONS
the social spider, Anelosimus domingo (Araneae,
Theridiidae),insouthwesternPeru.J.Arachnol., 17: 1. Fixtheegg. Withaveryfineneedle,pokeasmall
368. hole in the eggand with the egg so skewered, place it
Seibt, U. &W. Wickler. 1988. Bionomicsandsocial in a drop offixative (3 parts absolute ethanol:1 part
structure of“Family Spiders” ofthe genus Stego- glacial acetic acid). We used electrochemically-sharp-
dyphus,withspecialreferencetotheAfricanspecies enedtungstenwireneedles. Limbbuds, ifpresent, ap-
S. dumicola and S', mimosarum (Araneidae, Eresi- pear as a series ofsmall white lumps as the fixative
dae). Verb, naturwiss. Ver. Hamburg, 30:255-303. enterstheegg(ablackbackgroundenhancesvisibility).
Smith, D. R. 1982. Population genetics oiAnelosi- Remove the tip ofthe needle from the egg, and press
mus eximius (Araneae, Theridiidae). J. Arachnol., onthesideoftheeggsoastoforcethecontentsthrough
14:201-217. thesmallhole. Iftheeggisyoungenoughthecontents
Suzuki, S. 1954. Cytological studies in spiders. III. can be squirted into a long thin string, which aids in
Studies on the chromosomes offifty-seven species rapidfixingandlaterinbreakingthecontentsinsmall
ofspidersbelongingtoseventeenfamilies,withgen- pieces.Tissuewithnucleiiswhite;yolkwithoutnuclei
eral considerations on chromosomal evolution. J. is yellowish; ifthere is much yolk then some can be
Sci. Hiroshima Univ., ser. B, 15:23-136. discarded. Discardtheemptychorion. Fix for 30 sec-
Tapia, Y. & T. De Vries. 1980. Tolerancia y coop- onds.
eracion en la arana socialAnelosimusjucundusdel 2. Squashthetissue. Place the fixedcontentsofthe
bosque tropical Rio Palenque, Ecuador. Rev. de la egginasmalldropof60%aceticacidonamicroscope
Univ. Catolica Quito, VII(27):51-74. slide. Withtwovery fineneedles,breakthetissueinto
Trivers, R. 1985. Social Evolution. The Benjamin/ smallpieces. Placeacoverslipontop,andsquashthe
Cummings PublishingCompany, Inc. tissue flat. This squashingis perhaps the mostcritical
Tugmon, C., J. D. Brown & N. V. Homer. 1990. stepin theprocedure: squashingtoosoftly, slidingthe
Karyotypes ofseventeen USA spiderspecies (Ara- cover slip sideways while squashing, and air bubbles
neae, Araneidae, Gnaphosidae, Loxoscelidae, Ly- shouldall be avoided.
cosidae, Oxyopidae, Philodromidae, Salticidaeand 3.Removethecoverslipandletdrythetissue.Freeze
Theridiidae). J. Arachnol., 18:41-48. the slide on dry ice at least several minutes and flip
Vollrath,F. 1982. Colonyfoundationinasocialspi- thecoverslipoffwitharazorblade.Washofftheacetic
der. Z. Tierpsychol., 60:313-324. acid for 20 seconds in a bath ofabsoluteethanol. Let
AVILES& MADDISON-SOCIALSPIDER SEX RATIO 135
the slide dry at least ten minutes. Ifneeded, the slide could be used; we have used primarily a 3-4 minute
can beleftin this condition overnight orlonger. bath ofacetocarmine. After staining, the slide can be
4. Stain the tissne and make permanent the prepa- rinsedwithappropriate solvents to prepare it forper-
ration. Aftertheslidehasdriedwell, itcan bestained manentmounting.
andmadepermanent. Nodoubtmanydifferentstains