Arm156 1.6

Behavioral Ecology Advance Access published April 11, 2008
Female praying mantids use sexual cannibalismas a foraging strategy to increase fecundity Katherine L. Barry, Gregory I. Holwell, and Marie E. HerbersteinDepartment of Biological Sciences, Macquarie University, NSW 2109, Australia Several hypotheses have been proposed to explain the evolution of sexual cannibalism. The foraging strategy hypothesis statesthat sexual cannibalism may arise as an adaptive foraging strategy, providing females with the nutrients to increase futurefecundity. Yet, very few studies have found that nourishment through cannibalism translates into increased fecundity. Oneexplanation for this may be that most of these investigations have concentrated on sexually dimorphic spider species with tinymales that do not significantly increase female body mass on consumption. The current study focuses on the praying mantid,Pseudomantis albofimbriata, a moderately size dimorphic species with relatively large males. Cannibalistic females of such speciesmay be more likely to gain nutritional benefits from male consumption, which translate into increased fecundity. Here, canni-balistic females substantially improved their body condition and subsequently produced heavier egg cases than their noncanni-balistic counterparts. An additional prediction of the foraging strategy hypothesis is that sexual cannibalism will increase withdecreasing female condition. We found that the prevalence of sexual cannibalism in this system was indeed affected by femalebody condition; females in poor condition were more likely to consume their potential mates than females in good condition.
Additional analysis of the data refuted the relevance of each of the remaining hypotheses for this species, providing clearevidence for the foraging strategy hypothesis as an explanation for the maintenance of sexual cannibalism in this species.
Key words: fecundity benefit, female body condition, foraging strategy, praying mantid, Pseudomantis albofimbriata, sexual cannibalism, sexual conflict. [Behav Ecol] Sexual cannibalism, where females consume courting males a smaller size and body mass (Schneider and Elgar 2002).
before, during, or immediately after mating, may represent And finally, precopulatory sexual cannibalism may arise the ultimate conflict of interest between the sexes (Darwin strictly as an adaptive foraging strategy, providing females with 1871; Elgar 1992). Although cannibalism during courtship/ nutrients that increase survival and/or fecundity (Newman mating occurs with relatively low frequency in the majority of animal taxa, it is comparatively widespread among the arthro- The foraging strategy hypothesis was first proposed by pods (Elgar 1992; Maxwell 1999a). Several hypotheses have Newman and Elgar (1991) to explain precopulatory cannibal- been proposed to explain the origin and maintenance of sex- ism by virgin females and suggests that females adaptively ual cannibalism in various insect and arachnid groups. The assess a male’s value as a mate versus his value as a meal.
costs and benefits of sexual cannibalism are typically viewed Therefore, the model predicts that sexually cannibalistic be- from the male or female perspective, and the timing of can- havior will increase with decreasing food availability and body nibalism in relation to insemination is also considered (Elgar condition; will be more common in mated as opposed to and Schneider 2004). Cannibalism that occurs after insemina- virgin females; will increase as male size increases; and will tion may be advantageous for males and/or females. The decrease as male density decreases. Numerous studies lend degree of conflict is minimized if a male’s cannibalized soma support to one or more of these predictions (Liske and Davis increases female fecundity, and the paternal investment com- 1987; Birkhead et al. 1988; Andrade 1996; Elgar and Fahey pensates for any future loss of reproductive opportunity. In 1996; Maxwell 2000; Schneider and Elgar 2001; Herberstein contrast, precopulatory cannibalism is the most extreme form et al. 2002; Johnson 2005a). However, contrasting evidence of sexual conflict; although the female may benefit from the suggests the frequency of sexual cannibalism is not affected encounter, the male forfeits all present and future reproduc- by female feeding condition (Herberstein et al. 2002; Johnson tive successes (Buskirk et al. 1984).
There are 4 hypotheses that have been proposed to explain (Schneider et al. 2000; Schneider and Elgar 2001), or male size the evolution of precopulatory sexual cannibalism. First, it is an (Fromhage et al. 2003; Johnson 2005b; Schneider et al. 2006).
extreme form of female mate choice, where males of preferred The foraging strategy hypothesis also assumes that increased phenotypes are allowed to copulate and sire offspring but in- feeding and nutrition translates into increased female fecun- ferior males are cannibalized before fertilization (Elgar and dity. To date, only 2 studies have found such a benefit from Nash 1988). Second, it occurs as a result of strong selection male consumption. Sexual cannibalism increased fecundity for female aggression in earlier life-history stages, even in female mantids, Hierodula membranacea (Birkhead et al.
though there may be costs at the adult stage (Arnqvist and 1988), and increased egg sac mass in female fishing spiders, Henriksson 1997). Third, sexual cannibalism is the side effect Dolomedes triton (Johnson 2005a). These investigations were of an increased foraging vigor of females that mature at restricted to species with relatively large, nutritious males. Incontrast, the studies that found no significant increase in femalereproductive output as a result of cannibalism focus on spider Address correspondence to K.L. Barry. E-mail: kbarry@bio.mq.edu.
species that are highly sexually size dimorphic—a typical male Received 25 August 2006; revised 28 June 2007; accepted 9 represents less than 10% of a female’s body mass (Table 1).
It is, therefore, not surprising that the male soma does notsubstantially contribute to a female’s diet in these cases.
Ó The Author 2008. Published by Oxford University Press on behalf ofthe International Society for Behavioral Ecology. All rights reserved.
For permissions, please e-mail: journals.permissions@oxfordjournals.org Table 1Studies that examined whether sexual cannibalism affords a female fecundity benefit Male approximately 27% of femalebody mass (see Johnson 2005a) Male approximately 40% of femalebody mass a Personal communication with the author/s. NS, nonsignificant.
In the present study, we used the sexually cannibalistic false and after each individual’s mating trial. Because pronotum garden mantid, Pseudomantis albofimbriata, to test the assump- length remains constant throughout adulthood and can only tion and first prediction of the foraging strategy hypothesis.
be influenced by feeding during the juvenile stages (Barry KL, Pseudomantis albofimbriata exhibits only moderate sexual size in preparation), it is used as a measure of ‘‘fixed size.’’ Body dimorphism—males are approximately 40% the body mass mass is, instead, variable and depends on size and feeding lev- of females—and we predict that cannibalistic females of such els during adulthood. Male P. albofimbriata represent approx- species will increase body condition and fecundity as a result imately 40% the body mass of females (female body of male consumption. We also predict that the frequency of mass = 0.665 6 0.030 g, N = 38; male body mass = 0.265 6 sexually cannibalistic attacks will increase with decreasing 0.006 g, N = 38). Finally, we used each female’s mass divided food availability/female condition. Because it is possible that by her fixed size as an index of body condition (see Jakob the 4 hypotheses proposed to explain the evolution of sexual et al. 1996 for discussion of other indices).
cannibalism are not mutually exclusive, we carried out furtheranalyses to test the relevance of each remaining hypothesis forthis species.
Female foraging strategy hypothesis: fecundity benefit A virgin male and female were randomly chosen from the lab- oratory population the day after feeding had occurred, so thatfemales were neither hungry nor satiated during mating experiments. On the day of trial, males were 24.95 6 1.52 days Individual P. albofimbriata were collected from various sites postadult emergence and females were 24.25 6 1.38 days post- around Sydney and Canberra, Australia, during the summer adult emergence. Because the majority of mantids were col- months of 2003/2004 and 2005/2006 (December–February).
lected from Lomandra spp. bushes in the field, we placed The majority of individuals were found in Lomandra spp.
each experimental pair onto 1 of 5 potted Lomandra plants, bushes at Kuringai Bicentennial Park, West Pymble, and on placed outdoors so as to simulate natural conditions. Approx- the Australian National University campus, Canberra. Juvenile imately 10 min after the initial introduction of a female, a male mantids were collected from the study sites and reared to was placed onto the plant at least 20 cm behind the female.
adulthood on a diet of 2 small crickets (Acheta domestica) 3 If an interaction did not occur within 3 h, the trial was termi- times a week and water daily. Animals were housed individu- nated. If copulation or cannibalism occurred, measures of ally within well-ventilated 425-ml transparent cups in the female body mass and body condition were again obtained laboratory, at a temperature of 24–26 °C. Experiments were on completion of the experiment. Males that failed to mate carried out during daylight hours (typically 8 AM–8 PM) in were given at least 5 days before being paired with another February–April 2004 and February–April 2006.
Although the onset of cannibalism occurs prior to copula- tion in this species, males are able to initiate copulation and transfer sperm while being consumed (Barry 2004). There- The pronotum length of mantids was recorded after the final fore, we compared females that copulated (no cannibalism) molt, whereas body mass was measured immediately preceding with females that cannibalized and copulated to determine Barry, et al. • Sexual cannibalism in a praying mantid if sexual cannibalism afforded an immediate fecundity bene-fit. Female fecundity was estimated using female condition(after trial) and the mass of the first ootheca (egg case). Fe-male body condition is positively related to fecundity in manyspiders and mantids (Rubenstein 1987; Birkhead et al. 1988;Wise 2006), and preliminary analyses of P. albofimbriata haveshown that the mass of an ootheca is positively related to thenumber of offspring that hatch from that ootheca (Pearson’sR = 0.601, N = 22, P = 0.003). Females were starved after cop-ulation until ootheca production (time to first oviposition =9.890 6 0.900 days, N = 27), so that any change in fecunditywas due solely to consumption of a conspecific and not sub-sequent consumption of another prey item. Oothecae wereremoved from cups 5 days after deposition to allow completehardening and easy measurement using electronic scales.
Female foraging strategy hypothesis: female condition In order to determine if the frequency of cannibalism in- creased with decreasing female condition, we first compared Female body condition before and after mating trials, compared the body condition of cannibalistic (N = 16) and noncanni- between cannibals (N = 8) and noncannibals (N = 19). Cannibalistic balistic females (N = 19) at the commencement of trial.
females were in significantly poorer condition at the start of trial, but Our next experiment involved manipulating the feeding re- this difference disappeared when condition was compared after trial.
gime of a different group of virgin females (N = 19) immedi-ately before a second round of mating trials. These femaleswere starved for 10 days so as to empty their gut (see Maxwell2000) and then randomly allocated to weeklong feeding treat- ments. Females on the ‘‘low-quantity’’ feeding treatment Female foraging strategy hypothesis: fecundity benefit (N = 9) were given 1 small cricket (body mass = 0.037 60.003 g, N = 27) 3 times during the week preceding experi- There was a significant difference in condition change (before mentation, whereas females on the ‘‘high-quantity’’ treatment and after trial) between the cannibalistic and noncannibalistic (N = 10) were fed 1 large cricket (body mass = 0.300 6 0.023 g, females (t-test: t16.924 = 11.883, P , 0.001; Figure 1); all fe- N = 30) 3 times over the same week. All animals received males that consumed male conspecifics substantially im- water daily. Females were weighed prior to, and immediately proved their condition, whereas all noncannibalistic females after, feeding regimes, and high-quantity females were in sig- saw a decrease in body condition. Furthermore, female body nificantly better condition (0.051 6 0.005 g) after feeding condition significantly predicted ootheca mass (linear regres- treatments than low-quantity females (0.035 6 0.004 g; t-test: sion: r2 = 0.437, F1,26 = 4.405, P , 0.001; Figure 2), and can- t17 = 22.593, P = 0.019). The treatment groups were, there- nibalistic females tended to be in better condition after trial fore, renamed ‘‘good condition’’ and ‘‘poor condition.’’ Mat- than their noncannibalistic counterparts (t-test: t36 = 1.418, ings were then carried out as per the initial unmanipulated trials, and the frequency of sexual cannibalism was comparedbetween treatments.
Although the primary intention of this study was to determineif the foraging strategy hypothesis explains sexual cannibalismin this system, we also wanted to test the major assumptionsassociated with each of the remaining 3 hypotheses. First,we compared various phenotypic characters of males thatapproached cannibalistic and noncannibalistic females (matechoice hypothesis: Elgar and Nash 1988). Second, we com-pared the fixed adult size of cannibals and noncannibals todetermine whether females that consumed their mates weresmaller at maturity (life-history strategy hypothesis: Schneiderand Elgar 2002) or larger at maturity (aggressive spilloverhypothesis: Arnqvist and Henriksson 1997).
Data were analyzed using SPSS 11.0 for Windows and werechecked for normal distribution (Kolmogorov–Smirnov test)before further statistical analysis. The data collected during2 different mating seasons were subsequently pooled because there was no significant difference between any of the measure- Female body condition had a significant positive effect on ootheca ments used in analyses. Unless otherwise stated, all values are mass in Pseudomantis albofimbriata (regression equation = 2 0.06 1 mean 6 standard error, and all statistical tests are 2 tailed.
4.51 3 female body condition after trial).
Foraging strategy hypothesis: fecundity benefit As predicted, sexual cannibalism had a significant positive ef-fect on both female body condition and the mass of the firstootheca. Furthermore, a female’s body condition after trialwas a significant predictor of her subsequent reproductiveoutput, suggesting that females gain an immediate fecunditybenefit as a direct result of male consumption. Pseudomantisalbofimbriata males represent a relatively large proportion ofa conspecific female’s body mass, so it is not surprising thatsexual cannibalism can boost the reproductive output of afemale by up to 40%. A similar increase in body conditionand/or fecundity has been shown for other cannibalistic spe-cies with relatively large males (see Table 1). For example,male garden spiders, Araneus diadetamus, represent approxi-mately 25% a female’s body mass, and cannibalistic femalesincrease their body mass as a result of consuming a male(Elgar and Nash 1988). Birkhead et al. (1988) found that Mass of the first ootheca compared between cannibalistic (N = 8) a greater maximum weight and a significantly greater ootheca and noncannibalistic (N = 19) females. Sexual cannibalism had weight than females that were prevented from eating a male.
a significant effect on mean ootheca mass, so that the oothecae of There are, however, species that exhibit a similar degree of cannibalistic females were heavier than those of noncannibalistic moderate sexual size dimorphism but have shown little bene- fit to female fecundity as a result of sexual cannibalism(Spence et al. 1996; Maxwell 2000). For example, Maxwell(2000) showed that female Iris oratoria maintained on Most importantly, sexual cannibalism had a significant pos- a high-quantity feeding regime were more fecund; however, itive effect on the mass of oothecae (analysis of covariance: a single cannibalistic event did not have the same significant F1,26 = 6.631, P = 0.017; Figure 3), as did the covariates of impact. It may be that the rate of resource utilization is dif- female condition before the mating trial (F1,26 = 19.934, P , ferent for this species, so that male consumption has a delayed 0.001) and copulation duration (F1,26 = 7.639, P = 0.011; effect on reproductive output. Perhaps, eggs contained in the whole-model test: r2 = 0.540, F3,26 = 9.000, P , 0.001). Using first egg case are manufactured during the weeks prior to the regression equation given in Figure 2, the difference in copulation, and those derived from the male soma are laid ootheca mass between the groups indicates that sexual canni- in subsequent egg cases (see Johnson 2001, e.g., in fishing balism could increase female reproductive output by up to 40%. Interestingly, sexual cannibalism no longer influences Although the fecundity of P. albofimbriata females is im- the mass of oothecae if female body condition after trial re- proved as a direct result of sexual cannibalism, the possibility places body condition before trial as a covariate (body condi- of remaining unmated prevails. It, therefore, seems extraor- tion before trial: F1,26 = 28.247, P , 0.001; sexual cannibalism: dinary that a behavior with such detrimental possibilities for F1,26 = 0.25, P = 0.877; whole-model test: r2 = 0.615, F3,26 = both males and females has not been selected against. One explanation for the maintenance of sexual cannibalism is theevolution of alternative male strategies that substantially re-duce the risk of attack; the safety precautions undertaken by Female foraging strategy hypothesis: female condition most male mantids, such as mounting females from low-risk As predicted, females that cannibalized males during unma- positions (Maxwell 1998, 1999b, 2000; Barry et al. forthcom- nipulated mating trials were in significantly poorer condition ing) and maintaining a copulatory posture out of female at the start of trial than their noncannibalistic counterparts reach (Roeder 1935; Robinson MH and Robinson B 1979; (t-test: t36 = 22.090, P = 0.044; Figure 1). Furthermore, there Liske and Davis 1984, 1987; Maxwell 1998, 1999b; Barry was a significant difference in the frequency of sexual canni- et al. forthcoming), are evidence of such. The ability to in- balism between feeding treatments during the second exper- seminate a female even when cannibalism precedes genital iment (Fisher’s Exact test: N = 19, P = 0.0001): 89% (8/9) of contact may be an additional method of combating highly females in the poor-condition treatment attacked and subse- aggressive females—all P. albofimbriata males that are cannibal- quently cannibalized the male; however, none (0/10) of the ized attempt to copulate with their female attacker, and ap- females in good condition attempted to strike at a male con- proximately half are successful (Barry et al. forthcoming).
Foraging strategy hypothesis: female body condition In addition to increased fecundity, the foraging strategy hy- There was no significant difference in the characteristics of pothesis suggests that the likelihood of sexual cannibalism will males that were cannibalized and males that were not cannibal- be influenced by female body condition. As predicted, canni- ized (latency to approach, P = 0.082; male mass, P = 0.953; balistic females were in significantly poorer condition at the fixed male size, P = 0.546; male condition, P = 0.706; relative commencement of unmanipulated mating trials than noncan- mass, P = 0.184; relative size, P = 0.067; relative condition, nibalistic females but had a marginally better body condition P = 0.273). Furthermore, there was no significant difference after the completion of trial (Figure 1). Pseudomantis females in fixed adult size between cannibalistic (15.720 6 0.233 mm) in poor condition seem to use sexual cannibalism as a means of improving their condition, which positively influences their subsequent reproductive output. Furthermore, all females in Barry, et al. • Sexual cannibalism in a praying mantid the good-condition treatment copulated without consuming their mate, whereas all but one poor-condition female at- Macquarie University; Linnean Society of New South Wales.
tacked and subsequently cannibalized the male. Numerousstudies have similarly shown that sexually cannibalistic behav-ior increases with decreasing food availability and/or female Many thanks to Melissa Turoczy, Jody Tropiano, and Richard Peters body condition (Liske and Davis 1984; Birkhead et al. 1988; for their help with specimen collection and to Roberto Reyes and Kynaston et al. 1994; Andrade 1998; Maxwell 1998; Maxwell Rachael Dryden for help with experiments. We thank Mike Maxwellfor his helpful comments on the manuscript.
2000; Schneider and Elgar 2001). Again, the majority of thesestudies involve cannibalistic species with moderate sexual sizedimorphism.
Although sexual cannibalism occurs frequently in many lab- oratory studies, it may be relatively uncommon in natural populations. If alternative prey types are readily available, wild Andrade MCB. 1996. Sexual selection for male sacrifice in the females may never be food limited and cannibalism might be Australian redback spider. Science. 271:70–72.
infrequent. Further studies outlining female body size/condi- Andrade MCB. 1998. Female hunger can explain variation in canni- tion, the availability of prey, and the frequency of sexual balistic behavior despite male sacrifice in redback spiders. Behav cannibalism in natural populations are required before the evolutionary significance of this behavior can be totally Arnqvist G, Henriksson S. 1997. Sexual cannibalism in the fishing spider and a model for the evolution of sexual cannibalism based on genetic constraints. Evol Ecol. 11:255–273.
Barry KL. 2004. Mating behavior of a sexually cannibalistic praying man- tid, Pseudomantis albofimbriata [honours thesis]. Sydney (Australia):Macquarie University.
Because the hypotheses proposed to explain precopulatory Barry KL, Holwell GI, Herberstein ME. Forthcoming. Male mating sexual cannibalism are not necessarily mutually exclusive, we behaviour and sexual cannibalism in a praying mantid.
also tested the major assumptions associated with the mate Birkhead TR, Lee KE, Young P. 1988. Sexual cannibalism in the choice, aggressive spillover, and life-history strategy hypothe- praying mantis Hierodula membranacea. Behavior. 106:112–118.
ses. The mate choice hypothesis suggests that sexual cannibal- Buskirk RE, Frohlich C, Ross KG. 1984. The natural selection of sexual ism is an extreme form of female mate choice, where males of Darwin C. 1871. Sexual selection and the descent of man. London: preferred phenotypes are allowed to copulate and sire off- spring but inferior males are cannibalized before fertilization Elgar MA. 1992. Sexual cannibalism in spiders and other inverte- (Elgar and Nash 1988). However, we found no evidence of brates. In: Elgar MA, Crespi BJ, editors. Cannibalism: ecology and morphological or behavioral differences in males that were evolution among diverse taxa. Oxford: Oxford Science Publica- cannibalized versus males that were not cannibalized. The aggressive spillover hypothesis suggests that sexual cannibal- Elgar MA, Fahey BF. 1996. Sexual cannibalism, competition, and size ism occurs as a result of strong selection for female aggres- dimorphism in the orb-weaving spider Nephila plumipes Latreille siveness in earlier life-history stages; therefore cannibalistic (Araneae: Araneoidea). Behav Ecol. 7:195–198.
Elgar MA, Nash DR. 1988. Sexual cannibalism in the garden spider females tend to be larger at maturity than females that do Araneus diadematus. Anim Behav. 36:1511–1517.
not consume their mates (Arnqvist and Henriksson 1997).
Elgar MA, Schneider JM. 2004. Evolutionary significance of sexual Finally, the life-history strategy hypothesis suggests that sexual cannibalism. Adv Study Behav. 34:135–163 cannibalism is the side effect of an increased foraging vigor of Elgar MA, Schneider JM, Herberstein ME. 2000. Female control of females that mature at a smaller size and body mass paternity in the sexually cannibalistic spider Argiope keyserlingi.
(Schneider and Elgar 2002). However, our study suggests that Proc R Soc Lond B Biol Sci. 267:2439–2443.
there is no significant difference in the fixed size of cannibal- Fahey BF, Elgar MA. 1997. Sexual cohabitation as mate-guarding in istic and noncannibalistic P. albofimbriata females, making the leaf-curling spider Phonognatha graffei Keyserling. Behav Ecol neither of the latter hypotheses relevant as an explanation Fromhage L, Uhl G, Schneider JM. 2003. Fitness consequences of for sexual cannibalism in this system.
sexual cannibalism in female Argiope bruennichi. Behav Ecol Socio-biol. 55:60–64.
Herberstein ME, Schneider JM, Elgar MA. 2002. Costs of courtship and mating in a sexually cannibalistic orb-web spider: female Our results support the foraging strategy hypothesis as a pri- mating strategies and their consequences for males. Behav Ecol mary explanation for the maintenance of sexual cannibalism in this system. We found that sexual cannibalism had a signif- Jakob EM, Marshall SD, Uetz GW. 1996. Estimating fitness: a compar- icant positive effect on both female body condition and oo- ison of body condition indices. Oikos. 77:61–67.
theca mass and that female body condition was a significant Johnson JC. 2001. Sexual cannibalism in fishing spiders (Dolomedes triton): an evaluation of two explanations for female aggression predictor of ootheca mass. Finally, we found that females in towards potential mates. Anim Behav. 61:905–914.
poor condition were significantly more likely to attack and Johnson JC. 2005a. Cohabitation of juvenile females with mature cannibalize a male than females in good condition.
males promotes sexual cannibalism in fishing spiders. Behav Ecol. 16: In conclusion, our results confirm the lack of a general rationalization for the evolution and/or maintenance of sexual Johnson JC. 2005b. The role of body size in mating interactions of the cannibalism. Many factors, such as the timing of cannibalism in sexually cannibalistic fishing spider Dolomedes triton. Ethology.
relation to insemination and the degree of sexual size dimor- phism, vary between and within the mantids and arachnids, Kynaston SE, McErlain-Ward P, Mill PJ. 1994. Courtship, mating be- suggesting that sexual cannibalism has evolved multiple times havior and sexual cannibalism in the praying mantis, Sphodromantislineola. Anim Behav. 47:739–741.
and is maintained in numerous species for many different rea- Liske E, Davis WJ. 1984. Sexual behavior of the Chinese praying man- sons. The female foraging strategy hypothesis is more likely to explain sexual cannibalism in moderately sexually size dimor- Liske E, Davis WJ. 1987. Courtship and mating behavior of the phic species (i.e., most praying mantids), as the relatively large Chinese praying mantis, Tenodera aridifolia sinensis. Anim Behav. 35: males make for bigger and more nutritious meals.
Maxwell MR. 1998. Lifetime mating opportunities and male mating Rubenstein DI. 1987. Alternative reproductive tactics in the spider behavior in sexually cannibalistic praying mantids. Anim Behav. 55: Meta segmentata. Behav Ecol Sociobiol. 20:229–237.
Schneider JM, Elgar MA. 2001. Sexual cannibalism and sperm Maxwell MR. 1999a. Mating behavior. In: Prete FR, editor.
competition in the golden orb-web spider Nephila plumipes Praying mantids. Baltimore (MD): Johns Hopkins University Press.
(Araneoidea): female and male perspectives. Behav Ecol. 12: Maxwell MR. 1999b. The risk of cannibalism and male mating behav- Schneider JM, Elgar MA. 2002. Sexual cannibalism in Nephila plumipes ior in the Mediterranean praying mantid, Iris oratoria. Behaviour.
as a consequence of female life history strategies. J Evol Biol. 15: Maxwell MR. 2000. Does a single meal affect female reproductive out- Schneider JM, Gilberg S, Fromhage L, Uhl G. 2006. Sexual conflict put in the sexually cannibalistic praying mantid Iris oratoria? Ecol over copulation duration in a cannibalistic spider. Anim Behav.
Newman JA, Elgar MA. 1991. Sexual cannibalism in orb-weaving spi- Schneider JM, Herberstein ME, de Crespigny FC, Ramamurthy S, ders: an economic model. Am Nat. 138:1372–1395.
Elgar MA. 2000. Sperm competition and small size advantage for Prokop P, Vaclav R. 2005. Males respond to risk of sperm competi- males of the golden orb-web spider Nephila edulis. J Evol Biol. 13: tion in the sexually cannibalistic praying mantis, Mantis religiosa.
Spence JR, Zimmerman M, Wojcicki JP. 1996. Effects of food limita- Robinson MH, Robinson B. 1979. By dawn’s early light: matutinal tion and sexual cannibalism on reproductive output of the nursery mating and sex attractants in a neotropical mantid. Science. 205: web spider Dolomedes triton (Araneae: Pisauridae). Oikos. 75:373–382.
Wise DH. 2006. Cannibalism, food limitation, intraspecific competi- Roeder KD. 1935. An experimental analysis of the sexual behaviour of tion and the regulation of spider populations. Ann Rev Entomol. 51: the praying mantis (Mantis religiosa). Biol Bull. 69:203–220.

Source: http://mantodea.myspecies.info/sites/mantodea.myspecies.info/files/Barry%20et%20al,%20Behav%20Ecol%202008.pdf