988 Polyandry and tibial spur chewing in the Carolina ground cricket ( carolinus)

Edyta K. Piascik, Kevin A. Judge, and Darryl T. Gwynne

Abstract: During mating, the female Carolina ground cricket ( (Scudder, 1877)) chews specialized spurs on the male’s hind tibia for access to his hemolymph. One potential benefit to spur chewing includes nutritional ac- quisition from male hemolymph. A method for testing this hypothesis is to manipulate food quality or quantity, with the prediction that mating rate will increase as food quality or quantity decreases. We manipulated diet quality in adult fe- males and provided them with four consecutive mating opportunities. We measured four aspects of mating behaviour (mating rate, latency to copulate, copulation duration, and spur chewing duration) and three of female fitness (egg number, egg-laying rate, and life span). Females of the two diet treatments did not differ significantly in any of the measured mat- ing behaviours, although females fed a low-quality diet lived longer. Male life span did not correlate with any measured variable, although males that experienced more matings and longer total times of copulation and spur chewing lost more mass. These results suggest that spur chewing may be costly for males, although we detected no evidence that this behav- iour was a benefit to the female or represented a form of male coercion. Re´sume´ : Durant l’accouplement, la femelle de la ne´mobie de Caroline (Eunemobius carolinus (Scudder, 1877)) mordille des e´perons spe´cialise´s sur les tibias poste´rieurs du maˆle pour avoir acce`sa` son he´molymphe. Un be´ne´fice potentiel de ce mordillement des e´perons est l’acquisition de nutriments de l’he´molymphe du maˆle. Un moyen de tester cette hypothe`se est de manipuler la qualite´ ou la quantite´ de la nourriture en pre´disant que le taux d’accouplement va augmenter avec la diminution de la qualite´ ou de la quantite´ de la nourriture. Nous avons manipule´ la qualite´ du re´gime alimentaire de femel- les adultes et leur avons fourni quatre occasions successives de s’accoupler. Nous avons mesure´ quatre variables du com- portement d’accouplement (taux d’accouplement, pe´riode de latence entre les accouplements, dure´e des copulations et dure´e du mordillement des e´perons) et trois de la fitness des femelles (nombre d’œufs, taux de ponte et longe´vite´). Les fe- melles des deux re´gimes alimentaires n’ont montre´ aucune diffe´rence significative dans les comportements reproducteurs observe´s, bien que les femelles nourries du re´gime infe´rieur aient ve´cu plus longtemps. Il n’y a aucune corre´lation entre la dure´e de vie des maˆles et les variables mesure´es, bien que les maˆles qui ont connu un nombre plus e´leve´ de copulations, des accouplements plus longs et des mordillements prolonge´s de leurs e´perons aient perdu plus de masse. Nos re´sultats in- diquent que le mordillement des e´perons entraıˆne un couˆt chez les maˆles, bien que nous n’ayons obtenu aucun indice qui laisse croire que ce comportement soit be´ne´fique aux femelles ou repre´sente une forme de coercition par le maˆle. [Traduit par la Re´daction]

Introduction have conflicting interests in the outcome of their reproduc- tive strategies (Williams 1966; Dawkins 1976; Parker 1979; Females of many acquire nutrients from secre- Arnqvist and Rowe 2005). Females of many nuptial gift- tions provided by the mating male either by direct consump- giving species have evolved to mate with multiple partners tion or by absorbing them in their genital tract (Vahed 1998; at a rate beyond what is required to successfully fertilize all Gwynne 2008). These nuptial gifts include prey, seeds, glan- of their eggs (Arnqvist and Nilsson 2000). Given that mating dular products, and even body tissues of the mating male can be costly for females (Daly 1978), such polyandrous (Vahed 1998; Gwynne 2008). Nuptial gifts are commonly mating suggests that nuptial gifts benefit the female and viewed as a material benefit that increases female fitness. overcome those costs. But this has been challenged with the suggestion that gift in- There is evidence in some species that nutritious gifts pro- gredients may manipulate or coerce the female by increasing vide females with nutrients that increase the fitness of the male fitness but not hers (Arnqvist and Nilsson 2000; Vahed male’s own offspring (reviewed in Vahed 1998; Gwynne 2007). This follows from the hypothesis that the two sexes 2008), and female fitness in nuptial gift-giving in-

Received 3 March 2010. Accepted 4 August 2010. Published on the NRC Research Press Web site at cjz.nrc.ca on 28 September 2010. E.K. Piascik1,2 and D.T. Gwynne. Department of Biology, University of Toronto Mississauga, 3359 Mississauga Road, Mississauga, ON L5L 1C6, Canada. K.A. Judge. Department of Biology, University of Toronto Mississauga, 3359 Mississauga Road, Mississauga, ON L5L 1C6, Canada; Department of Biological Sciences, University of Lethbridge, 4401 University Drive West, Lethbridge, AB T1K 3M4, Canada. 1Corresponding author (e-mail: [email protected]). 2Present address: Department of Zoology, The University of British Columbia, Vancouver, BC V6T 1Z4, Canada.

Can. J. Zool. 88: 988–994 (2010) doi:10.1139/Z10-065 Published by NRC Research Press Piascik et al. 989 creased monotonically with increasing mating rate (Arnqvist Materials and methods and Nilsson 2000). However, nuptial gifts provided by males are also likely selected for their ability to manipulate the fe- Study animals male’s mating rate (Arnqvist and Nilsson 2000; Vahed In September 2008, we collected adult E. carolinus from 2007). In support of the coercion hypothesis, at least one the campus of the University of Toronto Mississauga study concludes that substances in the gift decrease female (43832’58@N, 79839’57@W). We used the first generation off- receptivity to additional matings (Sakaluk et al. 2006). spring of collected adults in the experimental trials. The ex- Effects of nuptial gifts on paternal investment and male perimental population was kept in a controlled environment relative mating success are not necessarily mutually exclu- room (25 8C, 70% relative humidity, 12 h light : 12 h dark sive, and in addition to selective advantages to males, it is photoperiod). We isolated the nymphs into individual plastic important to study the effects of nuptial gifts on females containers (9 cm diameter, 4.5 cm height) on the day that (Vahed 1998; Arnqvist and Nilsson 2000; Gwynne 2008). they hatched and weighed each to the nearest 0.1 mg (Met- If nuptial gifts provide direct benefits available from other tler AE50 electronic balance) on the following day, hereafter sources (e.g., nutrients and (or) water), then there should be termed moult mass. Each container had one vial of water a negative correlation between female mating rate and direct plugged with cotton for moisture and a piece of egg carton benefit received from other sources (Boggs 1990). A general for shelter. We fed the nymphs ground rabbit chow (Little method of testing this hypothesis involves manipulating the Friends Rabbit Food, Martin Mills Inc.) for the first 2 weeks, putative material benefit that the female receives from the after which they were fed whole pellets. Each day, we nuptial gift (Gwynne 2008). Under the material benefits hy- checked the nymphs for eclosion to adulthood and weighed pothesis, resource-deprived females are predicted to increase any new adults. their mating rate, decrease their latency to copulate, increase their copulation duration, and (or) increase their gift con- Female diet manipulation sumption. For example, poorly nourished female katydids Upon adult eclosion, we randomly assigned adult females (Kawanaphila nartee Rentz, 1993) increase their refractory to one of two diet treatments. The low-diet treatment con- period only when allowed to consume the spermatophylax sisted of 67% rabbit chow and 33% cellulose powder donated by the mating male (Simmons and Gwynne 1991). (Keycel 200-CT; Canada Colours and Chemicals Limited, In another example, Ursprung et al. (2009) showed that fe- Brampton, Ontario, Canada), whereas the high-diet treat- male seed beetles (Callosobruchus maculatus (Fabricius, ment consisted of 99% rabbit chow and 1% cellulose pow- 1775)) with access to nutrients and water increase their mat- der. Cellulose contains no nutritional material and its use in ing frequency only in response to water deprivation (see also the high-diet treatment was to control for any effects of the Edvardsson 2007; Fox and Moya-Laran˜o 2009). This study substance on the females. The proportion of cellulose pow- suggests that polyandry in this beetle species provides mate- der in our low-quality diet followed the methods of Wagner rial benefits to females that may offset some of the costs of and Hoback (1999) who reported a diet effect on male call- associating or mating with males (den Hollander and ing effort in the field cricket Gryllus lineaticeps Sta˚l, 1861 Gwynne 2009). Finally, in a nursery web spider (Pisaura (see also Wagner and Harper 2003). We fed adult males the mirabilis Clerck, 1757) that uses nuptial prey, starved fe- nymphal diet of 100% rabbit chow pellets. males accepted more copulations than fed individuals (Prokop and Maxwell 2009). Mating trials Males of some ground crickets (: , For the mating trials, we used male crickets between 4- ) donate hemolymph to their mates by allowing and 11-days postadult eclosion and females between 5- and the copulating female to chew on specialized spurs on his 12-days postadult eclosion. We systematically paired a fe- hind legs (Mays 1971). Female nemobiine crickets will male with a different male each day. Each female was given mate with more than one male and females of the southern an opportunity to mate with one male once a day for four ground cricket (Allonemobius socius (Scudder, 1877)) gain consecutive days. Experimental pairings were carried out in material benefits from multiple mating (Fedorka and Mous- blocks of 16 females (8 high diet and 8 low diet) and 16 seau 2002a). There is also evidence that tibial spur chewing males. We orchestrated pairings so that individual males causes longer copulation duration in striped ground cricket and females were paired only once and that each male in (Allonemobius fasciatus (De Geer, 1773)), possibly indicat- the block experienced 1 of the 16 possible sequences of fe- ing male control over insemination duration (Bidochka and males from each diet treatment (e.g., HLHL or LLLH; see Snedden 1985). In this study, we tested for material benefits Table 1). In this way, we could both control for the effects of tibial spur chewing in the Carolina ground cricket (Eune- of order of female presentation on male mating behaviour, mobius carolinus (Scudder, 1877)) by manipulating the nu- and investigate the effects of exposure to different females tritional content of female diet. We predicted that if females on male life span. We gave all individuals only one mating gain nutritional benefits from spur chewing, then females opportunity per day to allow males to replenish their nuptial fed on a low-quality diet will mate more and sooner and gift contribution. We gave pairs 90 min to mate based on will increase copulation and spur chewing durations relative preliminary observations of latency to copulate and copula- to females fed a high-quality diet. In addition, as the one tion duration in E. carolinus (E.K. Piascik, unpublished publication to describe mating behaviour in this species data). The mating trials were filmed in a sound-attenuating (Mays 1971) relied on observations of only two matings, room using video cameras equipped with infrared lighting the current study provides useful information on the basic (models DCR-HC28, DCR-TRV38, DCR-TRV330, DCR- mating behaviour of E. carolinus. TRV740; Sony Corp., Tokyo, Japan). We placed each mat-

Published by NRC Research Press 990 Can. J. Zool. Vol. 88, 2010

Table 1. Design for each block of experimental pairings of Carolina ground crickets (Eunemobius carolinus).

Male Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 HHHHHHHH L L L L L L L L 2 HHHHLL L L H H H H L L L L 3 HHLLHHLLHHLLHHLL 4 HLHLHLHLHLHLHLHL Note: H and L refer to females from the high- and low-quality diets, respectively. ing pair in a small plastic Petri dish (5.5 cm in diameter and Results 1 cm high) with a filter-paper lining. We covered each indi- vidual with a small plastic cap for 2 min prior to the trials to Mating behaviour allow for acclimation. The containers were positioned in sets In the single previous description of mating behaviour of of four beneath each of the four tripod-mounted cameras (i.e., E. carolinus, Mays (1971) observed two pairings. Here we four pairs per camera). We weighed all individuals to the describe mating encounters based on 70 complete mating se- nearest 0.1 mg (Mettler AE50 electronic balance; Mettler- quences. Toledo Inc., Columbus, Ohio, USA) before and after each Upon first encounter with the female, the male stridulated mating trial to assess mass changes associated with mating. and antennated her while swaying his body side to side. The We noted whether copulation occurred from the mating- male turned around and backed towards and then under the trial videos, and if so, then we recorded the latency to first female. This lasted for a few seconds before the female dis- successful copulation, which is measured as the time from mounted from the male and the pair separated whereupon introduction to a male until the start of copulation, and the the male continued to court the female for 33–86 min. copulation and spur chewing durations. When the pair came in close contact again, the female re- mounted the male and engaged genitalia. It appeared as Fecundity and life span though the male only brought his hind leg forward after the Once mated, we gave each female a new water vial genitalia became engaged. Once the leg was near the fe- plugged with gauze, which replaced the water vial plugged male’s mouthparts, she mouthed the area and chewed the with cotton. The gauze was used for oviposition and also as hind tibial spur. Chewing commonly continued throughout a water source, and we replaced it once a week. We meas- copulation on both hind spurs. No females fed without also ured total egg count as the total number of eggs collected copulating. In mating trials, copulation duration varied from from the gauze over the lifetime of the female. To measure 0.9 to 47.1 min. Copulation ended when the female turned the egg-laying rate, we divided the total number of eggs laid around and dismounted. The female appeared to drag the by each female by her life span (days) measured from the male for a few minutes while still coupled. During the brief day after her fourth mating opportunity. We monitored sur- periods where the female stopped dragging the male, he pro- vival daily and replaced the food and water weekly. duced a fast jerking motion that pushed her upwards and de- tached genitalia. After copulation, the female consumed the Statistical analyses spermatophore while the male remained near but facing To test the prediction that nutrient-deprived females will away. On two occasions a female cannibalized the male increase their mating rate, we used a univariate general lin- when they came in contact. The two males in question ap- ear model (GLM) with mating rate as the dependent varia- peared less than vigorous prior to the mating trial. ble, diet as the independent variable, and female body size (mass at adult molt) and female age as covariates. To test Effects of diet for differences between the two diets in mating behaviour of those females who mated, we used a multivariate GLM There were no significant differences between the low- with latency to copulate, copulation duration, spur chewing quality and the high-quality diet treatments in female mass, duration, and female mass change as dependent variables as calculated from mass at the adult moult to mass before and diet as the independent variable. To determine any ef- first mating opportunity (masshigh-quality diet = 11.9 ± 0.9 mg fects that the nutrition treatment had on total egg count and (mean ± SE), masslow-quality diet = 10.7 ± 0.9 mg; Student’s t egg-laying rate, we used Student’s t tests; for treatment ef- test, t[48] = 0.922, p = 0.361; Fig. 1). None of the interac- fects on female life span, we conducted a Kaplan–Meier sur- tions between female molt mass, age, and diet were statisti- vival analysis. To test for effects on male life span, we used cally significant, and so were removed from the final GLM pairwise Spearman’s rank correlations between the follow- before evaluation of the main effects on mating rate. The fe- ing variables: male postexperiment life span; number of males in the two diet treatments did not differ significantly matings; total mass change, total copulation duration, and in their mating rates (mating ratehigh-quality diet = 0.32 ± total spur chewing duration over the course of the experi- 0.04 mates/day, mating ratelow-quality diet = 0.36 ± 0.04 mates/ ment; and body size (moult mass). Where appropriate, we day; GLM, F[1,48] = 0.510, p = 0.479), and female body size adjusted for multiple testing using the sequential Bonferroni did not affect mating rate (GLM, F[1,48] = 0.038, p = 0.846), procedure (Holm 1979). All analyses were conducted using but there was a positive relationship between female age and SPSS version 10 (SPSS Inc., Chicago, Illinois, USA). mating rate (GLM, F[1,48] = 5.457, p = 0.024). Females that

Published by NRC Research Press Piascik et al. 991

Fig. 1. Change in mass (mean ± SE) for female Carolina ground Fig. 2. Relationship between female mating rate and female age at crickets (Eunemobius carolinus) fed either a high-quality diet (N = the start of the mating experiment of Carolina ground crickets (Eu- 26) or a low-quality diet (N = 24). Change in mass was the differ- nemobius carolinus). Diameter of data points is related to the num- ence between females’ molt mass and their mass just before their ber of observations with that combination of values (smallest = 1, first mating opportunity. largest = 6).

p = 0.027) (total fecunditymated = 71.9 ± 6.4 eggs, total were older at the start of the mating experiment were more fecundityunmated = 16.8 ± 6.1 eggs; Student’s t test, likely to mate than younger females (Fig. 2). t[25.8] = 6.216, sequential Bonferroni, p < 0.001), although Given the opportunity to mate with a male for four con- they did not differ in their egg-laying rate (egg-laying secutive days, only four females mated three times and no ratemated = 3.59 ± 0.31 eggs/day, egg-laying rateunmated = females mated four times. Multivariate GLM for both the 3.73 ± 1.55 eggs/day; Student’s t test, t[7.6] = 0.085, first and second matings with female age and moult mass as sequential Bonferroni, p = 0.934). We note, however, that covariates revealed no significant three- or two-way interac- individuals were not experimentally assigned different tion effects. These were removed from the final models. Re- numbers of mates and so any differences between females sults of the GLM lacking interaction terms revealed no that mated and those that did not may not be due to mat- significant effect of diet treatment on any of the mating var- ing per se and are likely due to some other factor that may also be the cause of the different number of mates. iables in either the first (Pillai’s Trace = 0.017, F[4,37] = 0.165, p = 0.955; Table 2) or second matings (Pillai’s Correlations among mating behaviours and male life Trace = 0.235, F[4,15] = 1.118, p = 0.385; Table 2). Although not significant but consistent with our predictions during span their second mating, there was a trend for females on the None of the measured variables correlated significantly low-quality diet to copulate sooner than those on the high- with male life span (Table 3). However, the number of mat- quality diet (Table 2, Fig. 3). ings correlated positively with the copulation duration and spur chewing duration of each mating, and male mass Treatment effects on female fitness change correlated negatively with the male’s number of mat- ings, copulation duration, spur chewing duration, and moult There were no significant differences between females fed mass (Table 3). Removal of males that failed to mate at the low-quality and high-quality diets in either their total fe- least once did not change the direction or magnitude of any cundity (total fecundityhigh-quality diet = 67.3 ± 9.2 eggs, total correlations, although in some cases the p value was only fecunditylow-quality diet = 76.5 ± 9.1 eggs; Student’s t test, significant at a = 0.1 (data not shown). t[42] = 0.711, sequential Bonferroni, p = 0.481) or their egg- laying rate (egg-laying rate = 3.94 ± 0.49 eggs/ high-quality diet Discussion day, egg-laying ratelow-quality diet = 3.25 ± 0.39 eggs/day; Student’s t test, t[42] = 1.107, sequential Bonferroni, p = By manipulating the material benefits females receive 0.550). However, females fed low-quality diet lived sig- during or directly after mating, it is possible to test whether nificantly longer than individuals fed high-quality diet nuptial feeding provides material benefits to females. If nup- (life spanhigh-quality diet = 24 ± 2 days, life spanlow-quality diet = tial gifts donated by the male are providing general nutrients 33 ± 3 days; Log Rank Statistic = 5.98, sequential Bonferroni, to the females, i.e., those available in the environment, then p = 0.045; Fig. 4). In addition, females that mated at least mating rates should increase when general sources are lim- once lived longer and laid more eggs than those that failed ited (Gwynne 2008; see Boggs 1990). Our results with the to mate (life spanmated = 28 ± 2 days, life spanunmated = nuptial feeding E. carolinus show that females fed a low- 14 ± 6 days; Log Rank Statistic = 6.12, sequential Bonferroni, quality diet did not mate more often or sooner than females

Published by NRC Research Press 992 Can. J. Zool. Vol. 88, 2010

Table 2. Descriptive statistics (mean ± SE) and statistical tests for the behavioural measures of mating effort in female Carolina ground crickets (Eunemobius carolinus) when assigned to either a high-quality or a low-quality diet treatment.

Diet Behaviour High-quality* Low-quality* F df p First mating Latency to copulate (min.) 45.2±1.9 46.3±2.4 0.104 1, 40 0.749 Copulation duration (min.) 32.8±2.4 34.2±2.0 0.097 1, 40 0.757 Spur chewing duration (min.) 24.6±2.1 24.6±2.3 0.004 1, 40 0.952 Mass change (mg) –2.6±0.3 –2.4±0.4 0.264 1, 40 0.610 Second mating Latency to copulate (min.) 62.9±3.9 55.4±2.2 1.945 1, 18 0.180 Copulation duration (min.) 26.3±3.7 30.4±2.1 0.457 1, 18 0.508 Spur chewing duration (min.) 19.4±3.2 20.1±1.3 0.023 1, 18 0.881 Mass change (mg) –2.5±0.5 –2.7±0.4 0.059 1, 18 0.811 *Sample sizes are 23 (high-quality diet) and 21 (low-quality diet) for the first mating, and 12 (high-quality diet) and 10 (low-quality diet) for the second mating.

Fig. 3. Latency to copulate (mean ± SE) during the second mating Fig. 4. Proportion of female Carolina ground crickets (Eunemobius for female Carolina ground crickets (Eunemobius carolinus) fed carolinus) left alive following the four mating opportunities for fe- either a high-quality diet (N = 12) or a low-quality diet (N = 10). males fed a high-quality diet (solid line, N = 22) and a low-quality diet (broken line, N = 21).

fed a high-quality diet, and when they did mate, they did not brush crickets (Cyphoderris strepitans Morris and Gwynne, copulate or feed for longer. This suggests that females of 1978), a species in which females feed on male hindwings this species may not receive direct benefits from multiple during mating, nonvirgin males are less likely to remate as matings. In contrast, a study on A. socius did show that mul- a result of female wing-feeding. Future work on E. carolinus tiple mating provided a material benefit to females and that could examine the condition dependence of male haemo- the size of the male’s gift predicted female fitness (Fedorka lymph gifts, as well as the effect of tibial spur chewing, on and Mousseau 2002a). Furthermore large male A. socius male immune function and life span. produce bigger gifts than small males (Fedorka and Mous- In E. carolinus, there remains the possibility that special- seau 2002b). ized nutrients (unavailable elsewhere) are transferred during Although our findings suggest that spur chewing does not spur chewing, or that males pass chemicals in the ingested provide direct benefits to females, this behaviour may be hemolymph that function as a chemical signal to the fe- costly to males. Our study showed that male mass loss cor- males, possibly as a form of coercion to increase male fit- related with both copulation and spur chewing durations. ness. If coercion is the case and the benefits to males of Therefore, longer copulations or spur chewing may result in coercion are high, then females should suffer some fitness males taking longer breaks from mating as they recoup lost cost as a result of mating and feeding on male tibial spurs energy reserves, although we were unable to detect a rela- (Clutton-Brock and Parker 1995). In the field cricket Gryllus tionship between either copulation duration, spur chewing bimaculatus De Geer, 1773, females exposed to courtship duration, or male size that would support this hypothesis and copulation had shorter life spans (Bateman et al. 2006). (data not shown). Sakaluk et al. (2004) found that in sage- However in our study, females that mated actually lived lon-

Published by NRC Research Press Piascik et al. 993

Table 3. Spearman’s rank correlation matrix of variables hypothesized to be related to the life span (N = 55 in all cells) of male Carolina ground crickets (Eunemobius carolinus).

Number of Copulation Spur chewing Variable Life span matings Mass change duration duration Number of matings 0.057 Mass change –0.038 –0.278* Copulation duration 0.007 0.955** –0.310* Spur chewing duration 0.057 0.908** –0.298* 0.945** Moult mass 0.102 –0.164 –0.283* –0.154 –0.116 Note: *, p < 0.05; **, p < 0.01. ger and laid more eggs than those that failed to mate. Tibial age effect is consistent with other studies showing that older spur chewing may also represent a form of male control virgin females are more likely to mate than younger ones over insemination (i.e., spermatophore attachment) duration (e.g., Kodric-Brown and Nicoletto 2001; Moore and Moore (Bidochka and Snedden 1985). However, we were unable to 2001). However, if females do mate not long after sexual evaluate this possibility because we did not measure sperma- maturation, then this effect is of dubious interest (Judge et tophore attachment time. Future tests of any negative effect al. 2010). Little is known about the mating rate and survi- of ingesting male hemolymph should be investigated by ma- vorship of E. carolinus in the wild, although we have heard nipulating the amount of hemolymph consumed by females males singing at our collection site from late summer until and measuring the duration of insemination. the first severe frost (K.A. Judge, personal observation). Ti- An alternative explanation for the lack of a diet effect on bial spur chewing provides a handy assay of male mating female mating behaviour is that the nutrient restriction in success, which could be used to measure selection on male our low-quality diet treatment was insufficient to limit the E. carolinus in the wild as has been done for the ground hypothesized material benefit in female E. carolinus. cricket A. socius (Fedorka and Mousseau 2002c). Although in the predicted direction, the effect of our diet This study provides further insight into the role of spur treatment on female body mass was not statistically signifi- chewing during mating in E. carolinus by examining the ef- cant (Fig. 1). However, this was possibly due to the large fects on both female and male fitness. Although we did not measurement error caused by (i) the overall small body size find spur chewing to provide direct benefits to the female or of E. carolinus (mean mass at adult molt = 28.3 ± 0.7 mg) represent a form of male coercion, our results do suggest and the relative imprecision of our mass measurements that spur chewing may be costly for males given that male (±0.1 mg) and (ii) variations in ambient humidity causing mass loss correlated with both copulation and spur chewing random fluctuations in individual water loss. Although we durations. To gain further understanding on the role of spur failed to detect any statistically significant difference be- chewing in this cricket species, future studies should exam- tween treatments in mass gain by female E. carolinus, the ine the relationship of both the gift size and the gift’s chem- fact that females fed low-quality diet lived longer than fe- ical composition to female fitness. males fed high-quality diet does indicate food stress because dietary restriction is known to extend the life span of ani- Acknowledgements mals (see Partridge et al. 2005). Reduced life span and con- Thank you to Laura Robson for help in collecting ground comitant increased fecundity from overeating can in theory crickets, to Glenn Morris for the use of his sound- be coerced by male seminal products (tested and refuted for attenuating room, and to the D.T. Gwynne laboratory for the fruit fly Drosophila melanogaster Meigen, 1830 by discussion of experimental design and ideas. Barnes et al. 2008). However, the hypothesis that food re- striction in female crickets limited potential overfeeding in- References duced by specialized substances in the hemolymph meal is not supported given our finding of no diet effects on fecund- Arnqvist, G., and Nilsson, T. 2000. The evolution of polyandry: ity. multiple mating and female fitness in insects. Anim. Behav. The diet-related differences in life span (Fig. 4) but not 60(2): 145–164. doi:10.1006/anbe.2000.1446. PMID:10973716. egg laying rate or fecundity could be the result of lower Arnqvist, G., and Rowe, L. 2005. Sexual conflict. Princeton Uni- power to detect differences in the latter two traits relative to versity Press, Princeton, N.J. Barnes, A.I., Wigby, S., Boone, J.M., Partridge, L., and Chapman, life span. It is likely easier to detect differences in life span T. 2008. Feeding, fecundity and lifespan in female Drosophila than in fecundity because the accuracy and precision of life- melanogaster. Proc. R. Soc. Lond. B Biol. Sci. 275(1643): span measurements increases with the effort of the experi- 1675–1683. doi:10.1098/rspb.2008.0139. menter (i.e., with the frequency of checking for dead indi- Bateman, P.W., Ferguson, J.W.H., and Yetman, C.A. 2006. Court- viduals), whereas fecundity measurements are probably ship and copulation, but not ejaculates, reduce the longevity of more sensitive to environmental factors such as the suitabil- female field crickets (Gryllus bimaculatus). J. Zool. (Lond.), ity of the oviposition substrate. 268(4): 341–346. doi:10.1111/j.1469-7998.2006.00054.x. In our study, female E. carolinus who experienced their Bidochka, M.J., and Snedden, W.A. 1985. Effect of nuptial feeding first mating opportunity at an older age had higher mating on the mating behaviour of female ground crickets. Can. J. Zool. rates than females who encountered their first male at 63(1): 207–208. doi:10.1139/z85-032. younger ages (Fig. 2) regardless of their diet treatment. This Boggs, C.L. 1990. A general model of the role of male-donated nu-

Published by NRC Research Press 994 Can. J. Zool. Vol. 88, 2010

trients in female insects’ reproduction. Am. Nat. 136(5): 598– Natl. Acad. Sci. U.S.A. 98(16): 9171–9176. doi:10.1073/pnas. 617. doi:10.1086/285118. 161154598. PMID:11470903. Clutton-Brock, T.H., and Parker, G.A. 1995. Sexual coercion in an- Parker, G.A. 1979. Sexual selection and sexual conflict. In Sexual imal societies. Anim. Behav. 49(5): 1345–1365. doi:10.1006/ selection and reproductive competition in insects. Edited by anbe.1995.0166. M.S. Blum and N.A. Blum. Academic Press, New York. Daly, M. 1978. Cost of mating. Am. Nat. 112(986): 771–774. pp. 123–166. doi:10.1086/283319. Partridge, L., Piper, M.D.W., and Mair, W. 2005. Dietary restric- Dawkins, R. 1976. The selfish gene. Oxford University Press, Oxford. tion in Drosophila. Mech. Ageing Dev. 126(9): 938–950. den Hollander, M., and Gwynne, D.T. 2009. Female fitness conse- doi:10.1016/j.mad.2005.03.023. PMID:15935441. quences of male harassment and copulation in seed beetles, Cal- Prokop, P., and Maxwell, M.R. 2009. Female feeding regime and losobruchus maculatus. Anim. Behav. 78(5): 1061–1070. doi:10. polyandry in the nuptially feeding nursery web spider, Pisaura 1016/j.anbehav.2009.06.036. mirabilis. Naturwissenschaften, 96(2): 259–265. doi:10.1007/ Edvardsson, M. 2007. Female Callosobruchus maculatus mate s00114-008-0477-6. PMID:19050843. when they are thirsty: resource-rich ejaculates as mating effort Sakaluk, S.K., Campbell, M.T.H., Clark, A.P., Johnson, J.C., and in a beetle. Anim. Behav. 74(2): 183–188. doi:10.1016/j. Keorpes, P.A. 2004. Hemolymph loss during nuptial feeding anbehav.2006.07.018. constrains male mating success in sagebrush crickets. Behav. Fedorka, K.M., and Mousseau, T.A. 2002a. Material and genetic Ecol. 15(5): 845–849. doi:10.1093/beheco/arh113. benefits of female multiple mating and polyandry. Anim. Behav. Sakaluk, S.K., Avery, R.L., and Weddle, C.B. 2006. Cryptic sexual 64(3): 361–367. doi:10.1006/anbe.2002.3052. conflict in gift-giving insects: chasing the chase-away. Am. Nat. Fedorka, K.M., and Mousseau, T.A. 2002b. Tibial spur chewing in 167(1): 94–104. doi:10.1086/498279. PMID:16475102. ground crickets: larger males contribute larger gifts (Orthoptera: Simmons, L.W., and Gwynne, D.T. 1991. The refractory period of Gryllidae). Fla. Entomol. 85(2): 317–323. doi:10.1653/0015- female katydids (Orthoptera: Tettigoniidae): sexual conflict over 4040(2002)085[0317:TSFIGC]2.0.CO;2. the remating interval? Behav. Ecol. 2(4): 276–282. doi:10.1093/ Fedorka, K.M., and Mousseau, T.A. 2002c. Nuptial gifts and the beheco/2.4.276. evolution of male body size. Evolution, 56(3): 590–596. doi:10. Ursprung, C., den Hollander, M., and Gwynne, D.T. 2009. Female 1111/j.0014-3820.2002.tb01369.x. PMID:11989688. seed beetles, Callosobruchus maculatus, remate for male- Fox, C.W., and Moya-Laran˜o, J. 2009. Diet affects female mating behaviour in a seed-feeding beetle. Physiol. Entomol. 34(4): supplied water rather than ejaculate nutrition. Behav. Ecol. 370–378. doi:10.1111/j.1365-3032.2009.00700.x. Sociobiol. 63(6): 781–788. doi:10.1007/s00265-009-0711-z. Gwynne, D.T. 2008. Sexual conflict over nuptial gifts in insects. Vahed, K. 1998. The function of nuptial feeding in insects: a re- Annu. Rev. Entomol. 53(1): 83–101. doi:10.1146/annurev.ento. view of empirical studies. Biol. Rev. Camb. Philos. Soc. 73(1): 53.103106.093423. PMID:17680720. 43–78. doi:10.1017/S0006323197005112. Holm, S. 1979. A simple sequentially rejective multiple test proce- Vahed, K. 2007. All that glisters is not gold: sensory bias, sexual dure. Scand. J. Stat. 6(2): 65–70. conflict and nuptial feeding in insects and spiders. Ethology, Judge, K.A., Tran, K.-C., and Gwynne, D.T. 2010. The relative ef- 113(2): 105–127. doi:10.1111/j.1439-0310.2006.01312.x. fects of mating status and age on the mating behaviour of fe- Wagner, W.E., Jr., and Harper, C.J. 2003. Female life span and male field crickets. Can. J. Zool. 88(2): 219–223. doi:10.1139/ fertility are increased by the ejaculates of preferred males. Z09-139. Evolution, 57(9): 2054–2066. doi:10.1111/j.0014-3820.2003. Kodric-Brown, A., and Nicoletto, P.F. 2001. Age and experience tb00385.x. PMID:14575327. affect female choice in the guppy (Poecilia reticulata). Am. Wagner, W.E., Jr., and Hoback, W.W. 1999. Nutritional effects on Nat. 157(3): 316–323. doi:10.1086/319191. PMID:18707292. male calling behaviour in the variable field cricket. Anim. Mays, D.L. 1971. Mating behavior of Nemobiine crickets — Behav. 57(1): 89–95. doi:10.1006/anbe.1998.0964. PMID: Hygronemobius, Nemobius, and Pteronemobius (Orthoptera: 10053075. Gryllidae). Fla. Entomol. 54(2): 113–126. doi:10.2307/3493557. Williams, G.C. 1966. Adaptation and natural selection: a critique of Moore, P.J., and Moore, A.J. 2001. Reproductive aging and mating: some current evolutionary thought. Princeton University Press, the ticking of the biological clock in female cockroaches. Proc. Princeton, N.J.

Published by NRC Research Press