Predation on reproducing wolf spiders: access to information has differential effects on male and female survival

Animal Behaviour 128 (2017) 165e173

Contents lists available at ScienceDirect

Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav

Predation on reproducing wolf spiders: access to information has differential effects on male and female survival Ann L. Rypstra a, *, Chad D. Hoefler b, 1, Matthew H. Persons c, 2 a

Department of Biology, Miami University, Hamilton, OH, U.S.A. Department of Biology, Arcadia University, Glenside, PA, U.S.A. c Department of Biology, Susquehanna University, Selinsgrove, PA, U.S.A. b

a r t i c l e i n f o Article history: Received 13 September 2016 Initial acceptance 8 November 2016 Final acceptance 24 March 2017 MS. number: A16-00806R2 Keywords: chemical cue courtship differential predation limited attention predation risk sensory ecology wolf spider

Predation has widespread influences on animal behaviour, and reproductive activities can be particularly dangerous. Males and females differ in their reactions to sensory stimuli from predators and potential mates, which affects the risk experienced by each sex. Thus, the information available can cause differential survival and have profound implications for mating opportunities and population structure. The wolf spider, Pardosa milvina, detects and responds in a risk-sensitive manner to chemotactile information from a larger predator, the wolf spider Tigrosa helluo. Male P. milvina use similar chemotactile cues to find females whereas female P. milvina focus on the visual, and perhaps vibratory, aspects of the male display. Our aim was to document the risk posed by T. helluo predators on P. milvina during reproduction and to determine whether augmenting chemotactile information would affect that outcome. In the laboratory, we explored the effects of adding predator and/or female cues on the predatory success of T. helluo on P. milvina males or observing females. Additional cues from prospective mates or from predators enhanced male survival. The addition of female cues increased predation on females whereas predator cues augmented female survival. In field enclosures, we documented the impact of T. helluo, with and without additional predator cues, on the sex ratio of survivors and the reproductive success of females. Additional predator cues shifted the sex ratio towards males, however, 90% of the remaining females in that treatment produced eggsacs whereas less than 60% reproduced in female-biased populations. Thus, augmenting the available predator information shifted the risk from males to females, presumably due to differences in their sensory priorities. By altering the availability of potential mates, this shift appears to have influenced the intensity of sexual selection for this spider. © 2017 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Males and females experience different levels of predation risk due to the contrasting sex roles that are implicit in any breeding
kely, system (Emlen & Oring, 1977; Kokko & Jennions, 2008; Sze
kely, Weissing, & Liker, Freckleton, Fichtel, & Kappeler, 2014; Sze Komdeur, 2014). By necessity, the behavioural responses of each sex are distinct but, because males and females must ultimately come together for reproduction, the reactions of one can affect the success of the other. Thus, these interactions may influence differential predation and affect the adult sex ratio of the population
kely, Liker et al., 2014; Sze
kely, Weissing et al., 2014). Ulti(Sze mately, any shifts in the sex ratio can feedback to influence the

* Correspondence: A. L. Rypstra, Department of Biology, 1601 University Blvd, Miami University, Hamilton, OH 45011, U.S.A. E-mail address: [email protected] (A. L. Rypstra). 1 E-mail address: hoefl[email protected] (C. D. Hoefler). 2 E-mail address: [email protected] (M. H. Persons).

efficacy of the mating system and further distinguish the sex roles
kely, 2013). (Fitze & Le Galliard, 2008; Liker, Freckleton, & Sze The detectability of the signals exchanged by males and females is a major factor that drives differential predation. In species with ‘classical’ sex roles, males have more outlandish characters and/or engage in more conspicuous activities that are meant to persuade females and, as a result, they are putatively under more predation pressure (Clark, Zeeff, Karson, Roberts, & Uetz, 2016; Costantini,
kely, Weissing et al., 2014; Bruner, Fanfani, Dell'Omo, 2007; Sze Zuk & Kolluru, 1998). On the other hand, while these prominent male features attract the unwanted attention of predators, they may distract approaching females who are then targeted by the predators (Hughes, Kelley, & Banks, 2009, 2012). For example, the calls of male crickets (Gryllodes supplicans) are intended as an advertisement to females but end up attracting gecko (Hemidactylus tursicus) predators that then preferentially prey on the females (Sakaluk & Belwood, 1984). Similarly, pike cichlids (Crenicichla alta) approach the male mating displays of guppies, Poecilia

http://dx.doi.org/10.1016/j.anbehav.2017.03.032 0003-3472/© 2017 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

166

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

reticulata, and then consume the drab females nearby (Pocklington & Dill, 1995). Thus, the relative risk experienced by males and females during courtship and mating are not necessarily easy to predict, even when there are obvious differences in the detectability of their signals. Some proportion of the susceptibility of animals is related to the need to communicate, which typically involves the exchange of a diversity of messages that engage multiple sensory modalities (Hebets & Papaj, 2005; Higham & Hebets, 2013; Partan & Marler, 1999, 2005; Uetz & Roberts, 2002). Cognitive limitations place constraints on the amount of information that can be processed and interpreted, thus the attention focused on one set of signals reduces the ability of animals to receive and react to other inputs (Dukas, 2002, 2004; Schmidt, Dall, & van Gils, 2010). Existing evidence suggests that the complexity of the sensory involvement, including the spatial and temporal sequence in which cues are received, affects whether and when they elicit a reaction (Clark et al., 2016; Hebets & Papaj, 2005; Munoz & Blumstein, 2012; Stephenson, 2016). In some cases, an initial stimulus serves to alert the recipient and enhance the detectability and discriminability of subsequent signals that enlist additional sensory modalities (Driver & Spence, 2004; Munoz & Blumstein, 2012; Rowe, 1999). For example, when glowlight tetras (Hemigrammus erythrozonus) are exposed to chemical cues from predators, they respond in a stronger and more specific manner to visual cues (Brown, Poirier, & Adrian, 2004; Wisenden, Vollbrecht, & Brown, 2004). In other instances, one sensory modality may take precedence as the primary source of environmental information or even distract individuals from making biologically relevant assessments (Blumstein, 2014; Dukas, 2004; Hartman & Abrahams, 2000). For example, when male noctuid moths (Spodoptera littolalis) focus their attention on the quality and quantity of female sex pheromones, they are effectively deaf to the sonar signals from predatory bats (Skals, €fstedt, & Surlykke, 2005). These examples Anderson, Kanneworff, Lo underscore the importance of cognitive capacity of males and females and the manner in which they prioritize their limited attention. Sexbased differences in sensory modalities, timing of information transfer and ability to react to appropriately to predator cues may translate into differences in predation risk. In this way, the sensory landscape can influence the adult sex ratio and potentially place selective pressure on the breeding system (Bro-Jørgensen, 2010). Our goal was to examine how the availability of various cues would affect the success of a predator housed with males and females during courtship. We then documented whether the differences in predation on the sexes were sufficient to affect the adult sex ratio and reproductive success of females in seminatural populations. Wolf spiders (Lycosidae) are a useful group with which to investigate complex signalling and its impact on ecology (Clark et al., 2016; Hebets, 2011; Hebets & Papaj, 2005; Roberts, Taylor, & Uetz, 2007; Uetz, 2000). Members of this group communicate using multiple sensory modalities during foraging (Persons, 1999; Persons & Uetz, 1996), courtship (Hebets & Papaj, 2005) and mating (Uetz & Roberts, 2002). In addition, wolf spiders are amenable to manipulative studies aimed to tease apart how each one of these factors affects mating success (Rypstra, Wieg, Walker, & Persons, 2003), foraging (Persons, Walker, & Rypstra, 2002) and their susceptibility to predation (Persons, Walker, Rypstra, & Marshall, 2001). We deployed a well-characterized wolf spider system where females attract males with substrate-borne chemical and tactile cues that cause those males to respond with a conspicuous visual courtship display, possibly accompanied by vibratory signals (Rypstra et al., 2003). Both males and females can extract specific information about a common coexisting predator, also a wolf spider, from its chemical and tactile cues (Bell, Rypstra, & Persons, 2006; Persons & Rypstra, 2001; Persons et al., 2001). Thus, the nature of the cues used to detect this predator are similar to those

that are used by females to attract males, but the cues used by males to attract females stimulate different sensory modalities. We predicted that the presence of predator cues would alter the survival of males and females and result in a biased sex ratio that would also affect reproductive success. Specifically, the attention that males must focus on chemotactile cues as they search for females should allow them to detect and respond to risk sooner when the same type of predator information is available. On the other hand, females, with their attention directed towards assessing the male's conspicuous display, may be less likely or less able to respond to other environmental information. STUDY SYSTEM The wolf spider, Pardosa milvina (Araneae, Lycosidae), is a particularly apt species with which to address these questions. Females advertise to males using air- and substrate-borne chemical cues (Rypstra et al., 2003; Searcy, Persons, & Rypstra, 1999). Males garner information about the female's mating status and hunger level from chemotactile cues (silk, faeces and other excreta) deposited on a surface that has been occupied by a female (Rypstra, Schlosser, Sutton, & Persons, 2009; Rypstra et al., 2003; Schlosser, 2005). Once males detect female cues, they begin to court, which ultimately lures females out of hiding places to observe and possibly mate with the male (Rypstra, Walker, & Persons, 2016; Rypstra et al., 2003). Male and female Pardosa also detect air- and substrate-borne chemotactile cues from a common predator, the larger wolf spider, Tigrosa helluo (Araneae, Lycosidae) (Persons et al., 2001; Shonewolf, Bell, Rypstra, & Persons, 2006). Their response to the chemotactile cues of T. helluo is costly (Persons et al., 2002) but effectively increases survival (Persons et al., 2001). Indeed, the reactions of P. milvina are accurately gauged to the risk posed by the T. helluo individual that produces the cues; P. milvina detect and adjust their response in a threat-sensitive manner that reflects the potential predator's size (Persons & Rypstra, 2001), sex (Lehmann, Walker, & Persons, 2004), hunger level (Bell et al., 2006) and recent diet (Persons et al., 2001). Notably, the courtship display of male P. milvina render them more susceptible to attack by T. helluo, but courting males in good condition are better able to survive than those in poor condition (Hoefler, Persons, & Rypstra, 2008). Here we report the results of two experiments that aimed to explore the effects of chemotactile cues on the relative success of T. helluo preying on male or female P. milvina during courtship and mating. In our designs, we added additional cues in order to ensure that the subjects could detect and react to the information immediately upon entering the experimental arena. In a laboratory experiment, we tested the hypothesis that abundant chemical information regarding females and/or predators would have differential effects on the mortality of male and female P. milvina during courtship. Because the impact of T. helluo cues on behaviour and sexual selection in P. milvina has been documented in a variety of other situations (Hoefler et al., 2008; Persons et al., 2002, 2001; Rypstra et al., 2016), we conducted a field experiment to determine whether the effect of abundant predator information that has been observed in the laboratory was sufficient to affect the adult sex ratio and mating success of P. milvina populations housed with T. helluo in a more natural situation. METHODS Basic Laboratory Maintenance All spiders were collected from fields at Miami University's Ecology Research Center, Oxford, Ohio, U.S.A. (39
31052.6800 ,

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

84
430 22.0800 ). When in the laboratory, they were held in an environmental room on a 13:11 h light:dark cycle set at 25
C with 50% relative humidity. Animals were housed individually and regularly provided a combination of crickets (Acheta domesticus) and fruit flies (Drosophila melanogaster) as prey. The specific conditions mimicked those deployed in past experiments (Persons et al., 2001; Rypstra et al., 2003). All P. milvina used in experiments were collected as immatures and held in isolation until they moulted to adulthood so that we could be certain that they had not mated. Experiments were run 1e3 weeks after the spiders' final moult. All T. helluo were adult females that had been caught in fields at the Ecology Research Center at least 3 weeks prior to experimentation. No spider was involved in more than one aspect of one experiment. We washed containers, enclosures and all other reusable items that came in contact with spiders or their cues, wiped them down with 70% ethanol and allowed them to dry completely between each use. Disposable items, such as the paper or straw used to present cues to animals during experiments, were discarded after one use. Ethical Note None of the species we used in experiments are endangered or threatened and there are no state or federal regulations (Ohio, U.S.A.) governing their care. When in the laboratory, animals were maintained under conditions that have been demonstrated to result in high survivorship and reproductive capacity. Spiders are predators and, as such, their success means that some other invertebrates were sacrificed. The spiders were fed individually to allow for an efficient capture and to minimize stress to the prey. Experiments were closely monitored and terminated when subjects were attacked. After experimentation, the survivors were either returned to the field from where they were collected, maintained under standard laboratory conditions until death, or anaesthetized by cooling and then placed in 70% ethanol. Laboratory Experiment The goal of this experiment was to determine whether a surplus of information regarding predators and/or virgin females would affect predatory success of T. helluo on P. milvina males or females during courtship. We used chemotactile cues because they function to increase predator detection and aid in the identification of potential mates for P. milvina. We added them to the experimental arena prior to introducing the spiders so that all participants could react to the signals from the onset of a trial. Because male P. milvina prioritize chemical information in their search for females, we predicted that they would be more responsive to these added cues. The experiment included four treatments: (1) added cues from P. milvina females (N ¼ 20); (2) added cues from T. helluo predators (N ¼ 20); (3) added cues from both females and predators arranged in a patch work (N ¼ 20) (Fig. 1); (4) control with no added cues (N ¼ 20). Experiments were conducted in cylindrical arenas (19 cm in diameter) lined with filter paper (Whatman No. 1) that was used to deliver the chemotactile cues. The floor of all arenas was completely covered with one large piece of filter paper. For treatments with female cues, a female P. milvina was housed in the arena for 24 h prior to trials and allowed to deposit silk, faeces and other excreta on the filter paper as she roamed the arena (as in Rypstra et al., 2009, 2003). In all trials, nine smaller filter paper disks, each 3.5 cm in diameter, were arranged on top of the sheet that covered the arena floor (Fig. 1). For treatments with predator cues, a female T. helluo was housed in a separate arena (12 cm in diameter) with the nine filter paper disks arranged in a nonoverlapping pattern on the bottom. That spider was allowed to deposit silk,

167

Female cues (or no cues)

Predator cues (or no cues)

Figure 1. Schematic of laboratory test arena indicating how chemotactile cues (if present) from predators and from females were arranged.

faeces and other excreta on the disks for 24 h (as in Persons et al., 2002, 2001). These disks were placed in the experimental arena no more than 5 min prior to the commencement of a trial (Fig. 1). We standardized the hunger of the participants from each spider species prior to experimentation. Two days before the experiment, virgin P. milvina males and females were provided crickets ad libitum to reduce the likelihood of cannibalism. On the other hand, we wanted the predators to be motivated to feed, so we held T. helluo individuals with access to water but no food for 10 days prior to the experiment. After the last feeding, we measured the carapace and abdomen width of all spiders using an ocular micrometer accurate to 0.1 mm. The cephalothorax only changes at moult and is a measure of spider size, but the abdomen expands as a spider feeds and thus can be used to estimate recent feeding history or condition (Anderson, 1974). We verified that the spiders selected for the various treatments did not differ in size by comparing the cephalothorax width using separate one-way ANOVAs for T. helluo (F3,76 ¼ 0.22, P ¼ 0.88), P. milvina males (F3,76 ¼ 0.13, P ¼ 0.94) and P. milvina females (F3,76 ¼ 0.20, P ¼ 0.90). We then looked for differences in the condition of spiders assigned to treatments using ANCOVAs of the abdomen width with cephalothorax width as a covariate to control for spider size (Garcia-Berthou, 2001). We ran separate analyses for T. helluo (F4,75 ¼ 0.19, P ¼ 0.90), P. milvina males (F3,75 ¼ 0.07, P ¼ 0.98) and P. milvina females (F3,75 ¼ 0.18, P ¼ 0.91). Experiments were run in batches with representatives of all treatments included each time. The arena was prepared by arranging the nine small disks of filter paper (with predator cues or blank) on top of the larger sheet that completely covered the bottom (with female cues or blank) (Fig. 1). Virgin male and female P. milvina were then introduced to opposites sides of the arena sequestered under opaque vials that were 1.5 cm in diameter and 2 cm tall. The predator T. helluo was placed under a translucent plastic vial, 3 cm in diameter, in the centre of the arena. All spiders were positioned in the arena such that they could contact any and all cues. After a 2 min acclimation period, the P. milvina individuals were released and, 1 min later, the T. helluo predator was released. We watched the arena for 30 min and recorded whether and when the T. helluo captured a male or female P. milvina. We used nominal logistic regression to test the hypothesis that added chemotactile information regarding the presence of a female or a potential predator would have differential effects on male and

168

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

female mortality during courtship. We then compared the timing and frequency of male versus female kills across all treatments using a right-censored proportional hazard test. To explore the specific effects of our treatments on each sex, we conduced separate two-way proportional hazard tests to determine the effect of predator cues, female cues and any interaction between cue types on the timing and frequency of predation on males and females. We followed these tests with pairwise comparisons among treatments using a familywise error rate of 0.008 (Haynes, 2013). All analyses were conducted in JMP® Pro 11.2.0 (SAS Institute, Cary, NC, U.S.A.). Field Experiment We conducted a field experiment to test the hypothesis that abundant predator information would have differential effects on male and female P. milvina success during courtship and mating. Because of the differences in the sensory priorities of males and females, we predicted that this added information would shift the predation risk from active, but chemically aware, males to females. We determined whether differential predation was sufficient to affect the sex ratio and, in so doing, affect the ability of females to mate and produce eggsacs. In this experiment, we introduced virgin male and female P. milvina into mesocosms and allowed them to interact and mate for 1 week. Treatments included: (1) control: P. milvina males and females with no T. helluo and no T. helluo cues; (2) predator: live T. helluo allowed to freely interact with P. milvina males and females; (3) cues added: live T. helluo along with additional T. helluo cues introduced into the enclosure with the P. milvina males and females. These treatments allowed us to determine whether advanced warning of the presence of T. helluo predators through the chemotactile cues affected the success of reproductively active P. milvina males and females in a more natural situation. The field experiment was conducted in 15 circular mesocosms (2 m2 in area, 1.6 m in diameter) constructed of aluminium flashing and installed in fields at the Ecology Research Center. Enclosure walls were inserted 10 cm into the soil so that the spiders could not burrow under the sides. The walls extended 40 cm above the surface and the top 5 cm was treated with RainX® (ITW Global Brands, Houston, TX, U.S.A.) to reduce spider escape. We installed the enclosures 5e7 days after the fields had been cultivated, so the interior of each enclosure was free of living vegetation. To create a favourable habitat and encourage a robust prey community for the spiders, we added a 7e10 cm layer of straw mulch sprinkled with 50 g of dry fruit fly medium (Carolina Biological Supply, Burlington, NC, U.S.A.) 10e14 days before the experiment (as in Walker & Rypstra, 2003). We completed three runs of the experiment on 1e7 June 2008, 5e12 July 2008 and 2e8 August 2008. In each run, five of the 15 mesocosms were randomly assigned to each of the three treatments. The mesocosms were removed, washed and installed in a different field between each experimental period. One week before an experimental run, we standardized the hunger level of the spiders by allowing them to feed ad libitum for 24 h. We then measured the cephalothorax width, as a measure of body size, and abdomen width of each spider using an ocular micrometer accurate to 0.1 mm. Spiders were randomly assigned to treatments and the body measurements enabled us to verify that the size and condition of spiders assigned to a given treatment were similar at the beginning of the experiment. We used one-way ANOVAs to compare the size, represented as cephalothorax width, of T. helluo (F1,179 ¼ 0.13, P ¼ 0.72), P. milvina males (F2,672 ¼ 0.38, P ¼ 0.69) and P. milvina females (F2,672 ¼ 0.34, P ¼ 0.71). We then analysed abdomen width, as a measure of feeding history, in separate ANCOVAs that included

cephalothorax width as a covariate for T. helluo (F1,178 ¼ 0.25, P ¼ 0.81), P. milvina males (F2,671 ¼ 0.18, P ¼ 0.84) and P. milvina females (F2,671 ¼ 0.22, P ¼ 0.80) (Garcia-Berthou, 2001). After measurement, we marked each spider with a dot of nontoxic paint (Testors®, Vernon Hills, IL, U.S.A.) on their carapace or abdomen so that we could identify each individual. They were then returned to their home container and held in the laboratory with access to water but no additional food. For each mesocosm in the added-cue treatment, we collected chemotactile cues on straw from four T. helluo females that were not otherwise involved in the experiment. Twenty-four hours before the cues were to be added to a field enclosure, we placed each of these T. helluo females in a plastic container, 20 cm in diameter and 8 cm deep, with 25 g of autoclaved straw. A moist cotton ball affixed to the side provided the spider with water. We removed the spider from the container and introduced the cueladen straw into the field enclosure twice, on day 1 and day 4 of the 7-day experimental period. We introduced four clumps of clean autoclaved straw to enclosures assigned to the control and predator-only treatments at the same time as we introduced the straw containing predator cues to the added-cue treatment. For all experimental runs, we established the treatments in all enclosures between 0600 and 0900 hours on the first day. We started the experiment by adding two T. helluo females to each of the mesocosms assigned to the predator treatment and the addedcue treatment. We then added four 25 g clumps of straw to each enclosure. The straw in the added-cue treatment was laden with T. helluo cues and the straw for control and predator treatments was clean. Finally, 30 virgin P. milvina individuals, 15 males and 15 females, were released in each mesocosm at a site that was at least 10 cm away from any other spider. On day 4, we replenished the cues in the added-cue enclosures with four clumps of cue-laden straw. We controlled for that manipulation by adding four clumps of clean autoclaved straw to the rest of the enclosures. On day 7, we searched each enclosure comprehensively and captured all spiders (as in Walker & Rypstra, 2003). We kept all P. milvina females in the laboratory where they were fed twice weekly and monitored for 30 days. Any eggsacs produced were weighed and opened and the eggs counted. For each enclosure, we represented the survivorship of P. milvina males and females as the proportion of each sex recovered alive out of the 15 individuals that we originally introduced. The final sex ratio was represented as the proportion of all the individuals we recovered that were male. Since all the P. milvina females were virgins at the outset of the experiment, female mating success was estimated by calculating the proportion of all the surviving females that produced eggsacs. For those females that produced eggsacs, we calculated an average clutch size per enclosure and used that as a measure of reproductive success. We looked for differences in the survival of males and females across treatments using MANOVA with the males and females recovered from the same enclosure as a repeated measure. We used the logit transformation with a constant 3 of 0.001 (Warton & Hui, 2011). We used the approach discussed in Warton and Hui (2011) to verify that this strategy successfully reduced heteroscedasticty, eliminated overdispersion and normalized the error structure of the proportional data. We evaluated treatment effects on male survivorship, female survivorship, sex ratio, mating success and average clutch size per enclosure with a linear mixed model using the restricted maximum likelihood (REML) approach. Treatment (control, predator, added cue) was the fixed effect and the experimental run (June, July, August) was entered as a random effect. Tukey HSD post hoc tests were used to identify specific treatment differences. All analyses were conducted in JMP® Pro 11.2.0 (SAS Institute).

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

169

RESULTS

Field Experiment

Laboratory Experiment

Survivorship of both species of spiders introduced into field mesocosms was high. We recovered 58 of the 60 T. helluo involved in the experiment and approximately 80% of the P. milvina in control enclosures. However, the predator treatments affected male and female survival differently (MANOVA: container)treatment: F2,40 ¼ 15.51, P < 0.0001). Female survival was lowest in the addedcue treatment whereas male survival was lowest in the predator treatment (with no added cues) (Table 2, Fig. 3). This differential shifted the sex ratio towards females in the predator treatment and towards males in the added-cue treatment (predators along with additional cues) (Table 2, Fig. 3c). Our manipulations also affected reproduction in field enclosures. Ninety-six per cent of the females that we recovered from the added-cue treatment produced eggsacs, versus 75% in the control treatment and 57% in the predator treatment (Table 2, Fig. 4a). Clutch sizes for females in both the predator treatment and the added-cue treatment were significantly smaller (by about 25%) than those for females in the control enclosures where there was no T. helluo or T. helluo information whatsoever (Table 2, Fig. 4b).

Over all of our treatments, male and female P. milvina were equally likely to be predated (Fisher exact test: P ¼ 0.871); males were killed in 38 trials, females were killed in 35 trials and no spiders were killed in 7 of the 80 trials. However, males survived significantly longer than females (c22 ¼ 42.85, P < 0.0001). On average, T. helluo captured females in 8.7 ± 1.0 min, but it took them 14.2 ± 1.0 min to catch males. Logistic regression indicated that the addition of chemotactile cues had different effects on the mortality of male and female P. milvina (Table 1). Notably, males were captured more often than females (>50% of captures) except in the trials where female cues were added to the arena without predator cues. In the female cue treatment, females were predated in 16 of the 20 trials (80%) whereas males were captured in only three instances (15%). When we added predator cues along with female cues, T. helluo captured males in 10 (50%) and females in 8 (40%) of the 20 trials. Thus, the effect of female cues was eliminated, which produced a significant interaction between the two cues types in the model (Table 1). When we explored the treatment effects on the timing with which males were apprehended, it was evident that the cue treatments affected both the timing and frequency of capture (Table 1, Fig. 2a). Specifically, there was a significant interaction between cue types in the proportional hazard model, since the addition of cues of any type, separately or together, helped males to survive longer (Table 1, Fig. 2a). On average, males lasted for 8.4 ± 1.6 min in trials where no cues were added to the arena, but they escaped predation twice as long (16.8 ± 1.1 min) when predator cues, female cues or both types of cues were included in the arena before the trial commenced (Fig. 2a). Both types of cues significantly affected the timing and frequency of female captures, but the interaction between the two treatments was not significant in the proportional hazard model (Table 1, Fig. 2b). Overall, the presence of predator information enhanced female survival, whereas the addition of their own cues reduced survival time and frequency (Fig. 2b). When female cues were presented alone, T. helluo predators captured females in 4.8 ± 1.4 min, but it took them 12.6 ± 2.8 min when no cues were added, and 15.5 ± 1.9 min when predator cues were present with or without additional female information (Fig. 2b).

Table 1 Tests for the effects of the addition of chemotactile cues on T. helluo predatory success when housed with male and female P. milvina in the laboratory Model and factor tested

df

c2

P

Nominal logistic regression for which sex captured Whole model 6 24.21 0.0005 Female cues 2 8.25 0.0161 Predator cues 2 5.84 0.0540 Female cues)predator cues 2 7.89 0.0194 Proportional hazard tests on timing of predation for each sex separately Male death Whole model 3 15.00 0.0018 Female cues 1 3.32 0.0682 Predator cues 1 2.13 0.1442 Female cues)predator cues 1 6.78 0.0092 Female death Whole model 3 20.41 0.0001 Female cues 1 4.59 0.0321 Predator cues 1 11.46 0.0007 Female cues)predator cues 1 0.07 0.7896

DISCUSSION Male and female P. milvina bore different levels of risk during courtship and mating, and the additional information contained in chemotactile cues shifted that risk in ways that affected reproductive success. Over all of our laboratory treatments, T. helluo predators were equally likely to capture males and females. Additional cues, whether they were from females or predators, extended life for males (Fig. 2a). In field enclosures, this differential translated into survival differences (Fig. 3a and b), resulting in male-biased sex ratios in enclosures where T. helluo predators were present along with additional predator Information (Fig. 3c). With more males available as potential mates, females were more likely to reproduce successfully in the added-cue treatment (Fig. 4a). Thus, the different responses of males and females to the environmental information available had the potential to influence the landscape for sexual selection and, ultimately, the success of the mating system. In the laboratory, the addition of any chemical information gave P. milvina males more time before they were attacked (Fig. 2a) and, in the field, added predator cues increased male survival (Fig. 3a). The ‘alerting’ hypothesis posits that an initial stimulus makes recipients more attentive to additional cues, especially those that arrive in other sensory modalities (Rowe, 1999; Stephenson, 2016). Pardosa milvina males use substrate-borne chemical signals to find and identify receptive females (Rypstra et al., 2009, 2003), and because T. helluo predators are also wolf spiders, the cues they deposit in the environment would stimulate the same receptors as the cues of P. milvina females (Rypstra et al., 2016). Upon recognizing signs from either predators or females, P. milvina males would benefit by being more attentive to their surroundings as specific locational information is critical to both predator evasion and mate attraction. Increased surveillance would allow P. milvina males to identify T. helluo predators sooner, even when female cues had elicited their heightened awareness. Hence, these data suggest that mature P. milvina males prioritize substrate-borne chemotactile information and, once it is detected, they scrutinize the area and react appropriately when they discover the precise location of potential mates or predators (Rowe, 1999; Rypstra et al., 2009). It is not surprising that more P. milvina females were captured more quickly when their own cues were included in the arena (Table 1, Fig. 2b). Tigrosa helluo detects prey using these same substrate-borne chemotactile cues (Persons & Rypstra, 2000), thus

170

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

(a) Male survival

(b) Female survival

Proportion alive

1

No cues Female cues Predator cues Both cues

0.8

0.6

1

0.8 A

0.6 A A

0.4

0.4

A

0.2

A

0.2

B

B

A,B 0

5

10

15

20

25

0

30

5

10

15

20

25

30

Time (min) Figure 2. Proportion (±SE) of P. milvina (a) males and (b) females surviving in the presence of the predator, T. helluo, over a 30 min experimental period in the laboratory. Treatments differed in the chemotactile information (none, female P. milvina, T. helluo predator, or both female P. milvina and predator T. helluo) that was added to the arena before the start of the trial. Treatments indicated with the same letter were not significantly different from one another using a familywise error rate of 0.008. Full analyses are presented in Table 1.

Table 2 Results of linear mixed model of field enclosure data Variable

df

Proportion of males surviving Proportion of females surviving Sex ratio (proportion of males out of all surviving) Proportion of females producing eggs Average clutch size

2, 2, 2, 2, 2,

40 40 40 40 37

F

P

10.73 5.62 4.68 18.50 35.88

0.0002 0.0071 0.0150 <0.0001 <0.0001

In each case, treatments included control, predator present and cues added with predator. Month was entered as a random effect. Application of the logit transformation before analysis successfully normalized the proportional data.

the presence of P. milvina female signals likely caused this predator to focus their search efforts. It seems likely that T. helluo use the density and distribution of cues to refine their search area and discover females directly or detect the activity of the courting male and find the observing female easier to catch (Hughes et al., 2012). On the other side of the predatoreprey arms race, when P. milvina

(a) Male survival

(c) Sex ratio

(b) Female survival 1

A

A

1 A

0.8

0.8

0.6

0.6

A B

B 0.4

0.4

0.2

0.2

0

Control

Predator Cues added treatment

0

Proportion male

1

Proportion

females are attempting to assess courting males, they may be less capable of responding to other signals in the environment. In general, effective courtship displays are specifically selected to attract and sustain the female's attention (Dukas, 2004) and, specifically, P. milvina females integrate subtle features of the male's morphology and behaviour in their mating decision (Rypstra et al., 2016, 2003). If the female's attention is primarily directed towards the male's features and performance, then the female's ability to attend to other stimuli, such as an approaching predator, is likely to be diminished. The ‘alerting’ hypothesis might suggest that, upon detection of predator cues, P. milvina females would direct their attention towards isolating and identifying the potential threat. While this might explain the fact that females lived longer when predator cues were present in our laboratory trials (Fig. 2b), it is not consistent with the low survival of females in field enclosures with added predator information (Fig. 3b). Likewise, if the surviving females were distracted by the predator signals in the field enclosures, we

0.8 A 0.6

A,B B

0.4 0.2

Control

Predator Cues added treatment

0

Control

Predator Cues added treatment

Figure 3. Survival of P. milvina (a) males and (b) females, and (c) the final adult sex ratio (as represented by the proportion of all survivors that were male) recovered from field enclosures. Treatments: control (with no T. helluo predators and no predator cues); predator (with two T. helluo predators and no additional cues); cues added (with two T. helluo predators along with additional T. helluo cues). Bars indicated with the same letter were not significantly different from one another according to Tukey's HSD test (P < 0.05). Full analyses are presented in Table 2.

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

(a) Females with eggsacs

(b) Clutch size A

1

171

25 A Average clutch size

Proportion

0.8

B B

0.6

0.4

B B

15

10

5

0.2

0

20

Control

Predator treatment

Cues added

0

Control

Predator treatment

Cues added

Figure 4. Reproductive success of female P. milvina recovered from field enclosures including (a) the proportion that produced eggsacs and (b) the average clutch size of the eggsacs produced. Treatments: control (with no T. helluo predators and no predator cues); predator (with two T. helluo predators and no additional cues); cues added (with two T. helluo predators along with additional T. helluo cues). Bars indicated with the same letter were not significantly different from one another according to Tukey's HSD test (P < 0.05). Full analyses are presented in Table 2.

would have expected them to devote less attention towards courting males and possibly even delay reproduction until they were in a more predator-free situation. However, nearly all of the females that survived in the added-cue treatment produced an eggsac, which verifies that they were still willing to tend to the displaying male and mate successfully. Indeed, past laboratory studies indicate that the presence of T. helluo chemotactile cues alone has no impact on the timing or frequency with which virgin P. milvina males and females mate (Rypstra et al., 2016). Rather than causing females to search for potential predators, T. helluo cues seem to cause P. milvina females to shift their attention away from the conspicuous visual courtship dance, with its commensurate vibratory components, towards subtle morphological features of the males that are highlighted by the courtship displays (Rypstra et al., 2016). The results of that study along with the data reported here suggest that the visual display of males, along with any vibratory signals that the activity produces, is powerful enough to demand the attention of virgin P. milvina females even in the presence of pervasive predator signals. One goal of this study was to understand whether access to information could have sufficiently different impacts on males and females to influence the efficacy of reproductive strategy. The presence of predators lowered population density in field studies, but the presence or absence of additional chemotactile cues affected the sex ratio of P. milvina mating populations. When the cues were present, more females were lost, but nearly all the survivors (>95%) reproduced successfully. This finding is consistent with prevailing theory that assumes increasingly biased sex ratio reduces the reproductive variance for the rarer sex, here females, and imposes stronger sexual selection on the more common sex, here males (Emlen & Oring, 1977; Kokko, Klug, & Jennions, 2012; Shuster & Wade, 2003). Interestingly, when we did not include additional predator cues, males became the rare sex and female survivors were much less likely to produce an eggsac (only 57%); this low success rate suggests that the more common females were experiencing more intense sexual selection pressure and were possibly competing for available males. An important aspect of this study is that, not only the presence of predators, but also the nature of information available about those predators while the potential prey were engaged in courtship and mating altered the opportunity

for sexual selection (Jones, 2009). Indeed, the reproductive advantage appeared to shift away from females when male mortality was higher and differential mortality was driven by subtle differences in the manner in which each of the animals involved responded to environmental stimuli. Conclusion
kely, Weissing et al. (2014) argued for a more dynamic view Sze of breeding system evolution that explicitly incorporates ecological feedbacks and helps generate predictions regarding the evolutionary trajectory of populations. We generally assume that males with conspicuous displays face the greatest risk and, indeed, that has been documented for the species under study here (Hoefler et al., 2008). However, here we show that the actual risk is dependent on the nature of the information available as well as the attentional priorities of the males, the females they are attempting to attract, and any nearby predators. In addition, the data from our field enclosures suggest that these factors are sufficient to affect the adult sex ratio and the likelihood that females will successfully reproduce. In this way, environmental information can potentially affect the relative importance of natural versus sexual selection in driving the mating system and reproductive success of this species. Our data underscore the intricate connection between environmental features and eavesdropping predators that influences the success of the breeding system and mating dynamics. Further study to explicitly connect attentional abilities of males and females with the sensory information they require to accurately assess mates and detect predators will be key to understanding the risks that they face and shed light on the long-term processes that direct the balancing act between conspicuousness and defence. Acknowledgments We are indebted to K. Carter, J. Cheek, S. C. Evans, E. Hetzel, S. A. Nagy and M. Yazdani for assistance with field and laboratory work. Fields at the Miami University's Ecology Research Center were prepared by Rodney J. Kolb. This manuscript benefited from comments from A. Davis, L. Erickson, J. Godfrey, J. C. Johnson, L. Latham, C. Lawson, A. Singer, M. Stanley and several anonymous referees.

172

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173

Funding was provided by the National Science Foundation (grants DBI 0216776, DBI 0116947), as well as Miami University's Ecology Research Center, Department of Zoology and Hamilton Campus.

References Anderson, J. F. (1974). Responses to starvation in spiders Lycosa lenta Hentz and Filistata hibernalis (Hentz). Ecology, 55, 576e585. http://dx.doi.org/10.2307/ 1935148. Bell, R. D., Rypstra, A. L., & Persons, M. H. (2006). The effect of predator hunger on chemically-mediated antipredator responses and survival in the wolf spider Pardosa milvina (Araneae: Lycosidae). Ethology, 112, 903e910. http://dx.doi.org/ 10.1111/j.1439-0310.2006.01244.x. Blumstein, D. T. (2014). Attention, habituation, and antipredator behavior: Implications for urban birds. In D. Gil, & H. Brumm (Eds.), Avian urban ecology (pp. 41e53). Oxford, U.K.: Oxford University Press. Bro-Jørgensen, J. (2010). Dynamics of multiple signaling systems: Animal communication in a world in flux. Trends in Ecology & Evolution, 25, 292e300. http:// dx.doi.org/10.1016/j.tree.2009.11.003. Brown, G. E., Poirier, J.-F., & Adrian, J. C. (2004). Assessment of local predation risk: The role of sublethal concentrations of chemical alarm cues. Behavioral Ecology, 15, 810e815. http://dx.doi.org/10.1093/beheco/arh084. Clark, D. L., Zeeff, C. K., Karson, A., Roberts, J. A., & Uetz, G. W. (2016). Risky courtship: Background contrast, ornamentation, and display behavior of wolf spiders affect visual detection by tad predators. Ethology, 122, 364e375. http:// dx.doi.org/10.1111/eth.12476. Costantini, D., Bruner, E., Fanfani, A., & Dell’Omo, G. (2007). Male-biased predation of western green lizards by Eurasian kestrels. Naturwissenschaften, 94, 1015e1020. http://dx.doi.org/10.1007/s00114-007-0284-5. Driver, J., & Spence, C. (2004). Crossmodal spatial attention: Evidence from human performance. In C. Spence, & J. Driver (Eds.), Crossmodal space and crossmodal attention (pp. 179e220). Oxford, U.K.: Oxford University Press. Dukas, R. (2002). Behavioral and ecological consequences of limited attention. Philosophical Transactions of the Royal Society B: Biological Sciences, 357, 1539e1547. http://dx.doi.org/10.1098/rstb.2002.1063. Dukas, R. (2004). Causes and consequences of limited attention. Brain, Behavior and Evolution, 63, 197e210. http://dx.doi.org/10.1159/000076781. Emlen, S., & Oring, L. W. (1977). Ecology, sexual selection, and the evolution of mating systems. Science, 197, 215e223. http://dx.doi.org/10.1126/ science.327542. Fitze, P. S., & Le Galliard, J. F. (2008). Operational sex ratios, sexual conflict and the intensity of sexual selection. Ecology Letters, 11, 432e439. http://dx.doi.org/ 10.1111/j.1461-0248.2008.01158.x. Garcia-Berthou, E. (2001). On the misuse of residuals in ecology: Testing regression residuals vs. the analysis of covariance. Journal of Animal Ecology, 70, 708e711. http://dx.doi.org/10.1046/j.1365-2656.2001.00524.x. Hartman, E. J., & Abrahams, M. V. (2000). Sensory compensation and the detection of predators: The interaction between chemical and visual information. Proceedings of the Royal Society B: Biological Sciences, 267, 571e575. http:// dx.doi.org/10.1098/rspb.2000.1039. Haynes, W. (2013). Bonferroni correction. In W. Dubitsky, O. Wolkenhauer, K.H. Cho, & H. Yokota (Eds.), Encyclopedia of systems biology (p. 154). New York, NY: Springer Science and Business Media. Hebets, E. A. (2011). Current status and future directions of research in complex signaling. Current Zoology, 57, 175e186. Hebets, E. A., & Papaj, D. R. (2005). Complex signal function: Developing a framework of testable hypotheses. Behavioral Ecology and Sociobiology, 57, 197e214. http://dx.doi.org/10.1007/s00265-004-0865-7. Higham, J. P., & Hebets, E. A. (2013). An introduction to multimodal communication. Behavioral Ecology and Sociobiology, 67, 1381e1388. http://dx.doi.org/10.1007/ s00265-013-1590-x. Hoefler, C. D., Persons, M. H., & Rypstra, A. L. (2008). Evolutionarily costly courtship displays in a wolf spider: A test of viability indicator theory. Behavioral Ecology, 19, 974e979. http://dx.doi.org/10.1093/beheco/arn055. Hughes, N. K., Kelley, J. L., & Banks, P. B. (2009). Receiving behaviour is sensitive to risks from eavesdropping predators. Oecologia, 160, 609e617. http://dx.doi.org/ 10.1007/s00442-009-1320-2. Hughes, N. K., Kelley, J. L., & Banks, P. B. (2012). Dangerous liaisons: The predation risks of receiving social signals. Ecology Letters, 15, 1326e1339. http://dx.doi.org/ 10.1111/j.1461-0248.2012.01856. Jones, A. G. (2009). On the opportunity for sexual selection, the Bateman gradient and the maximum intensity of sexual selection. Evolution, 63, 1673e1684. http://dx.doi.org/10.1111/j.1558-5646.2009.00664.x. Kokko, H., & Jennions, M. D. (2008). Parental investment, sexual selection and sex ratios. Journal of Evolutionary Biology, 21, 919e948. http://dx.doi.org/10.1111/ j.1420-9101.2008.01540.x. Kokko, H., Klug, H., & Jennions, M. D. (2012). Unifying cornerstones of sexual selection: Operational sex ratio, Bateman gradient and the scope for competitive investment. Ecology Letters, 15, 1340e1351. http://dx.doi.org/10.1111/j.14610248.2012.01859.x. Lehmann, L., Walker, S. E., & Persons, M. H. (2004). The influence of predator sex on chemically-mediated antipredator response in the wolf spider Pardosa milvina

(Araneae: Lycosidae). Ethology, 110, 1e17. http://dx.doi.org/10.1111/j.14390310.2004.00972.x.
kely, T. (2013). The evolution of sex roles in birds is Liker, A., Freckleton, R. P., & Sze related to adult sex ratio. Nature Communications, 4, 1587. http://dx.doi.org/ 10.1038/ncomms2600. Munoz, N. E., & Blumstein, D. T. (2012). Multisensory perception in uncertain environments. Behavioral Ecology, 23, 457e462. http://dx.doi.org/10.1093/beheco/ arr220. Partan, S., & Marler, P. (1999). Behavior: Communication goes multimodal. Science, 283, 1272e1273. http://dx.doi.org/10.1126/science.283.5406.1272. Partan, S., & Marler, P. (2005). Issues in the classification of multimodal communication signals. American Naturalist, 166, 231e245. http://dx.doi.org/10.1086/ 431246. Persons, M. H. (1999). Hunger effects on the foraging responses to perceptual cues in immature and adult wolf spiders (Lycosidae). Animal Behaviour, 57, 81e88. Persons, M. H., & Rypstra, A. L. (2000). Preference for chemical cues associated with recent prey in the wolf spider Hogna helluo (Araneae: Lycosidae). Ethology, 106, 27e35. http://dx.doi.org/10.1046/j.1439-0310.2000.00496.x. Persons, M. H., & Rypstra, A. L. (2001). Wolf spiders show graded antipredator behavior in the presence of chemical cues from different sized predators. Journal of Chemical Ecology, 27, 2493e2504. http://dx.doi.org/10.1023/A: 1013679532070. Persons, M. H., & Uetz, G. W. (1996). The influence of sensory information on patch residence times in wolf spiders. Animal Behaviour, 51, 1285e1293. Persons, M. H., Walker, S. E., & Rypstra, A. L. (2002). Fitness costs and benefits of antipredator behavior mediated by chemotactile cues in the wolf spider Pardosa milvina (Araneae: Lycosidae). Behavioral Ecology, 13, 386e392. http://dx.doi.org/ 10.1093/beheco/13.3.386. Persons, M. H., Walker, S. E., Rypstra, A. L., & Marshall, S. D. (2001). Wolf spider predator avoidance tactics and survival in the presence of diet-associated predator cues (Araneae: Lycosidae). Animal Behaviour, 61, 43e51. http:// dx.doi.org/10.1006/anbe.2000.1594. Pocklington, R., & Dill, L. M. (1995). Predation on females or males: Who pays for bright male traits? Animal Behaviour, 49, 1122e1124. http://dx.doi.org/10.1006/ anbe.1995.0141. Roberts, J. A., Taylor, P. W., & Uetz, G. W. (2007). Consequences of complex signalling: Predator detection of multimodal cues. Behavioral Ecology, 18, 236e240. http://dx.doi.org/10.1093/beheco/ar1079. Rowe, C. (1999). Receiver psychology and the evolution of multicomponent signals. Animal Behaviour, 58, 921e931. http://dx.doi.org/10.1006/anbe.1999.1242. Rypstra, A. L., Schlosser, A. M., Sutton, P. L., & Persons, M. H. (2009). Multimodal signaling: The relative importance of chemical and visual cues from females to the behaviour of male wolf spiders (Lycosidae). Animal Behaviour, 77, 937e947. http://dx.doi.org/10.1016/j.anbehav.2008.12.026. Rypstra, A. L., Walker, S. E., & Persons, M. H. (2016). Cautious versus desperado males: Predation risk affects courtship intensity but not female choice in a wolf spider. Behavioral Ecology, 27, 876e885. http://dx.doi.org/10.1093/beheco/ arv234. Rypstra, A. L., Wieg, C., Walker, S. E., & Persons, M. H. (2003). Mutual mate assessment in wolf spiders: Differences in cues used by males and females. Ethology, 109, 315e325. http://dx.doi.org/10.1046/j.1439-0310.2003.00874.x. Sakaluk, S. K., & Belwood, J. J. (1984). Gecko phonotaxis to cricket calling song: A case of satellite predation. Animal Behaviour, 32, 659e662. Schlosser, A. M. (2005). The effects of female body condition, female cue and predator cue presence on the locomotive and reproductive behaviour of the male wolf spider, Pardosa milvina (Araneae; Lycosidae) (Master's thesis). Oxford, OH: Miami University. Schmidt, K. A., Dall, S. R., & van Gils, J. A. (2010). The ecology of information: An overview on the ecological significance of making informed decisions. Oikos, 119, 245e253. http://dx.doi.org/10.1111/j.1600-0706.2009.17824.x. Searcy, L. E., Rypstra, A. L., & Persons, M. H. (1999). Airborne chemical communication in the wolf spider, Pardosa milvina. Journal of Chemical Ecology, 25, 2527e2533. http://dx.doi.org/10.1023/A:1020878225553. Shonewolf, K. W., Bell, R., Rypstra, A. L., & Persons, M. H. (2006). Field evidence of an airborne enemy-avoidance kairomone in wolf spiders. Journal of Chemical Ecology, 32, 1565e1576. http://dx.doi.org/10.1007/s10886-006-9070-7. Shuster, S. M., & Wade, J. M. (2003). Mating systems and strategies. Princeton, NJ: Princeton University Press. €fstedt, C., & Surlykke, A. (2005). Her Skals, N., Anderson, P., Kanneworff, M., Lo odours make him deaf: Crossmodal modulation of olfaction and hearing in a male moth. Journal of Experimental Biology, 208, 595e601. http://dx.doi.org/ 10.1242/jeb.01400. Stephenson, J. (2016). Keeping eyes peeled: Guppies exposed to chemical alarm cue are more responsive to ambiguous visual cues. Behavioral Ecology and Sociobiology, 70. http://dx.doi.org/10.1007/s00265-016-2076-4, 757e584.
kely, T., Liker, A., Freckleton, R. P., Fichtel, C., & Kappeler, P. M. (2014). Sex-biased Sze survival predicts adult sex ratio variation in wild birds. Proceedings of the Royal Society B: Biological Science, 281. http://dx.doi.org/10.1098/rspb.2014.0342, 20140342.
kely, T., Weissing, F. J., & Komdeur, J. (2014). Adult sex ratio variation: ImplicaSze tions for breeding system evolution. Journal of Evolutionary Biology, 27, 1500e1512. http://dx.doi.org/10.1111/jeb.12415. Uetz, G. W. (2000). Signals and multimodal signalling in spider communication. In Y. Espmark, T. Amundsen, & G. Rosenqvist (Eds.), Animal signalling and signal

A. L. Rypstra et al. / Animal Behaviour 128 (2017) 165e173 design in animal communication (pp. 387e405). Trondheim, Norway: Tapir Press. Uetz, G. W., & Roberts, J. A. (2002). Multisensory cues and multimodal communication in spiders: Insights from video/audio playback studies. Brain, Behavior and Evolution, 59, 222e230. http://dx.doi.org/10.1159/000064909. Walker, S. E., & Rypstra, A. L. (2003). Sexual dimorphism and the differential mortality model. Is behaviour related to survival? Biological Journal of the Linnaean Society, 78, 97e103. http://dx.doi.org/10.1046/j.1095-8312.2003.00134.x.

173

Warton, D. I., & Hui, F. K. (2011). The arcsine is asinine: The analysis of proportions in ecology. Ecology, 92, 3e10. http://dx.doi.org/10.1890/10-0340.1. Wisenden, B. D., Vollbrecht, K. A., & Brown, J. L. (2004). Is there a fish alarm cue? Affirming evidence from a wild study. Animal Behaviour, 67, 59e67. http:// dx.doi.org/10.1016/j.anbehav.2003.02.010. Zuk, M., & Kolluru, G. R. (1998). Exploitation of sexual signals by predators and parasitoids. Quarterly Review of Biology, 73, 415e438. http://dx.doi.org/10.1086/ 420412.