Printer Friendly

Meral-Spot Reflectance Signals Weapon Performance in the Mantis Shrimp Neogonodactylus oerstedii (Stomatopoda).

Introduction

Animals commonly compete over resources such as food, territories, or mates. In these situations, one individual will receive a benefit from winning, and one will experience a cost from losing (Maynard Smith, 1979; Searcy and Nowicki, 2005); but both may experience a cost if they are injured during the competition. Therefore, they each benefit from accurately assessing their opponent and resolving a contest without physical combat. This could be achieved by first signaling their fighting ability, commonly termed their "resource holding potential" (Parker, 1974) or their aggressive intent (Enquist, 1985). It might seem that the signaler would benefit from deceiving an opponent by signaling greater fighting ability or aggressive intent than it possesses. Conversely, a receiver will benefit from responding to the signal only if that signal is honest. If cheating became sufficiently common, receivers would no longer benefit from responding to the signal, and the signaling system would no longer function (Maynard Smith, 1979; Searcy and Nowicki, 2005). Thus, signaling theory predicts that these signals must be "honest on average" to persist in the population (Maynard Smith and Parker, 1976; Johnstone and Graf en, 1993; Searcy and Nowicki, 2005).

Color is commonly used as a signal of resource holding potential in agonistic encounters throughout the animal kingdom, including animals such as birds (Pryke et al., 2001; Siefferman and Hill, 2005), reptiles (Whiting et al, 2006; Ligon and McGraw, 2016), fish (Gerlach et al, 2014; Kelley et al., 2016), and cephalopods (Adamo and Hanlon, 1996; Scheel et al., 2016). In these cases, the honesty of the signal is maintained because producing the color is costly, due to physiological constraints (Lozano, 2001), social costs (Ligon and McGraw, 2016), increased predation risk (Moodie, 1972), or nutritional or energetic limitations (Frischknecht, 1993). For example, many red or orange patches are produced from carotenoid pigments. These carotenoids must be ingested, and in addition to creating colored patches, they are also required for immune function or other physiological functions (Britton, 2008; Svensson and Wong, 2011). Thus, carotenoids are limited, and animals must resolve a trade-off between creating a deeply colored patch and maintaining adequate immunity. Although there are many studies investigating color signal honesty in fish (Candolin, 2000; Balzarini et al., 2016) and in terrestrial animals including birds (Pryke et al., 2001; McGraw, 2007), reptiles (Whiting et al., 2006; Ligon and McGraw, 2016), and insects (Lindstedt et al., 2010; Blount et al., 2012), there are few examples in aquatic invertebrates (but see Cortesi and Cheney, 2010). Such information about agonistic signals is essential to understand whether color signaling theory extends across more taxa than are currently studied.

One aquatic invertebrate group that likely uses color signals is the stomatopod group (mantis shrimps). Previous work has shown that they probably use color signals during contests for possession of refuges (Caldwell and Dingle, 1975, 1976; Franklin et al., 2016, 2017). Refuges, such as holes in coral rubble or rock, are required to process food, escape predators, mate, and brood eggs (Caldwell, 1987). During contests, stomatopods often strike their opponent with their raptorial appendages. Stomatopod strikes are extremely powerful and fast, generating forces thousands of times the stomatopod's body weight (Patek et al., 2004; Patek and Caldwell, 2005); and these strikes may injure or even kill an opponent (Caldwell and Dingle, 1975; Berzins and Caldwell, 1983). Before physical blows, stomatopods often perform a threat display called a "meral spread" (Fig. 1 A; Dingle, 1969). This posture shows the color of a patch, called the meral spot, which seems to indicate fighting ability or aggressive intent (Caldwell and Dingle, 1975, 1976; Franklin et al., 2016, 2017).

Despite the high cost of fighting, we do not know whether stomatopods first signal fighting ability to opponents. Indeed, Green and Patek (2015) found that the meral spread was not used to avoid escalation in size-matched pairs of the stomatopod Neogonodactylus bredini. They found that nearly all conflicts between individuals of similar size and fighting ability escalated to strikes, even after a meral spread. Moreover, the winners of a contest did not have the strongest strike; instead, they struck the loser more times (Green and Patek, 2015). However, N. bredini has a white meral spot, not a colored patch. Other stomatopod species, such as Neogonodactylus oerstedii (studied here; Hansen, 1895), have a colored meral spot that may signal resource holding potential (Franklin et al., 2016, 2017).

Stomatopod species with colored patches are known to use color signals during contests over refuges (Franklin et al., 2016, 2017), but the information content of these signals remains a mystery. Recent research that manipulated signal color in N. oerstedii demonstrated that receivers increase agonistic behaviors toward opponents with lighter-colored spots (i.e., increased total reflectance) or toward those with decreased ultraviolet (UV) reflectance (Franklin et al., 2016, 2017). Thus, behaviorally, it appears that the meral spot color signals fighting ability or aggressive intent; but does it represent an honest signal?

To investigate signal honesty, we determined whether aspects of the signal correlate with the whole organism's performance Whole-organism performance refers to an animal's ability to conduct an ecologically relevant task (e.g., sprinting rapidly or biting forcefully), as opposed to measurements of physiological effects at lower functional levels (e.g., producing enzymes; Irschick et al., 2008; Lailvaux and Husak, 2014). For example, higher bite force in collared lizards is correlated with components of the gape display and with blue coloration (Lappin et al., 2006; Plasman et al., 2015). Many other studies across different taxa have finked whole-organism performance with components of threat signals (Lailvaux et al., 2005; Lappin et al., 2006; Bywater et al., 2008; Henningsen and Irschick, 2012; Mowles and Briffa, 2012). Although whole-organism performance is not the same as fighting ability (i.e., the animal's ability to win a contest), studies have demonstrated a link between whole-organism performance, including weapon performance, and contest outcome (e.g., crayfish, Bywater et al., 2008; hermit crabs, Mowles et al., 2010).

To investigate the relationship between components of threat signals and performance, it is essential to consider the reliability of performance measurements. For example, recordings of lizard bite force are affected by the measuring device, where in the mouth the bite is applied, low sample size, lizard temperature, and lizard motivation (Losos et al., 2002; Anderson et al., 2008). We attempted to minimize these errors by using reliable equipment, strict requirements for usable measurements, a minimum number of measurements per individual, and constant experimental conditions. Despite these requirements, maximum measures of performance can still be affected by animal motivation and by variable numbers of recordings per individual (Losos et al., 2002; Hagey et al., 2016). To overcome these issues, we modeled each stomatopod's performance using the Weibull distribution, instead of using the maximum measurement recorded (following the procedure in Hagey et al., 2016).

As our measurement of whole-organism performance, we recorded peak force (N) and impulse (N*s) from stomatopod strikes (following Patek and Caldwell, 2005). The peak force is related to the stress in the material that the stomatopod is striking, which then contributes to deformation and, potentially, failure. It is also related to the material properties of the material; the very stiff steel sensor used here records very high peak forces for a very short duration, but a softer material would have a smaller peak force absorbed over a longer duration (Patek and Caldwell, 2005). The impulse, the integral of force through time, helps to compare these two cases. Tests on circular steel plates showed that deformation and failure of the plate are linearly proportional to the applied impulse, not the peak force (Florence, 1966; Corbett et al., 1996). Thus, a large force over a short time or a small force over a long time may have the same impulse and the same potential to damage the object that the stomatopod is striking.

We used these measures of whole-organism performance to investigate meral spot signal honesty in TV. oerstedii. Meral spot color in N. oerstedii is likely based on carotenoproteins (Newbigin, 1897; Zagalsky et al., 1970; Bandaranayake, 2006), which are carotenoids that have formed chemical complexes with proteins. Because carotenoids must be ingested, a darker meral spot that would require more pigment may indicate a stomatopod that has had more success foraging and that is in better condition. Therefore, we predicted that strike force and strike impulse would increase as the darkness of the meral spot increases.

Materials and Methods

Stomatopod husbandry

Stomatopods (Neogonodactylus oerstedii (Hansen, 1895)) were ordered commercially (KB Marine Life, Big Pine Key, FL) and shipped overnight to Tufts University, Medford, Massachusetts. They were housed under a 12h : 12h light: dark cycle in artificial seawater maintained at 33-35 ppt salinity and 22-25 [degrees]C. Aquaria (75 L) were divided into 3 compartments, with 1 stomatopod housed per compartment. Stomatopods could not see other individuals. Water was circulated through a carbon filter and flowed through all three compartments. Each stomatopod was provided with a 3-inch section of 0.5-inch polyvinyl chloride tube as a refuge. Twice per week stomatopods were fed pollock, and one-third of the water was changed. When stomatopods were to be tested, they were transferred in their refuge to a 1-L bucket 3 days before force measurements were to be recorded. These buckets were maintained at 22 [degrees]C and aerated with air pumped through an air stone, and the twice weekly water changes continued. The 1-L buckets let us transfer buckets to the testing area with minimal disturbance to the stomatopods. Stomatopods were housed in captivity for at least one month to allow for acclimation and to ensure that we were testing stomatopods that had not recently molted (i.e., intermolt). To avoid any physiological deterioration due to time in captivity, we only analyzed data from stomatopods that were housed in the lab for fewer than 100 days.

Force measurements

To measure strike force, we used a one-axis force sensor (model 200B02, PCB Piezotronics, Depew, NY), which was placed near the entrance to the refuge (Fig. 1C). The sensor had one layer of yellow Octavia laminating film (10 [micro]m thick, GBC, Lake Zurich, IL) stuck to it, because this helped to elicit a strike. To further encourage a stomatopod to strike, we put shrimp paste on the sensor. If the shrimp paste did not work, we also waved a thin strip of yellow laminating film in between the sensor and the refuge. This plastic decoy was removed before the strike, so the stomatopod did not strike through the decoy. Data were recorded at 500 kHz, using Lab View (ver. 15.0.1, National Instruments, Austin, TX) with a USB-6343 data acquisition system (National Instruments). From these force recordings, we first subtracted a baseline reading, averaged in the window from 500 [micro]s to 10 [micro]s before the strike. Then we calculated strike force (the amplitude of the first peak relative to the baseline) and impulse (the integral of force over time from 50 [micro]s before the first peak, up to the time of the first peak), using custom MATLAB code (vR2014a, Mathworks, Natick, MA; Fig. ID). Each recording had oscillations due to the resonant frequency of the sensor (Fig. 1D), but it was not possible to get sensors with higher resonant frequencies. All trials were conducted under bright lights and were filmed at 16,000 frames per second, 60 [micro]s exposure time, and 320 x 240 resolution, using a Phantom Miro M120 camera (Vision Research, Wayne, NJ). From the video footage, we recorded whether the stomatopod struck with 1 or 2 appendages, whether the strike was near the center of the sensor (> ~1 mm from the edge), and whether the strike was perpendicular to the sensor (i.e., did not glance across the sensor; Video 1, available online). The video footage was synchronized to the force recording to distinguish force peaks from each appendage and from cavitation. Strikes were used in subsequent analysis only if they were perpendicular, near the center of the sensor, and if we could calculate the force from one appendage only (i.e., force peaks matching the video footage were discernible in the output for each appendage and/or cavitation). When stomatopods struck with two appendages, only data from the first appendage that contacted the sensor were used. Three stomatopods were excluded from analysis because we did not obtain five usable strikes. We obtained 5-11 suitable strikes from 4 males and 14 females.

Meral spot color and size

After strike measurements were recorded, we recorded the spectral reflectance and size of each of the stomatopod's meral spots. Crustacean colors can change dramatically after a molt (Wade et al., 2005), and meral spot total reflectance is greater immediately after a molt (Reaka, 1975; Franklin et al., 2017). Thus, to ensure that we were recording intermolt meral spot color similar to wild populations, only animals that had not molted since arriving in the laboratory (i.e., for at least one month) were included in the analysis. No stomatopods molted within the week after testing procedures concluded. To record meral spot measurements, live stomatopods were anesthetized by cooling them in the freezer (-13 [degrees]C) until they stopped moving. This took about 20 min in 25 mL of seawater. Spectral measurements were recorded using a JAZ spectrophotometer (Ocean Optics, Largo, FL) with a pulsed xenon light source. Reflectance was recorded between 300 and 750 nm and was measured relative to a WS-1 white standard. The light source and collecting probes were 600 [micro]m UV-visible fiber-optic cables (Ocean Optics) with collimating lenses attached to the end. Both were fixed 20 mm from the meral spot, in a position that mimicked natural lighting conditions (Endler, 1990), with the light source at 45[degrees] to the meral spot surface and the collecting probe perpendicular to it (Fig. 1B). The collecting probe was focused to a diameter of 1 mm, which is smaller than the diameter of meral spots of stomatopods used in this study. In a petri dish, stomatopods were placed on one side, and the lower raptorial appendage was pulled posteriorly and pinned so that the meral spot was exposed and facing up. Two measurements each were taken of both the left and right meral spot in air. Meral spot color does not vary between left and right meral spots (Franklin et al., 2016); therefore, these four spectra were averaged to obtain one spectral measurement for each stomatopod. While stomatopods were in this position, we used a Dino-Lite 5.0-MP USB microscope (New Taipei City, Taiwan) to take two photos of each meral spot. Meral spots in N. oerstedii consist of a purple patch surrounded by a white ring (Fig. 1A). From these photos, we measured the area of the entire meral spot and just the purple patch, using Image J (https://imagej.nih.gov/ij). We then calculated the proportion of the meral spot that was pigmented and averaged to obtain one measurement for each stomatopod.

Statistical analysis

To analyze the force data, we first investigated variation in strike force and strike impulse. We determined whether these parameters were correlated by using Pearson's moment correlation coefficient. Then, for both strike force and strike impulse, we investigated whether there was a difference between experimental days (i.e., duration in lab), between strikes within days, or between one-appendage and two-appendage strikes. Linear mixed models were used, with day number, strike number (within day), and appendage number as predictor variables and stomatopod ID as a random effect. The response variable was either strike force (N) or impulse (N*s), and both were log transformed to meet the homoscedasticity assumption of a linear model. To investigate whether the force or impulse of one strike influenced the subsequent strike, we plotted autocorrelation function estimates. Marginal hypothesis tests using the chi-square statistic were conducted to assess the significance of each term.

We then used a Weibull distribution to estimate peak force and impulse from all strikes recorded from an individual stomatopod (Hagey et al., 2016). Because we did not obtain the same number of strikes for every stomatopod, using the maximum force or the maximum impulse recorded for each stomatopod was a biased estimate (Head et al., 2012). The Weibull method allowed us to account for unequal sample sizes and downweighted any extreme values that might influence estimates (Hagey et al., 2016). For each individual, we fitted a Weibull distribution to its strike force recordings and used the scale parameter of this distribution as an estimate of peak force (Hagey et al., 2016). We followed the same procedure to estimate impulse for each stomatopod. From here on, we will refer to these Weibull estimates as peak force or impulse.

From the raw spectral reflectance data (Fig. 2), we obtained traditional measures of total reflectance and chroma. Total reflectance (%) was calculated as the mean reflectance for the interval 300-750 nm ([R.sub.avg]; Delhey et al., 2003), and chroma was calculated as ([R.sub.300] [lambda]([R.sub.50]) - [R.sub.[lambda]]([R.sub.50])-750)/[R.sub.300-750], where [lambda]([R.sub.50]) is the wavelength at which the reflectance is halfway between the maximum and minimum reflectances (Andersson et al., 2002; Whiting et al., 2006). The spectral range of 300-750 nm corresponds to the visual sensitivity of N. oerstedii (Marshall and Oberwinkler, 1999; Franklin et al., 2016). Hue (wavelength of peak reflectance) is also commonly measured in studies investigating color signals. However, our previous research suggests that total reflectance and/or chroma are more likely to be used as a signal (Franklin et al., 2017). Thus, we did not include hue, because we did not want to overparameterize the statistical model. There was no correlation between time in captivity and meral spot total reflectance or chroma (total reflectance: [F.sub.1,16] = 3.8, P = 0.070; chroma: [F.sub.1,16] = 2.0, P = 0.18). Total reflectance and chroma also did not vary significantly between the sexes (total reflectance: [F.sub.1,16] = 0.65, P = 0.43; chroma: [F.sub.1,16] = 0.06, P = 0.81). Therefore, we did not include sex in any subsequent analyses, because of the small number of males in the sample (4 of 18).

Linear models were used to investigate the relationship between performance estimates (peak force and impulse), meral spot total reflectance or chroma, body length, and duration of captivity. The proportion of the meral spot that was pigmented was also initially included as a predictor; however, this parameter negatively correlated with total reflectance ([F.sub.1,16] = 5.2, P = 0.037, [r.sup.2] = 0.25) and was not correlated with peak force ([chi square] = 0.05, df = 1, P = 0.81) or impulse ([chi square] = 0.04, df = 1, P = 0.85), and so it was removed to avoid multi-collinearity. We calculated variance inflation factors for the remaining predictor variables, and all were less than 2, indicating no further collinearity issues (Zuur et al., 2010). Strike force in stomatopods has been shown to increase with overall body size as well as with measurements of raptorial appendage size (Patek and Caldwell, 2005; Zack et al., 2009; Claverie et al., 2011). Following Claverie et al. (2011), we elected to include body length as a predictor in our models, rather than scale our peak force or impulse measurements. We used a lognormal probability error distribution for both peak force and impulse models because quantile-quantile plots, residual plots, and the Akaike information criterion suggested that this distribution fitted the data best. To investigate the influence of each term, marginal hypothesis tests were conducted using the chi-squared statistic to compare the fit of the full model with that of the nested model (without one term). For significant terms, we also calculated a pseudo-[R.sup.2] as [R.sup.2] = 1 - ([D.sub.A]/[D.sub.0]), where [D.sub.0] is the deviance of the null model and [D.sub.A] is the deviance of the model including the parameter of interest (Cameron and Windmeijer, 1997; Coxe et al., 2013). The coefficient of determination ([r.sup.2]) is not calculated because generalized linear models use a maximum likelihood estimator rather than ordinary least squares regression. Pseudo-[R.sup.2] is a measure of improvement in model fit, with the addition of the parameter of interest, and it is not directly comparable to [r.sup.2]. Following this, we conducted a post hoc test investigating whether total reflectance values differed between stomatopods that strike primarily with one or two appendages. All analyses were conducted in R 3.3.1 (R Core Team, 2014). The survival package (Therneau and Grambsch, 2000) was used to fit Weibull distributions, the lme4 package (Bates et al, 2015) to fit the linear mixed models, and the Anova function in the car package for marginal hypothesis tests (Fox and Weisberg, 2011).

Results

We found a strong correlation between strike force and strike impulse (t = 12.3, df = 162, P < 0.001, [r.sup.2] = 0.48; Fig. 3). However, the variation within this correlation was large (Fig. 3), so we elected to conduct all analyses on both strike impulse and strike force. Neither parameter was affected by experimental day number (strike impulse: [chi square] (=) 1.71, df = 1, P = 0.19; strike force: [chi square] - 2.69, df = 1, P = 0.10) or strike number within day (strike impulse: [chi square] = 0.01, df = 1, P = 0.93; strike force: [chi square] = 0.68, df = 1, P = 0.41; Table 1). Furthermore, there was no autocorrelation in the residuals, indicating that the force or impulse of one strike did not influence the next strike. Strike force was not influenced by whether the stomatopod struck with one or two appendages ([chi square] = 0.37, df = 1, P = 0.54); but one-appendage strikes had lower strike impulse than two-appendage strikes, even for strikes from the same individual ([chi square] = 10.91, df = 1, P < 0.001; Table 1).

Stomatopods with darker meral spots struck with a greater impulse, and there was a trend toward these individuals also striking with greater peak force (Fig. 4; impulse: [chi square] (=) 5.44, df = 1, P = 0.020, pseudo-[R.sup.2] = 0.17; peak force: [chi square] = 3.75, df = 1, P = 0.053, pseudo-[R.sup.2] = 0.20; Fig. 4; Table 2). Impulse and peak force were also positively correlated with stomatopod body length (impulse: [chi square] = 10.28, df = 1, P = 0.001, pseudo-[R.sup.2] = 0.20; peak force: [chi square] = 5.36, df = 1, P = 0.021, pseudo-[R.sup.2] = 0.10; Table 2; Fig. Al). Peak force also negatively correlated with duration of captivity ([chi square] = 4.01, df = 1,P = 0.045, pseudo-[R.sup.2] = 0.30; Fig. A2), whereas impulse did not correlate with duration of captivity ([chi square] = 2.32, df = 1, P = 0.13). Neither impulse nor peak force correlated with chroma (impulse: [chi square] (=) 0.015, df = 1, P = 0.90; peak force: [chi square] = 0.002, df = 1, P = 0.96; Table 2). Stomatopod length had the largest effect in our model, accounting for a change of 120.1 N and 2.1 mN*s across the range of stomatopods we measured (37.4-53.6 mm). Meral spot total reflectance had the second-largest effect, accounting for a change of 106.3 N force and 1.4 mN*s over the range of spectral intensities measured. On average, strike force also decreased by 101.2 N over the testing period, but impulse did not change significantly.

When stomatopods struck with both appendages, the appendages usually did not contact the sensor simultaneously. The results reported above indicated that for two-appendage strikes, the first appendage produced a larger strike impulse than that from a one-appendage strike. Therefore, we conducted a post hoc analysis to investigate whether mean total reflectance correlated with the number of appendages used in the strike. Stomatopods that struck with two appendages more often had darker meral spots than those that struck with one appendage more often ([F.sub.1,143] = 5.59, df = 1, P = 0.019).

Discussion

Here we demonstrate for the first time that stomatopod weapon performance correlates with the total reflectance of stomatopods' meral spots. Stomatopods with darker meral spots had higher strike impulse, and there was a trend toward higher strike force. If strike impulse or force predicts competitive success, these differences suggest that meral spot total reflectance signals fighting ability during agonistic encounters. Moreover, stomatopods with darker spots tended to strike more often with both appendages; and the strike impulse of the first appendage in a two-appendage strike tended to be higher than the strike impulse of a single-appendage strike, even for strikes from the same individual. Thus, stomatopods with darker meral spots tend to invest more metabolic energy in each strike than do stomatopods with lighter meral spots. Together, these results suggest that the total reflectance of the meral spot provides an honest signal of opponent weapon performance.

There is evidence that stomatopods may use this information to assess opponents in contests. They are known to assess total reflectance of the meral spot during territorial contests (Franklin et al., 2017). Franklin and colleagues experimentally increased total reflectance of the meral spot by using an ultrafast laser. The range in meral spot total reflectance of both treatment and control stomatopods was similar to the range reported here (current study: 2.3%-7.8%; 2017 study control: 1.8%-4.6%; 2017 study treatment: 3.7%-5.8%, AMF, unpubl. data), suggesting that stomatopods can perceive this variation. Together, these results suggest that total reflectance signals weapon performance to an opponent, which may be an indicator of resource holding potential. This is likely to be valuable information to a stomatopod during a contest over possession of a refuge. During these contests, one stomatopod is partially concealed within the refuge, and it may be difficult for a challenger stomatopod to assess the resident's size. Meral spot total reflectance could provide an honest signal of resource holding potential, thus allowing stomatopods to avoid injuries from contests escalating to physical combat (Berzins and Caldwell, 1983) when opponents are not equally matched. To complete the link between strike force or impulse and resource holding potential, further studies of Neogonodactylus oerstedii should investigate whether stomatopods that win contests have greater strike force or strike impulse.

In the group of stomatopods studied here, it might appear that reflectance is separated into two groups (those with reflectance less than 5.5% and those with reflectance greater than 6.9%, in Fig. 4). Previous research has demonstrated that bimodal distributions can be biologically important (e.g., Nelson and Marler, 1989). However, the division was not related to any group within the data set, including sex or size class, and previous research with larger samples sizes has not detected two groups (Franklin et al., 2016). This suggests that the pattern is not biologically relevant but rather is a characteristic of our sample.

Moreover, meral spot chroma did not correlate with peak force or impulse, suggesting that chroma does not indicate performance. It is possible that other measures of color, such as hue or UV chroma, may correlate with performance. These measures were not included in the final statistical model because of the low sample size. However, behavioral wavelength discrimination tests in the laboratory suggest that stomatopods have relatively coarse color vision in the visible spectrum (Thoen et al., 2014). This result led Thoen and colleagues to suggest that stomatopods do not process color similarly to other organisms. While we do not know much about how stomatopods process chromatic information, this coarse spectral discrimination could indicate that stomatopods have a poor ability to detect small variations in chroma among meral spots. Although it is possible that stomatopods have greater spectral discrimination in different behavioral contexts (Jacobs, 1981; Neumeyer, 1991) or in other regions of the spectrum (e.g., UV vision), the results we present here do not provide any reliable evidence that meral spot chroma is an agonistic signal.

Both impulse and peak force increased with stomatopod body length. Muscle force and power are proportional to cross-sectional area and volume, respectively, both of which are generally larger in larger animals (Schmidt-Nielsen, 1984). By these simple scaling relations, larger stomatopods should produce more powerful strikes. Indeed, previous research into scaling in stomatopod strikes has shown that force increases with increasing muscle cross-sectional area (Mendoza Blanco and Patek, 2014), size of the raptorial appendage (striking appendage), and overall body size (Patek and Caldwell, 2005; Zack et al., 2009; Claverie et al., 2011). Our size range is smaller than the size range in Claverie et al. (2011) (this study: 37.4-53.6 mm; Claverie et al: 20-60 mm), but we detected a greater range of peak force measurements (63.1-357.2 N). This may indicate that these scaling relationships differ across species or with different morphologies.

For any given individual, strike force and impulse did not appear to be influenced by our experimental schedule. We did not find any evidence that striking with greater force or impulse influences subsequent strikes or that stomatopod strikes become weaker within or across testing days. This suggests that stomatopods can recover quickly from a strike. However, in contests stomatopods can strike in faster succession than the lag between strike recordings here. On a shorter timescale, a greater investment in one strike may influence subsequent strike force, strike impulse, or strike frequency. This could result in a trade-off between investment in the current strike and investment in future strikes.

However, the duration of time an individual was held in captivity did influence peak force and meral spot total reflectance, but there was no correlation between duration of captivity and impulse. In particular, stomatopods held in the lab for longer durations had lower peak force estimates and lower meral spot total reflectance. Because of this, we excluded stomatopods from our analysis that were in the lab for more than 100 days. The magnitude of this effect was similar to the decrease in peak force associated with increased meral spot total reflectance. This suggests that lab housing may result in a mismatch between weapon performance conveyed through meral spot total reflectance and actual weapon performance. This change in total reflectance may indicate that stomatopods in the lab are lacking dietary nutrients required for color formation (Bandaranayake, 2006). Thus, prolonged housing in the laboratory may disrupt stomatopod color signaling systems and should be a consideration for behavioral experiments. Although these results indicate that stomatopods may have deteriorated during laboratory housing, when these changes are controlled for in the statistical analyses, impulse is still correlated with meral spot total reflectance, and peak force shows a negative trend (P = 0.053). This suggests that the results relating to meral spot total reflectance are robust.

The mechanism that maintains the honesty of meral spot total reflectance as a signal of weapon performance may be due to the pigments that give the meral spot its color. The spot's color probably comes from carotenoproteins, a set of proteins that make many purples and blues in crustaceans (Newbigin, 1897; Zagalsky et al., 1970; Bandaranayake, 2006). These chemical complexes require carotenoids, which the animals cannot synthesize themselves but must ingest. The animals probably face a trade-off, because carotenoids are also required for immune and other physiological functions (Britton, 2008; Svensson and Wong, 2011). Allocating carotenoids for coloration means that they cannot be used for other physiological requirements. It follows that only individuals in good condition would invest heavily in ornamentation because only for them will the benefit of a stronger signal outweigh the costs (Lozano, 2001; Getty, 2006). Diet supplementation or pigment extraction coupled with high-performance liquid chromatography experiments could be conducted to determine whether this trade-off is the mechanism behind signal reliability in Neogonodactylus oerstedii. Such experiments could also provide insight into whether stomatopods can modify meral spot total reflectance within a molt cycle, and into the relationship between pigment density, body condition, and meral spot total reflectance.

Our results here may be generalizable to other stomatopod species with colored meral spots. Many stomatopod species have colored meral spots that they display during agonistic encounters (Caldwell and Dingle, 1975). Research suggests that colored meral spots can be assessed reliably across a range of depths, despite the reduction in short and long wavelengths associated with increasing depth (Cheroske and Cronin, 2005). Colored meral spots tend to be purple, orange, or red (to human eyes), colors that are likely created by carotenoproteins (Newbigin, 1897; Zagalsky et al., 1970; Bandaranayake, 2006). Immediately after molting, when the exoskeleton is soft and stomatopods are unable to strike, these meral spots are much lighter in color (Reaka, 1975; Franklin et al., 2017). Consequently, stomatopods with colored meral spots may not be able to bluff fighting ability after a molt, as has been observed in a species with a white meral spot (Adams and Caldwell, 1990). Thus, in these species meral spot total reflectance may be a reliable signal of weapon performance by indicating strike impulse (as demonstrated here) and by indicating a recent molt.

Species that have white meral spots are unlikely to signal weapon performance by using meral spot total reflectance. This is supported by observations that species with white meral spots tend to perform the meral spread less often during contests than species with colorful meral spots (Caldwell and Dingle, 1975, 1976). Furthermore, recent research in Neogonodactylus bredini, a species with a white meral spot, demonstrated that the meral spread is not used to resolve size-matched contests (Green and Patek, 2015). The mechanism of signal honesty we proposed above is based on pigments, which may not be present in white meral spots. White meral spots could potentially provide a signal in the UV, which has not been measured. However, the apparent lack of pigments, along with the behavioral evidence (Caldwell and Dingle, 1975, 1976; Green and Patek, 2015), suggests that white meral spots do not act as a signal of weapon performance. Instead, they may help amplify the size of the raptorial appendage, similar to markings that amplify the size of the abdomen in jumping spiders (Taylor et al., 2000). Alternatively, the bright white and UV reflectance of the meral spot may improve detection of the meral spread by a receiver. Greater contrast in spectral intensity between a signal and the background can improve signal detection (e.g., anoline lizards, Persons et al., 1999). Further research is needed into the reflectance variability of white meral spots and into stomatopod behavioral responses to any variability present.

Green and Patek (2015) also found that strike force did not predict contest outcome; rather, the stomatopod that delivered more strikes tended to win contests. However, strike force and strike impulse likely play a role in stomatopod conflict resolution in different conditions. Green and Patek staged contests between size-matched stomatopods (all pairs <4% difference in body length). Within these pairs, the strike force differed by only 2.48 N. However, if stomatopods compete with larger or smaller opponents, the difference between opponents in strike force and strike impulse is likely to be much larger. Here, we report differences of almost 300 N in peak strike force (range: 63.2-357.2 N) and 2.7 mN*s in peak strike impulse (range: 0.5-3.2 mN*s). Stomatopod size is known to influence contest outcome in several stomatopod species (Caldwell and Dingle, 1979; reviewed in Caldwell, 1987), and studies suggest that stomatopods are more likely to escalate contests if their opponent has a weak strike (Caldwell, 1979; Berzins and Caldwell, 1983). Thus, strike impulse and force may play a role in resolving contests when contestants are not so evenly matched. Interestingly, Green and Patek (2015) also did not find a strong correlation between size of striking appendage components and strike force, suggesting that the meral spread does not indicate fighting ability. However, studies with other species of stomatopods have found strong correlations ([r.sup.2] > 0.55) between components of the striking appendage (e.g., merus length, propodus and dactylus size) and strike force (Claverie et al., 2010) or spring force (Zack et al., 2009; Claverie et al., 2011). Thus, the function of the meral spread during territorial contests may vary across species. Future studies should investigate the role of colored meral spots, strike impulse, and strike force in asymmetric contests, as well as compare contest behaviors between stomatopod species.

Meral spots may also play additional roles in stomatopod ecology, including in courtship rituals, species recognition, and interactions with predators. Stomatopods perform the meral spread during courtship (Hatziolos and Caldwell, 1983), and it is possible that females assess the meral spot of potential mates. We have also previously noted that females tend to have lighter meral spots than males (Franklin et al., 2016). In the current study, we did not detect a difference between males and females in total reflectance; however, we had data from only four male stomatopods. Thus, more research will be required to determine whether the meral spot plays a role in sex identification. The meral spot probably also plays a role in species recognition among different stomatopod species. Meral spot color varies across stomatopod species, and stomatopods that overlap in range tend to have differently colored meral spots (Caldwell and Dingle, 1976). Lastly, the meral spot may also act as a diematic or startle display against predators (Stevens, 2005; Umbers et al., 2015). Visual models investigating how a trichromatic fish predator perceives N. oerstedii meral spots suggest that the meral spots contrast chromatically with seagrass and rubble backgrounds (AMF, unpubl. data). Thus, fish predators can likely distinguish the meral spots; but whether stomatopods use the meral spread and meral spots as a startle display has not been investigated. Consequently, further research will likely demonstrate that the meral spot is a signal that has different functions across different contexts (i.e., a pluripotent signal, Hebets et al., 2016).

Our research contributes to the growing literature documenting signals that advertise performance (e.g., Lailvaux et al., 2005; Lappin et al., 2006; By water et al., 2008; Henningsen and Irschick, 2012; Mowles and Briffa, 2012). Further research into stomatopods may identify a link between signal, performance, and ability, similar to several other studies that have demonstrated a link between performance and contest outcome (Robson and Miles, 2001; Lailvaux and Irschick, 2007; Bywateref et al., 2008; Mowles et al., 2010). Such studies provide support for current theory that signals should be honest on average if signals are to persist in the population (Maynard Smith and Parker, 1976; Johnstone and Grafen, 1993; Searcy and Nowicki, 2005).

Acknowledgments

We thank Michelle Ysrael and Grove Belden for assistance with data collection, Dr. Sheila Patek for advice on experimental equipment and setup, and Dr. Sara Lewis for laboratory support and helpful discussion. AMF was supported by a Fulbright Science and Technology Fellowship and received funding from the Animal Behaviour Society E. O. Wilson award. We thank two anonymous reviewers for constructive feedback.

Data Accessibility

Data are available at the ZMAPortal (http://zmaportal.org/zmaportal/), under study ZMA14.

Literature Cited

Adamo, S. A., and R. T. Hanlon. 1996. Do cuttlefish (Cephalopoda) signal their intentions to conspecifics during agonistic encounters? Anim. Behav. 52: 73-81.

Adams, E. S., and R. L. Caldwell. 1990. Deceptive communication in asymmetric fights of the stomatopod crustacean Gonodactylus bredini. Anim. Behav. 3: 706-716.

Anderson, R. A., L. D. McBrayer, and A. Herrel. 2008. Bite force in vertebrates: opportunities and caveats for use of a nonpareil whole animal performance measure. Biol. J. Linn. Soc. 93: 709-720.

Andersson, S., S. Pryke, J. Ornborg, M. Lawes, and M. Andersson. 2002. Multiple receivers, multiple ornaments, and a trade-off between agonistic and epigamic signaling in a widowbird. Am. Nat. 160: 683-691.

Balzarini, V., M. Taborsky, F. Villa, and J. G. Frommen. 2016. Computer animations of color markings reveal the function of visual threat signals in Neolamprologus puleher. Curr. Zool. 63: 45-54.

Bandaranayake, W. M. 2006. The nature and role of pigments of marine invertebrates. Nat. Prod. Rep. 23: 223-255.

Bates, D., M. Maechler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67: 1-48.

Berzins, I. K., and R. L. Caldwell. 1983. The effect of injury on the agonistic behavior of the stomatopod, Gonodactylus bredini (Manning). Mar. Behav. Physiol. 10: 83-96.

Blount, J. D., H. M. Rowland, F. P. Drijfhout, J. A. Endler, R. Inger, J. J. Sloggett, G. D. D. Hurst, D. J. Hodgson, and M. P. Speed. 2012. How the ladybird got its spots: effects of resource limitation on the honesty of aposematic signals. Fund. Ecol. 26: 334-342.

Britton, G. 2008. Functions of intact carotenoids. Pp. 189-212 in Carotenoids, G. Britton, S. Liaaen-Jensen. and H. Pfander, eds. Birkhauser, Basel.

Bywater, C. L., M. J. Angilletta, and R. S. Wilson. 2008. Weapon size is a reliable indicator of strength and social dominance in female slender crayfish (Cherax dispar). Funct. Ecol. 22: 311-316.

Caldwell, R. L. 1979. Cavity occupation and defensive behavior in the stomatopod Gonodactylus festae: evidence for chemically mediated individual recognition. Anim. Behav. 27: 194-201.

Caldwell, R. L. 1987. Assessment strategies in stomatopods. Bull. Mar. Sci. 41: 135-150.

Caldwell, R. L., and H. Dingle. 1975. Ecology and evolution of agonistic behavior in stomatopods. Naturwissenschaften 62: 214-222.

Caldwell, R. L., and H. Dingle. 1976. Stomatopods. Sci. Am. 234: 80-89.

Caldwell, R. L., and H. Dingle. 1979. The influence of size differential on agonistic encounters in the mantis shrimp. Gonodactylus viridis. Behavior 69: 255-264.

Cameron, A. C, and F. A. Windmeijer. 1997. An R-squared measure of goodness of fit for some common nonlinear regression models. J. Econom. 11: 329-342.

Candolin, U. 2000. Male-male competition ensures honest signaling of male parental ability in the three-spined stickleback (Gasterosteus aculeatus). Behav. Ecol. Sociobiol. 49: 57-61.

Cheroske, A. G., and T. W. Cronin. 2005. Variation in stomatopod (Gonodactylus smithii) color signal design associated with organismal condition and depth. Brain Behav. Evol. 66: 99-113.

Claverie, T., E. Chan, and S. N. Patek. 2011. Modularity and scaling in fast movements: power amplification in mantis shrimp. Evolution 65: 443-461.

Corbett, G. G., S. R. Reid, and W. Johnson. 1996. Impact loading of plates and shells by free-flying projectiles: a review. Int. J. Impact Eng. 18: 141-230.

Cortesi, F., and K. L. Cheney. 2010. Conspicuousness is correlated with toxicity in marine opisthobranchs. J. Exp. Biol. 7: 1509-1518.

Coxe, S., S. G. West, and L. S. Aiken. 2013. Generalized Linear Models. Oxford University Press. Oxford.

Delhey, K., A. Johnsen, A. Peters, S. Andersson, and B. Kempenaers. 2003. Paternity analysis reveals opposing selection pressures on crown coloration in the blue tit (Parus caendeus). Proc. R. Soc. Biol. Sci. B 270: 2057-2063.

Dingle, H. 1969. A statistical and informational analysis of aggressive communication in the mantis shrimp Gonodactylus bredini Manning. Anim. Behav. 17: 561-575.

Endler, J. A. 1990. On the measurement and classification of colour in studies of animal colour patterns. Biol. J. Linn. Soc. 41: 315-352.

Enquist, M. 1985. Communication during aggressive interactions with particular reference to variation in choice of behaviour. Anim. Behav. 33: 1152-1161.

Florence, A. L. 1966. Circular plate under a uniformly distributed impulse. Int. J. Solids Struct. 2: 37-47.

Fox, J., and S. Weisberg. 2011. An R Companion to Applied Regression, 2nd ed. Sage, Thousand Oaks, CA.

Franklin, A. M., N. J. Marshall, and S. M. Lewis. 2016. Multimodal signals: ultraviolet reflectance and chemical cues in stomatopod agonistic encounters. R. Soc. Open Sci. 3: 160329.

Franklin, A. M., M. B. Applegate, S. M. Lewis, and F. G. Omenetto. 2017. Stomatopods detect and assess achromatic cues in contests. Behav. Ecol. 28: 1329-1336.

Frischknecht, M. 1993. The breeding colouration of male three-spined sticklebacks (Gasterosteus aculeatus) as an indicator of energy investment in vigour. Evol. Ecol. 1: 439-450.

Gerlach, T., D. Sprenger, and N. K. Michiels. 2014. Fairy wrasses perceive and respond to their deep red fluorescent coloration. Proc. R. Soc. Biol. Sci. B 281: 20140787.

Getty, T. 2006. Sexually selected signals are not similar to sports handicaps. Trends Ecol. Evol. 21: 83-88.

Green, P. A., and S. N. Patek. 2015. Contests with deadly weapons: telson sparring in mantis shrimp (Stomatopoda). Biol. Lett. 11: 20150558.

Hagey, T. J., J. B. Puthoff, K. E. Crandell, K. Autumn, and L. J. Harmon. 2016. Modeling observed animal performance using the Weibull distribution. J. Exp. Biol. 219: 1603-1607.

Hansen, H. J. 1895. Isopoden, Cumaceen u. Stomatopoden der Plankton-Expedition. Lipsius & Tischer, Kiel.

Hatziolos, M. E., and R. L. Caldwell. 1983. Role reversal in courtship in the stomatopod Pseudosquilla ciliata (Crustacea). Anim. Behav. 31: 1077-1087.

Head, A. W., J. S. Hardin, and S. C. Adolph. 2012. Methods for estimating peak physiological performance and correlating performance measures. Environ. Ecol. Stat. 19: 127-137.

Hebets, E. A., A. B. Barron, C. N. Balakrishnan, M. E. Hauber, P. H. Mason, and K. L. Hoke. 2016. A systems approach to animal communication. Proc. R. Soc. Biol. Sci. B 283: 20152889.

Henningsen, J. P., and D. J. Irschick. 2012. An experimental test of the effect of signal size and performance capacity on dominance in the green anole lizard. Funct. Ecol. 26: 3-10.

Irschick, D. J., J. J. Meyers, J. F. Husak, and J.-F. Le Galliard. 2008. How does selection act on whole-organism functional performance capacities? A review and synthesis. Evol. Ecol. Res. 10: 177-196.

Jacobs, G. H. 1981. Comparative Color Vision. Academic Press. New York.

Johnstone, R. A., and A. Grafen. 1993. Dishonesty and the handicap principle. Anim. Behav. 46: 759-764.

Kelley, J. L., G. M. Rodgers, and L. J. Morrell. 2016. Conflict between background matching and social signalling in a colour-changing freshwater fish. R. Soc. Open Sci. 3: 160040.

Lailvaux, S. P., and J. F. Husak. 2014. The life history of whole-organism performance. Q. Rev. Biol. 89: 285-318.

Lailvaux, S. P., and D. J. Irschick. 2007. The evolution of performance-based male fighting ability in Caribbean Anolis lizards. Am. Nat. 170: 573-586.

Lailvaux, S. P., J. Hathway, J. Promfret, and R. J. Knell. 2005. Horn size predicts physical performance in the beetle Euoniticellus intermedius (Coleoptera: Scarabaeidae). Funct. Ecol. 19: 632-639.

Lappin, A. K., Y. Brandt, J. F. Husak, J. M. Macedonia, and D. J. Kemp. 2006. Gaping displays reveal and amplify a mechanically based index of weapon performance. Am. Nat. 168: 100-113.

Ligon, R. A., and K. J. McGraw. 2016. Social costs enforce honesty of a dynamic signal of motivation. Proc. R. Soc. Biol. Sci. B 283: 20161873.

Lindstedt, C, J. H. R. Talsma, E. Ihalainen, L. Lindstrom, and J. Mappes. 2010. Diet quality affects warning coloration indirectly: excretion costs in a generalist herbivore. Evolution 64: 68-78.

Losos, J. B., D. A. Creer, and J. A. I. Shulte. 2002. Cautionary comments on the measurement of maximum locomotor capabilities. J. Zool. 258: 57-61.

Lozano, G. 2001. Carotenoids, immunity, and sexual selection: comparing apples and oranges? Am. Nat. 158: 200-203.

Marshall, J. N., and J. Oberwinkler. 1999. Ultraviolet vision: the colourful world of the mantis shrimp. Nature 401: 873-874.

Maynard Smith, J. 1979. Game theory and the evolution of behaviour. Proc. R. Soc. Biol. Sci. B 205: 475-488.

Maynard Smith, J., and G. A. Parker. 1976. The logic of asymmetric contests. Anim. Behav. 24: 159-175.

McGraw, K. J. 2007. Dietary mineral content influences the expression of melanin-based ornamental coloration. Behav. Ecol. 18: 137-142.

Mendoza Blanco, M., and S. N. Patek. 2014. Muscle trade-offs in a power-amplified prey capture system. Evolution 68: 1399-1414.

Moodie, G. 1972. Predation. natural selection and adaptation in an unusual threespine stickleback. Heredity 28: 155-167.

Mowles, S. L., and M. Briffa. 2012. Forewarned is forearmed: early signals of RHP predict opponent fatigue in hermit crab shell fights. Behav. Ecol. 23: 1324-1329.

Mowles, S. L., P. A. Cotton, and M. Briffa. 2010. Whole-organism performance capacity predicts resource-holding potential in the hermit crab Pagurus bernhardus. Anim. Behav. 80: 277-282.

Nelson, D. A., and P. Marler. 1989. Categorical perception of a natural stimulus continuum: birdsong. Science 244: 976-978.

Neumeyer, C. 1991. Evolution of colour vision. Pp. 284-305 in Evolution of the Eye and Visual System: Vision and Visual Dysfunction, Vol. 2, J. Cronly-Dillon and R. L. Gregory, eds. Macmillan, London.

Newbigin, M. 1.1897. The pigments of the decapod Crustacea. J. Physiol. 21: 237-257.

Parker, G. A. 1974. Assessment strategy and the evolution of fighting behaviour. J. near. Biol. 47: 223-243.

Patek, S. N., and R. L. Caldwell. 2005. Extreme impact and cavitation forces of a biological hammer: strike forces of the peacock mantis shrimp Odontodactylus scyllarus. J. Exp. Biol. 208: 3655-3664.

Patek, S. N., W. L. Korff, and R. L. Caldwell. 2004. Deadly strike mechanism of a mantis shrimp. Nature 428: 819-820.

Persons, M. H., L. J. Fleishman, M. A. Frye, and M. E. Stimphill. 1999. Sensory response patterns and the evolution of visual signal design in anoline lizards. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 184: 585-607.

Plasman, M., V. H. Reynoso, L. Nicolas, and R. Torres. 2015. Multiple colour traits signal performance and immune response in the Dickerson's collared lizard Crotaphytus dickersonae. Behav. Ecol. Sociobiol. 69: 765-775.

Pryke, S. R., M. J. Lawes, and S. Andersson. 2001. Agonistic carotenoid signalling in male red-collared widowbirds: aggression related to the colour signal of both the territory owner and model intruder. Anim. Belmv. 62:695-704.

R Core Team. 2014. R: a language and environment for statistical computing. [Online]. R Foundation for Statistical Computing. Vienna. Available:http://www.R-project.org [2018, November 5].

Reaka, M. L. 1975. Molting in stomatopod crustaceans. I. Stages of the molt cycle, setagenesis, and morphology. J. Morphol. 146: 55-80.

Robson, M. A., and D. B., Miles. 2001. Locomotor performance and dominance in male tree lizards, Urosaurus omatus. Fund. Ecol. 14: 338-344.

Scheel, D., P. Godfrey-Smith, and M. Lawrence. 2016. Signal use by octopuses in agonistic interactions. Curr. Biol. 26: 377-382.

Schmidt-Nielsen, K. 1984. Scaling: Why Is Animal Size so Important? Cambridge University Press. Cambridge.

Searcy, W. A., and S. Nowicki. 2005. The Evolution of Animal Communication. Princeton University Press, Princeton, NJ.

Siefferman, L., and G. E. Hill. 2005. UV-blue structural coloration and competition for nestboxes in male eastern bluebirds. Anim. Behav. 69: 67-72.

Stevens, M. 2005. The role of eyespots as anti-predator mechanisms, principally demonstrated in the Lepidoptera. Biol. Rev. 80: 573-588.

Svensson, P. A., and B. B. M. Wong. 2011. Carotenoid-based signals in behavioural ecology: a review. Behaviour 148: 131-189.

Taylor, P. W., O. Hasson, and D. L. Clark. 2000. Body postures and patterns as amplifiers of physical condition. Proc. R. Soc. Biol. Sci. B 267: 917-922.

Therneau, T. M., and P. M. Grambsch. 2000. Modeling Survival Data: Extending the Cox Model. Springer, New York.

Thoen, H. H., M. J. How, T.-H. Chiou, and J. Marshall. 2014. A different form of color vision in mantis shrimp. Science 343: 411-413.

Umbers, K. D. L., J. Lehtonen, and J. Mappes. 2015. Deimatic displays. Curr. Biol. 25: R58-R59.

Wade, N" K. C. Goulter, K. J. Wilson, M. R. Hall, and B. M. Degnan. 2005. Esterified astaxanthin levels in lobster epithelia correlate with shell colour intensity: potential role in crustacean shell colour formation. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 141: 307-313.

Whiting, M. J., D. M. Stuart-Fox, D. O'Connor, D. Firth, N. C. Bennett, and S. P. Blomberg. 2006. Ultraviolet signals ultra-aggression in a lizard. Anim. Behav. 72: 353-363.

Zack, T. I., T. Claverie, and S. N. Patek. 2009. Elastic energy storage in the mantis shrimp's fast predatory strike. J. Exp. Biol. 212: 4002-4009.

Zagalsky, P. R., H. J. Ceccaldi, and R. Daumas. 1970. Comparative studies on some decapod crustacean carotenoproteins. Comp. Biochem. Physiol. 34: 579-607.

Zuur, A., E. N. Ieno, and C. S. Elphick. 2010. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 1: 3-14.

AMANDA M. FRANKLIN (*), CASSANDRA M. DONATELLI, CASEY R. CULLIGAN, AND ERIC D. TYTELL

Department of Biology, Tufts University, Medford, Massachusetts

DOI: 10.1086/700836

Received 18 January 2018; Accepted 18 September 2018; Published online 28 December 2018.

(*) To whom correspondence should be addressed. Email: am7franklin@gmail.com.

Online enhancement: video.
Table 1
Results of marginal hypothesis tests for the effects of day number,
punch number within day, and arm number on Neogonodactylus oerstedii
strike force and strike impulse

Variable        [chi square]  df   P

Strike force
  Day number     3.30         1    0.069
  Punch number   0.68         1    0.41
within day
  Arm number     0.40         1    0.53
Strike impulse
  Day number     2.70         1    0.10
  Punch number   0.01         1    0.92
within day
  Arm number    11.23         1   <0.001

Bold indicates P < 0.05.

Table 2
Results of marginal hypothesis tests for the effects of total
reflectance, chroma, length, and duration of captivity on
Neogonodactylus oerstedii peak force and peak impulse

Variable               [chi square]   df   P       Pseudo-[R.sup.2]

Peak force
  Total reflectance    3.75           1    0.053   0.20
  Chroma               0.003          1    0.96
  Length               5.36           1    0.021   0.10
  Days in lab          4.01           1    0.045   0.30
Peak impulse
  Total reflectance    5.44           1    0.020   0.17
  Chroma               0.015          1    0.90
  Length              10.28           1    0.001   0.20
  Days in lab          2.32           1    0.13

Pseudo-[R.sup.2] is calculated as [R.sup.2] = 1 -
([D.sub.A]/[D.sub.0]), where [D.sub.0] is the deviance of the
null model and [D.sub.A] is the deviance of the model including
the parameter of interest. Bold indicates P < 0.05; italics indicate P
< 0.06.
COPYRIGHT 2019 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Franklin, Amanda M.; Donatelli, Cassandra M.; Culligan, Casey R.; Tytel, Eric D.
Publication:The Biological Bulletin
Article Type:Report
Date:Feb 1, 2019
Words:8888
Previous Article:Hemoglobin Polymerization in Red Blood Cells of Marine Fishes: A Case of Adaptive Phenotypic Plasticity?
Next Article:Catecholic Compounds in Ctenophore Colloblast and Nerve Net Proteins Suggest a Structural Role for DOPA-Like Molecules in an Early-Diverging Animal...
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters