Consequences of Intraspecific Variation in Female Body Size in
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
BEHAVIOR Consequences of Intraspecific Variation in Female Body Size in Stagmomantis limbata (Mantodea: Mantidae): Feeding Ecology, Male Attraction, and Egg Production MICHAEL R. MAXWELL1 AND CAYLIN FRINCHABOY Department of Mathematics and Natural Sciences, National University, 11255 North Torrey Pines Road, La Jolla, CA 92037 Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Environ. Entomol. 43(1): 91Ð101 (2014); DOI: http://dx.doi.org/10.1603/EN12310 ABSTRACT Body size is an important feature of organisms, inßuencing many components of life history and Þtness, such as feeding success and reproductive output. Body size is considered especially salient for solitary predators, whose food intake hinges on individual predation success, which in turn is often driven by the relative sizes of predator and prey. The current study examined intraspeciÞc variation in adult female length and its Þtness consequences in a solitary predator, the praying mantid Stagmomantis limbata Hahn. Through a 5-yr integration of observational and experimental work in the Þeld and captivity, we investigated the relationship between female pronotum length and prey size, diet breadth, male attraction, and measures of egg production (fecundity and ootheca mass). We found that longer females ate longer prey in the Þeld and showed greater breadth of prey size than shorter females. Longer females did not necessarily feed at higher rates in the Þeld, as measured by the rate of abdominal expansion. Female length failed to show signiÞcant effects on male attraction or on the incidence of cannibalism. Longer females had higher fecundity (mature eggs in body at death) and laid heavier oothecae than shorter females. In nature, longer females consistently emerged as adults earlier in the season than shorter females. Shorter female adults emerged when feeding rates were higher in the Þeld, suggesting an incidental ecological beneÞt of shorter adult size. KEY WORDS body size, prey size, Þtness, praying mantis, Stagmomantis limbata Body size is an important feature of organisms, as it Solitary predators include praying mantids (Manto- inßuences many components of life history and Þtness dea), elongate insects that exhibit considerable inter- (Peters 1983, Roff 1992, Stearns 1992, Hone and Ben- speciÞc variation in body length, from 1 cm to ⬎15 cm ton 2005). In many organisms, increasing body size (Ehrmann 2002). Mantids have been the subject of results in increased diet breadth, longevity, and re- many studies on foraging theory and community ecol- productive output, both between and within species ogy (Charnov 1976, Holling et al. 1976, Hurd and (Peters 1983, Andersson 1994, Chown and Gaston Eisenberg 1990, Moran et al. 1996; reviewed in Hurd 2010). Research on predatory species typically reveals 1999, Fagan et al. 2002). Examinations of the ecolog- that prey size and diversity increase with predator size ical consequences of size variation in mantids have (Schoener 1971, Wilson 1975, Cohen et al. 1993, Costa focused primarily on interspeciÞc variation (Hurd and 2009). For prey, large body size may afford protection Eisenberg 1989a,b; reviewed in Hurd 1999). In a metaÐ from predation (Paine 1976, Polis 1988). Consider- analysis of the impacts of three mantid species on the ations of relative size between individuals are espe- Þeld assemblages of arthropod prey, Fagan et al. cially relevant for cannibalistic species, where the (2002) detected different impacts for different spe- consumer and consumed are often determined by cies, suggesting difference in body size as a probable relative size (Fox 1975, Polis 1981, Claessen et al. contributing factor. In a Þeld study of two other man- 2000). Within a species, large body size has been found tid species, Maxwell and Eitan (1998) demonstrated to correlate with fecundity and mating success interspeciÞc differences in body length measure- (Honek̆ 1993, Andersson 1994, Bonduriansky 2001). ments, with longer species eating longer prey than the Body size is considered to be particularly important shorter species. for solitary predators, whose food intake hinges on The Þtness consequences of intraspeciÞc variation individual predation success, which in turn is often in size in mantids have received relatively less atten- driven by the relative sizes of predator and prey (Wil- tion. A few studies have documented variation in size son 1975, Claessen et al. 2002, McCoy et al. 2011). in natural populations within nymphal stadia and among adult females (Matsura et al. 1975, Hurd and 1 Corresponding author, e-mail: mmaxwell@nu.edu. Eisenberg 1989a, Maxwell and Eitan 1998). Studies in 0046-225X/14/0091Ð0101$04.00/0 䉷 2014 Entomological Society of America
92 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 1 captivity show that variation in body mass and length of cannibalism (Moya-Laraño et al. 2003, Wilder and among early instars can be generated by differences in Rypstra 2008). To test our hypotheses, we integrated food per capita (e.g., Hurd and Rathet 1986, Dussé and naturalistic observations with Þeld experiments and Hurd 1997, Moran and Hurd 1997), as suggested by work in captivity. Þeld manipulations of nymphal density (Hurd and Eisenberg 1984, Fagan and Hurd 1994). The conse- Materials and Methods quences of intraspeciÞc size variation have been re- stricted to captivity. From HollingÕs morphometric Field Observations: Biometry, Feeding Ecology, models of mantids as predators (Holling 1964, Holling and Behavior. We measured and observed adult fe- et al. 1976), longer individuals can be expected to males at a Þeld site along Baker Creek in Big Pine, CA attack larger prey, as appears to be the case in captive (37⬚ 10⬘ N, 118⬚ 17⬘ W) for 4 yr (2008 Ð2011). At this studies on Tenodera aridifolia (Stoll, 1813) (Hurd site, populations are nonoverlapping and univoltine, 1988, Iwasaki 1991, Whitman and Vincent 2008). In- with hatching occurring in spring, adults Þrst appear- traspeciÞc size variation can have reproductive con- ing in August, persisting through late September, and Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 sequences, as Matsura et al. (1975) suggest a positive dying off in October with the onset of cold temper- relationship between body length and ootheca (egg atures. Mantids occur along the banks of the creek, case) mass in T. angustipennis Saussure. In staged mat- which are dominated by willow (Salix sp.), black lo- ing trials, however, neither adult female length nor the cust (Robinia pseudoacacia L.), rose hips (Rosa wood- degree of mate size dimorphism was found to affect sii Lindley), and long-leaved aster (Aster ascendens the occurrence of female-on-male cannibalism in the Lindley). In S. limbata, as in many mantids, the ßight- mantids Pseudomantis albofimbriata Stål and Stagmo- less females are somewhat sedentary, allowing for re- mantis limbata Hahn (Barry et al. 2008, Maxwell et al. peated measurements and observations of individuals 2010b). in nature (Maxwell 1998, Maxwell et al. 2010b). The current study examines the consequences of Field observations were conducted through weekly intraspeciÞc size variation in a Þeld setting for the censuses in 2008 and daily monitoring of marked fe- mantid S. limbata, which is native to North America, males from 2009 to 2011, based on methods described occurring from the western United States into South by Maxwell et al. (2010b). In 2008, we conducted six America (Roberts 1937, Ehrmann 2002). Females typ- weekly censuses between 20 August and 27 Septem- ically exceed 50 mm in body length and 1.5 g in body ber. Each census covered the same 847-m section of mass after a few weeks of feeding as adults, whereas the creekside. To conduct each census, we visually males rarely exceed 50 mm or 0.5 g as adults. Females searched for S. limbata between 0900 and 1600 hours, are ßightless, solitary ambush predators, while males gently shaking the vegetation within 2 m of the creek are ßight-capable and more mobile than the females. bank. For each adult found, we recorded sex, prono- Roberts (1937) documented variation in the number tum length (to 0.1 mm), dorsoventral abdominal of nymphal stadia within captive colonies, suggesting thickness (females only, to 0.1 mm), underwing col- intracohort variation in size among adults. In nature oration, and any intersexual interactions (i.e., mount- and captivity, variation in a measure of adult length, ing of the female by the male, copulation, and canni- pronotum length, exists between years (Maxwell and balism). For a given stage (i.e., nymphal stadium or Eitan 1998, Maxwell et al. 2010b). In S. limbata and adult), pronotum length is Þxed and does not expand other mantids, manipulations of the feeding regime as the mantid feeds, thereby serving as a standard within cohorts of adult females have demonstrated index of length in mantids (e.g., Lawrence 1992, Max- important effects on fecundity, the occurrence of in- well 1998). We used hind wing coloration to identify tersexual cannibalism, and male behavioral responses newly emerged adult females, as females have diffuse (Matsura and Morooka 1983; Birkhead et al. 1988; yellow underwings upon adult emergence, developing Lelito and Brown 2008; Barry et al. 2010; Maxwell et into a tessellated pattern of distinct yellow cross veins al. 2010a,b). The Þtness effects of variation in female by 4 Ð 8 d postemergence (Maxwell et al. 2010b). For length remain unknown in natural populations. a female found with diffuse yellow underwings, we The current study examines variation in female assigned the date of discovery as the femaleÕs date of length and its consequences in a multiyear study on S. adult emergence. Prey items being eaten by the man- limbata. SpeciÞcally, we investigate the relationship tids were taxonomically identiÞed and measured to 1 between adult female length and prey size, diet cm. All adults were individually marked with perma- breadth, male attraction, and measures of egg produc- nent ink on the pronotum for later identiÞcation, and tion (fecundity and ootheca mass). Based on the ex- were released at the spot of capture. To assess feeding isting literature, we expect longer females to eat larger rate, we measured abdominal thickness of all resighted prey, show greater diet breadth, and show higher females (Maxwell et al. 2010b). fecundity and ootheca mass. With regard to male at- From 2009 to 2011, we conducted daily monitoring traction, we test two competing predictions. Female of females. In each year, we found females for obser- length might correlate with the amount of pheromone vations through daily searches within the 847-m sec- produced, and longer females might be more attrac- tion of creekside by the methods described above (i.e., tive because of increased fecundity (Prenter et al. visually searching through the vegetation between 1994, Bonduriansky 2001). Alternatively, males might 0900 and 1600 hours). The 847-m section was divided avoid longer females if such females pose a greater risk into thirds, and one-third was searched each day such
February 2014 MAXWELL AND FRINCHABOY: INTRASPECIFIC SIZE VARIATION IN S. limbata 93 that each third was searched every 3 d. We marked We collected 24 recently emerged adult females from each adult for later identiÞcation and recorded mor- a site 300 m upstream from the 847-m study section of phological and behavioral data, as described above. In Baker Creek between 22 August and 8 September. 2009 and 2010, we measured the body length of each Each female had diffuse yellow underwings, indicat- adult upon Þrst sighting (front of head to end of ing recent adult emergence, and was assigned as either abdomen, to 1 mm). At the start of each daily census, “short” (pronotum length range ⫽ 18 Ð20 mm, n ⫽ 12) we checked the behavior of all previously marked or “long” (pronotum length range ⫽ 22Ð24 mm, n ⫽ females (i.e., intersexual interactions and prey items). 12). Each female was maintained in a ventilated 1.6- We measured the abdominal thickness of females liter polyethylene container in an outdoor shelter at 12Ð16 d after Þrst sighting to minimize physical han- the Owens Valley Laboratory (University of Califor- dling of the females. Daily censuses and monitoring of nia, Bishop, CA). Each female was provided one marked females occurred in the periods 11 August to cricket nymph (Acheta domesticus L.) every 2 d from 5 September 2009, 8 September to 15 October 2010, capture until 15 September. Over this feeding period, and 29 August to 29 September 2011. In 2009, we the abdomen of each female increased to at least 9 mm Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 stopped daily censuses on 6 September, but continued in thickness, indicating a mature clutch of eggs and to daily monitor previously marked females until 28 sexual receptivity (Maxwell et al. 2010b); mean ⫾ SE September. As we monitored these females, we abdominal thickness ⫽ 10.3 ⫾ 0.2 mm (n ⫽ 24). marked, measured, and recorded behavior of any en- We placed the females at the Baker Creek site on 15 countered unmarked adults. September at 1100 hours. Each female was placed in For observed predation events in the Þeld (2008 Ð a plastic mesh cage (17 by 17 by 5 cm) that was 2011), we examined the relationship between female overlain with four layers of light cotton gauze to re- pronotum length and prey length through correlation. duce visual cues and signals. In addition to the 12 short We analyzed the effect of female pronotum length on and 12 long females, we set out six empty covered diet breadth by Þrst sorting the females by pronotum cages to verify our expectation that males would not length, and constructing two subgroups of females for be attracted to the cages per se, based on previous comparison: the shortest third (i.e., the shortest 33%) and research (Maxwell et al. 2010a, 2010b). The 30 cages the longest third (longest 33%). For each subgroup of were placed at 20-m intervals along a 600-m contin- females, we deÞned diet breadth as the subgroupÕs sam- uous stretch of bank of Baker Creek, in the pattern ple variation in observed prey length. We then compared empty-short-long-short-long replicated sixfold. Each diet breadth between the two subgroups through Lev- cage was fastened to vegetation at a height 1.0 Ð1.5 m. eneÕs test of two variances (Zar 2010). We checked the cages at 1000 and 1430 hours on To assess feeding rates of adult females in the Þeld, each of the three consecutive days (16 Ð18 Septem- we analyzed the rate of abdominal expansion of young ber). We recorded any male found on a cage or within adults in all years from 2008 to 2011 (Maxwell et al. 30 cm of the cage. All males were removed and kept 2010b). Young adult females were identiÞed as those in isolated containers at the Owens Valley Laboratory having diffuse yellow underwings at Þrst sighting. The to eliminate possible effects of experience on male rate of abdominal expansion was calculated as: behavior. To replace these captured males, an equal number of males collected from the Owens Valley Rate ⫽ 共a 2 ⫺ a 1)/t, [1] Laboratory were released 50 m orthogonal to the fe- where a2 ⫽ dorsoventral abdominal thickness (milli- male cage at the midpoint of the linear 600-m stretch meters) 12Ð16 d after the Þrst sighting, a1 ⫽ dorso- at 1600 hours on each day. The experiment was ter- ventral abdominal thickness (millimeters) at Þrst minated after the check at 1430 hours on 18 Septem- sighting, and t ⫽ the actual number of days between ber. All captured males and females were retained for a2 and a1 measurements. This analysis was restricted to further work in captivity. All but 10 females were females measured within 12Ð16 d after the Þrst sight- maintained until they died in captivity; the 10 females ing, as this period is typically too early for a recently were released at Baker Creek on 29 September. emerged female to copulate and oviposit. In the cur- Field Observations: Female and Male Pairings. rent study, the average age of copulation was 15 d for Through the methods described above, we observed females in the Þeld (see Results). While the minimum pairings in the Þeld. We deÞned a pairing as a male age of copulation for these females was 4 d, the current mounted on a female, or a male and female in copu- study found that the average interval between the Þrst lation. Within each year, we compared the pronotum copulation and the Þrst oviposition was 12 d for well- lengths of females observed in at least one pairing with fed females in captivity (see Results). Thus, for the those of all other females measured at the Þeld site. young adult females observed in the Þeld, abdominal These comparisons incorporate the combined effects expansion during the Þrst 12Ð16 d was unlikely to be of all relevant stimuli (e.g., chemical, visual, tactile) on confounded by abdominal ßattening after oviposition. male attraction. Reproduction: Male Attraction, Natural Pairings, Female Fecundity and Ootheca Mass. We analyzed Fecundity, and Ootheca Mass. Field Experiment: Fe- the relationship between female pronotum length and male Length and Male Attraction. In 2011, we con- fecundity (number of mature eggs in body at death) ducted a Þeld experiment to examine male response to and ootheca mass in three generations of females females of different pronotum lengths (long vs. short), (2006, 2008, and 2010). In all years, we obtained virgin with emphasis on chemical attraction by the females. adult females through the collection and rearing of
94 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 1 late-instar nymphs and young adults from several Þeld sites in Bishop and Big Pine, CA, during July through September. Each nymph and adult was isolated in a ventilated 1.6-liter polyethylene container. In 2006, nymphs and adults were transported to an outdoor shelter on a private property near National University, La Jolla, CA (32⬚ 55⬘ N, 117⬚ 15⬘ W). In 2008 and 2010, nymphs and adults were maintained in an outdoor shelter at the Owens Valley Laboratory at Bishop until 28 September 2008 and 17 October 2010, respectively, after which all adults were transported to the shelter near National University. Each nymph was provided with misted water and one house cricket nymph every 2 d, supplemented with mealworms (Tenebrio molitor Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 L.). Mantid nymphs were checked daily for adult emergence. Through these methods, we obtained over 70 virgin females in all years (n ⫽ 73Ð151 per year). We measured pronotum length of all adult females. Adult females were provided a standardized diet of one house cricket nymph every 2 d. Fig. 1. Pronotum length (millimeters) for adults mea- We measured fecundity in 2006 and 2008 for captive sured in the Þeld, 2008Ð2011. White boxes indicate females; females that did not copulate, did not lay oothecae, black boxes indicate males. Horizontal bars indicate mean for and did not cannibalize males. Fecundity was mea- each sex within a year, with ⫾ SE indicated by box; whisker sured by dissecting the females upon death and count- lines indicate 5th and 95th percentiles. Different letters in- dicate signiÞcant difference between groups (Tukey com- ing mature eggs (i.e., 5Ð 6 mm in length and rich orange parisons, family-wise ␣ ⫽ 0.05). For each group, n ⫽ 56Ð185 in coloration). We analyzed females that lived for at adults. least 20 d postemergence, as S. limbata undergo a feeding period to develop a clutch of eggs (n ⫽ 18 in 2006, n ⫽ 12 in 2008). For the second measure of body condition, we used To obtain oothecae, we paired virgin females and the unstandardized residual values from the linear Þeld-collected adult males in 2010 (n ⫽ 26 pairings). regression of female body mass at mating trial on Pronotum length of the females spanned 17Ð22 mm. pronotum length. To standardize recent feeding and mating history Statistical Analyses. We performed statistical anal- among the captured males, each male was fed one yses on the software SPSS 17.0 (IBM Co., Chicago, IL). cricket nymph every 3 d for a minimum of 7 d before All inferential tests are two-tailed, and mean ⫾ SE use in a mating trial. Mating trials consisted of random values are reported for descriptive statistics. When pairings of the females and males between 18 Sep- multiple univariate tests are conducted for a particular tember and 16 October 2010 at the Owens Valley analysis, we lower alpha (␣) by the sequential Bon- Laboratory. At each trial, we weighed each female ferroni technique (Rice 1989). (0.01 g) and measured dorsoventral abdominal thick- ness (0.1 mm); mean ⫾ SE female age at trial ⫽ 19 ⫾ Results 1 d. Each pair was placed in a screen cage (45 by 45 by 50 cm) and was allowed to interact for 10 h. Each Field Observations: Pronotum Length. Mean pro- female copulated with her paired male and did not notum length for females was 21.0 mm (SE ⫽ 0.1, n ⫽ cannibalize him. We weighed the Þrst ootheca (0.01 g) 589, all years combined), and mean pronotum length laid by each female for analysis. for males was 14.9 mm (SE ⫽ 0.047, n ⫽ 359, all years In the analysis of ootheca mass, we included two combined). Females were consistently longer than measures of female body condition: slope-adjusted males in all years, with signiÞcant variation between ratio and regression residuals (Petrie 1983, Jakob et al. years within both sexes (Fig. 1; ANOVA: whole-model 1996, Moya-Laraño et al. 2008). The slope-adjusted F7,940 ⫽ 739.7; P ⬍ 0.001, year P ⬍ 0.001, sex P ⬍ 0.001, ratio uses the general equation: year ⫻ sex interaction P ⬍ 0.01). Pronotum length accurately reßected body length, m ⫽ k ⫻ l a, [2] as evidenced for adults in 2009 and 2010. For females, a strong, positive relationship was found between pro- where m ⫽ body mass at mating trial, l ⫽ pronotum notum length and body length measured at the Þrst length, and k and a are scaling parameters that vary sighting (Multiple regression: F2,232 ⫽ 212.6; adjusted according to species (Peters 1983). For S. limbata R2 ⫽ 0.64, whole-model P ⬍ 0.001, pronotum length females, k ⫽ 0.13 and a ⫽ 2.63 (Maxwell 1998), yield- P ⬍ 0.001, year P ⬎ 0.2; Pearson correlation between ing the following equation for body condition as a pronotum length and body length: r ⫽ 0.80, P ⬍ 0.001). slope-adjusted ratio: For both years combined, mean body length for fe- males was 57 mm (SE ⫽ 1 mm, range ⫽ 44 Ð 69 mm, n ⫽ Body condition ⫽ m/(0.13 ⫻ l 2.63) [3] 235). For males, pronotum length and body length
February 2014 MAXWELL AND FRINCHABOY: INTRASPECIFIC SIZE VARIATION IN S. limbata 95 Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Fig. 2. Pronotum length (millimeters) for newly emerged adult females in the Þeld. Date of adult emergence is on x-axis; numerical dates are Julian dates, where 1 August ⫽ 213. (a) 2008, n ⫽ 27 females. (b) 2009, n ⫽ 37 females. (c) 2010, n ⫽ 20 females. (d) 2011, n ⫽ 29 females. were similarly related (Multiple regression: F2,112 ⫽ Average prey lengths for each year are given in Fig. 155.1; adjusted R2 ⫽ 0.73; whole-model P ⬍ 0.001, 3a. For years with at least 10 predation events (i.e., pronotum length P ⬍ 0.001, year P ⬎ 0.7; Pearson 2009, 2010, and 2011), nonmale prey length failed to correlation: r ⫽ 0.86, P ⬍ 0.001). For both years com- differ signiÞcantly among years (Kruskal-Wallis test bined, mean body length for males was 46 mm (SE ⫽ with tied ranks: Hc ⫽ 2.89; df ⫽ 2; P ⬎ 0.2). The 1 mm; range ⫽ 39 Ð55 mm; n ⫽ 115). inclusion of conspeciÞc males increased the mean For newly emerged adult females in the Þeld, a prey length for 2011 (Fig. 3a), causing this year to negative relationship between pronotum length and inferred emergence date was evident in all years (Fig. Table 1. Predation events by adult female S. limbata in the 2). A negative correlation was signiÞcant in all years, field, 2008 –2011 with r ⬍ ⫺0.5, P ⬍ 0.01 in all years (range of Pearson correlations ⫽ ⫺0.76 to ⫺0.55). Prey No. Prey taxon (Order: Family: Genus and species) length Field Observations: Feeding Ecology. We observed (cm) events 73 females with prey items in the Þeld (Table 1). Ten prey items were conspeciÞc males, which involve the Diptera (misc. ßies) 1 9 Hymenoptera: Apidae: Apis mellifera 1 33 possible confounding inßuence of sexual attraction on (honeybee) predation success. Females that were observed to can- Hymenoptera: Vespidae (unidentiÞed yellow 1 3 nibalize males did not signiÞcantly differ in pronotum jackets) length from females that were not observed to can- Hymenoptera: Vespidae: Polistes sp. (paper 2 2 wasp) nibalize males (mean ⫾ SE: 21.3 ⫾ 0.4 mm, n ⫽ 10 Hymenoptera (unidentiÞed wasp) 1 1 cannibalistic females; 20.9 ⫾ 0.2 mm, n ⫽ 63 noncan- Lepidoptera (unidentiÞed butterßy) 2 1 nibalistic females; t-test: t71 ⫽ 0.8, P ⬎ 0.4). Three prey Mantodea: Mantidae: Stagmomantis limbata 6 3 items were conspeciÞc females. In two of these cases, female Mantodea: Mantidae: Stagmomantis limbata 5 10 the femalesÕ pronotum lengths were known, and the male consumed females had shorter pronota than the con- Orthoptera: Acrididae: Melanoplus femurrubrum 2Ð3a 10 sumer females. For the 63 nonmale prey items, the (red-legged grasshopper) majority (46 events, or 73%) were 1 cm in length, Orthoptera: Tettigoniidae (unidentiÞed 3 1 katydid) making assumptions regarding normality of data dif- Þcult to evaluate. Thus, we used nonparametric sta- All prey items are insects. tistical tests on prey length data. a Two prey lengths observed: 2 cm (n ⫽ 1), 3 cm (n ⫽ 9).
96 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 1 Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Fig. 3. Predation events by adult females in the Þeld, 2008Ð2011. (a) Prey length (centimeters) by year. Sample sizes are given in parentheses; the Þrst number represents all prey, except conspeciÞc males; the second number is the number of conspeciÞc males. Black diamonds: mean (⫾SE) prey length, excluding conspeciÞc males. White diamonds: mean (⫾SE) prey length, including conspeciÞc males. (b) Female pronotum length (millimeters) and prey length (centimeters). White squares: conspeciÞc males (n ⫽ 10); black diamonds: all other prey (n ⫽ 63). Vertical dashed lines divide the 63 nonmale predation events into thirds, sorted by female pronotum length. differ from the other 2 yr (Kruskal-Wallis test with tied To examine the effect of female pronotum length on ranks: Hc ⫽ 14.32; df ⫽ 2; P ⬍ 0.001). diet breadth, we examined the sample variations in For all years combined, adult female pronotum prey length for the shortest-third and longest-third length and observed prey length were positively cor- females (indicated in Fig. 3b). The shortest-third fe- related for nonmale prey (Fig. 3b; Spearman correla- males ate shorter prey than the longest-third females tion: rs ⫽ 0.38; P ⬍ 0.01; n ⫽ 63). This correlation was (mean prey length ⫽ 1.1 cm and 2.1 cm, respectively; maintained when cannibalized males were included in Mann-Whitney test: U ⫽ 305; n1 ⫽ n2 ⫽ 21; P ⬍ 0.01). analysis (Spearman correlation: rs ⫽ 0.34; P ⬍ 0.01; n ⫽ The shortest-third females showed a narrower 73). Given the negative correlation between adult breadth in prey length than the longest-third females, emergence date and pronotum length, we checked for evinced by smaller variance (0.2 vs. 2.5 cm2, respec- a possible inßuence of the date of predation on prey tively; LeveneÕs test, normality of data not assumed: length. Date of predation failed to show a signiÞcant t40 ⫽ ⫺5.00; P ⬍ 0.001). The statistical signiÞcance of effect, with female pronotum length showing a positive this comparison did not change when the cannibalized relationship with prey length (generalized linear model: males were included in analysis (LeveneÕs test, nor- likelihood ratio 2 ⫽ 8.0; whole-model P ⬍ 0.05; prono- mality of data not assumed: t46 ⫽ ⫺2.78; P ⬍ 0.01). tum P ⬍ 0.01; date P ⬎ 0.9). Similar results were obtained Over all years, we measured the rate of abdominal when conspeciÞc males were included in analysis (gen- expansion for 50 young adult females. Pronotum eralized linear model: likelihood ratio 2 ⫽ 6.9; whole- length did not affect the rate of abdominal expansion, model P ⬍ 0.05; pronotum P ⬍ 0.01; date P ⬎ 0.5). but date of Þrst measurement had a signiÞcant effect,
February 2014 MAXWELL AND FRINCHABOY: INTRASPECIFIC SIZE VARIATION IN S. limbata 97 Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Fig. 4. Date of Þrst measurement of dorsoventral abdominal girth (millimeters) for newly emerged adult females in the Þeld and subsequent rate of abdominal expansion (millimeters per day ), 2008Ð2011. Numerical dates are Julian dates, where 1 August ⫽ 213. Crosses: 2008; black circles: 2009; triangles: 2010; white circles: 2011. Possible outliers are indicated by the letters A, B, and C. with females maturing later in the season showing Female Fecundity and Ootheca Mass. For female fe- greater rates of expansion (Fig. 4; ANCOVA: F5,44 ⫽ cundity (2006 and 2008), egg number increased with 3.14; whole-model P ⬍ 0.05; date P ⬍ 0.001; pronotum pronotum length (Fig. 6a; Multiple regression: F2,27 ⫽ P ⬎ 0.2; year P ⬎ 0.1). Exclusion of three outliers (A, 8.5; adjusted R2 ⫽ 0.34; whole-model P ⬍ 0.001; pro- B, and C) reduced the statistical signiÞcance of the notum length P ⬍ 0.01; year P ⬎ 0.7). For the mass of overall ANCOVA model, but the signiÞcance of the the Þrst ootheca (2010), ootheca mass increased with three variables did not change (ANCOVA: F5,41 ⫽ both pronotum length and body condition (Fig. 6b; 2.32; whole-model P ⫽ 0.061; date P ⬍ 0.01; pronotum Stepwise multiple regression: F2,23 ⫽ 16.8; adjusted P ⬎ 0.3; year P ⬎ 0.2). R2 ⫽ 0.56; whole-model P ⬍ 0.001; pronotum length Reproduction: Male Attraction, Natural Pairings, P ⬍ 0.001; body condition index ratio P ⬍ 0.01). We Fecundity, and Ootheca Mass. Field Experiment: Fe- obtained statistically equivalent results when using male Length and Male Attraction. In total, we found 56 males on the femalesÕ cages. We did not Þnd any males on the six empty control cages, so we excluded these empty cages from analysis. The average number of males per female was similar for short females (mean ⫾ SE ⫽ 2.6 ⫾ 0.9 males; n ⫽ 12) and long females (mean ⫾ SE ⫽ 2.1 ⫾ 0.8 males, n ⫽ 12; t-test: t22 ⫽ 0.4, P ⬎ 0.6). To examine possible nonindepen- dence of arriving males, we analyzed male arrivals as presence-absence data (i.e., presence ⫽ one or more males per check). The 56 males can be regarded as 33 independent occurrences; that is, exactly one male was found on 22 checks, with two or more males found on 11 checks. For these presence-absence data, the average number of occurrences of at least one male was similar for short females (mean ⫾ SE ⫽ 1.6 ⫾ 0.3 occurrences; n ⫽ 12) and long females (mean ⫾ SE ⫽ 1.2 ⫾ 0.4 occurrences, n ⫽ 12; t-test: t22 ⫽ 0.9, P ⬎ 0.3). Field Observations: Female and Male Pairings. For natural pairings in the Þeld, paired females tended to be longer than unpaired females, but this difference Fig. 5. Pronotum length (millimeters) of females in was not statistically signiÞcant in any year (Fig. 5). For natural pairings in Þeld, 2008Ð2011. Mean ⫾ SE plotted. a given year, sample sizes were 9 Ð36 paired females White diamonds: unpaired females (n ⫽ 71Ð178 per year). and 71Ð178 unpaired females. For the combined years Black diamonds: paired females (n ⫽ 9Ð36 per year). Un- 2009 Ð2011, adult age was known for 20 of the paired paired t-tests compare pronotum length of unpaired and females (mean ⫾ SE ⫽ 15 ⫾ 2 d; range ⫽ 4 Ð28 d). paired females within each year; NS: P ⬎ 0.05.
98 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 1 Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Fig. 6. Female pronotum length (millimeters) and measures of reproduction. (a) Fecundity (number of mature eggs in body at death). White circles: 2006 (n ⫽ 18). Black circles: 2008 (n ⫽ 12). (b) Mass of Þrst ootheca (g), 2010 (n ⫽ 26). regression residuals as the measure of female body more successful in subduing longer prey. Recent work condition. The average interval from copulation to on mantids in captivity has begun to address this ques- Þrst ootheca was 12 d (SE ⫽ 2 d; n ⫽ 26). tion but from different perspectives. In the mantid T. aridifolia, Whitman and Vincent (2008) reported a positive correlation between prey size and adult man- Discussion tid mass, combining attacks by males and the larger The current study reveals several consequences of females. However, recent interspeciÞc comparisons of intraspeciÞc variation in body size, particularly with mantid species in captivity indicate that target sizes respect to prey size, prey breadth, female fecundity, that elicit strike behaviors are not related to predator and ootheca mass. Longer females were found with species size (Prete et al. 2011, 2013). Clearly, more longer prey in the Þeld, as well as greater breadth of experimental work on the effects of predator size on prey size, than shorter females. These results are in attack rates and capture success is warranted, with agreement with empirical studies on intraspeciÞc vari- attention to interspeciÞc and intraspeciÞc variation. ation in predator size, prey size, and diet breadth in One form of predation, intraspeciÞc predation many taxa, including insects (Schoener 1971, Wilson (cannibalism), was observed in the current study, in 1975, Cohen et al. 1993, Costa 2009), as well as Hol- two contexts: intersexual and intrasexual. With regard lingÕs predictions developed for mantids based on to intersexual cannibalism, females that were ob- foreleg morphology (Holling 1964, Holling et al. served to cannibalize males in the Þeld were of similar 1976). In the current study, we note that it is difÞcult length to noncannibalistic females. This result aligns to ascertain whether longer females were more likely with captive work on S. limbata and P. albofimbriata to strike at longer prey, or whether strike rates were (Barry et al. 2008, Barry 2010, Maxwell et al. 2010b). constant across female sizes and longer females were In these studies, poorly fed females were more likely
February 2014 MAXWELL AND FRINCHABOY: INTRASPECIFIC SIZE VARIATION IN S. limbata 99 to cannibalize males than were well-fed females, while feeding regime has been manipulated, with female neither female pronotum length nor male pronotum length being standardized or left unexamined (e.g., length inßuenced the occurrence of cannibalism. Fur- Lelito and Brown 2006, 2008; Maxwell et al. 2010a,b; thermore, Maxwell et al. (2010b) failed to determine Barry 2010; Barry et al. 2010). Therefore, it is difÞcult that the degree of length dimorphism between paired to draw generalizations about the importance of females and males inßuenced the occurrence of can- female length on male behavior at this stage. Further- nibalism. However, it may be premature to conclude more, research on cannibalistic arachnids reveals male that female length does not inßuence the occurrence preferences for large females (Prenter et al. 1994, of cannibalism in mantids. In spiders, varied results Hoeßer 2007), with at least one species preferring have been reported, with many species showing a small females (Moya-Laraño et al. 2003). Thus, female positive relationship between mate size dimorphism length, in addition to female feeding regime, merits and cannibalism, whereas others show no, or even a more consideration in studies of mantid mating be- negative, relationship (reviewed in Wilder and Ryp- havior. stra 2008). Female hunger also plays an important role In the current study, our expectations were con- Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 in the occurrence of cannibalism in some spiders, Þrmed regarding female length and measures of egg whereas other spider species show little effect of fe- production, with longer females having higher fecun- male hunger (reviewed in Wilder et al. 2009). Given dity and laying heavier oothecae. These results are this range of results across spider species, more con- consistent with the many empirical studies that dem- sideration of absolute length and sexual size dimor- onstrate that female fecundity increases with body phism is needed in studies of mating behavior and size within insect species (Honek̆ 1993, Whitman cannibalism in mantids. 2008) and other taxa (Peters 1983, Roff 1992). In Intrasexual cannibalism, speciÞcally cannibalism insects, it is reasonable to assume that female abdom- among adult females, was observed in the current inal length increases with body length or pronotum study, which is consistent with previous observations length, thereby enabling longer females to hold more in the Þeld for S. limbata and other mantids (Hurd et eggs. Preziosi et al. (1996) directly tested this assump- al. 1994, Maxwell and Eitan 1998). This form of can- tion in the water strider Aquarius remigis Say, dem- nibalism might reßect overall food limitation in the onstrating that larger females showed higher fecun- Þeld or individual variation in foraging success (Fagan dities because they had larger abdomens. Other and Hurd 1991). Alternatively, interfemale cannibal- authors indicate that larger abdomens may also pro- ism might be a manifestation of interference compe- vide increased space for accessory materials or organs tition over ambush sites, or possibly over sites for involved in egg production and oviposition (Wickman “calling” (pheromone emission) and oviposition. If and Karlsson 1989, Roff 1992). This consideration is such interference competition is occurring, then lon- especially relevant for mantids, which lay eggs in ooth- ger females are likely to overpower and consume shorter females (Fox 1975, Polis 1981), similar to how ecae, requiring the production and storage of the en- size difference between mantid nymphs reliably dif- veloping spume material. In mantids, longer females ferentiates the cannibal from the victim (Hurd 1988). may have greater abdominal space for more eggs as Although observations of femaleÐfemale cannibalism well as more oothecal spume. in the current study were too few to do a rigorous In addition to increased fecundity and ootheca analysis of the effect of relative length, we note that mass, longer females may have longer reproductive the cannibal female was longer than the victim female life spans in nature. In the current study, longer fe- in the two cases where relative lengths were known. males consistently matured earlier in the season, as The possibility of cannibalism in the context of inter- seen in the strong negative relationship between date female competition deserves further investigation of adult emergence and female pronotum length in all (Polis 1981, Maxwell 1999). years. These early emergence dates are expected to Regarding female length and male attraction, we contribute to longer reproductive life spans for longer held no a priori expectations regarding the effect of females. Early-maturing females, however, appear to female length, as males may prefer longer females for emerge during a more food-limited part of the season, increased fecundity, or shorter females for reduced as feeding rates in nature (measured by rates of ab- risk of cannibalism (Bonduriansky 2001, Wilder and dominal expansion) were higher later in the season. Rypstra 2008). Our experimental and observational Although smaller, later-emerging females might have results on mating behavior fail to reveal an effect of less time available for feeding, mating, and egg-laying, female length. In the Þeld experiment on male attrac- they might be reproductively active during a more tion, longer females did not attract more males than food-abundant part of the season. Late-season emer- shorter females did. Given that the females were ob- gence at small size, then, may involve the beneÞt of scured with cotton gauze, it could be argued that this increased prey abundance. While this might be a for- result suggests that pheromone emission was not af- tunate ecological coincidence in this study system, fected by female length, but the males might still favor small size can confer several additional advantages, as longer or shorter females once they make visual con- many empirical studies demonstrate that small size tact. The natural pairings in the Þeld, however, fail to can be less costly in terms of energetic and metabolic indicate a length preference expressed by males. In demands, movement costs, and susceptibility to water studies of male mating behavior in mantids, female and heat loss (Peters 1983, Hone and Benton 2005).
100 ENVIRONMENTAL ENTOMOLOGY Vol. 43, no. 1 The current study points to several ecological con- Dussé, K., and L. E. Hurd. 1997. Food limitation reduces sequences of large adult female size in the mantid S. body length in mantid nymphs, Tenodera sinensis Sau- limbata, particularly larger prey size, broader diet, and ssure (Mantodea: Mantidae): implications for Þtness. increased egg production. While longer females may Proc. Entomol. Soc. Wash. 99: 490 Ð 493. have longer reproductive life spans in nature, shorter Ehrmann, R. 2002. Mantodea: Gottesanbeterinnen der Welt. Natur und Tier, Münster, Germany. later-emerging females may reap the ecological ben- Fagan, W. F., and L. E. Hurd. 1991. Late season food level, eÞt of feeding at higher rates as adults. Furthermore, cannibalism, and oviposition in adult mantids (Or- shorter females were not necessarily at a disadvantage thoptera: Mantidae): sources of variability in a Þeld ex- in attracting males. Thus, while larger adult size ap- periment. Proc. Entomol. Soc. Wash. 93: 956 Ð961. pears to provide several Þtness beneÞts, large size Fagan, W. F., and L. E. Hurd. 1994. Hatch density variation might not effect other Þtness components, or might of a generalist arthropod predator: population conse- entail hidden costs. quences and community impact. Ecology 75: 2022Ð2032. Fagan, W. F., M. D. Moran, J. J. Rango, and L. E. Hurd. 2002. Community effects of praying mantids: a meta-analysis of Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Acknowledgments the inßuences of species identity and experimental de- sign. Ecol. Entomol. 27: 385Ð395. We thank Goggy Davidowitz, Larry Hurd, Chad Johnson, Fox, L. R. 1975. Cannibalism in naturalism populations. Doug Whitman, and anonymous reviewers for constructive Annu. Rev. Ecol. Syst. 6: 87Ð106. comments. Our gratitude extends to the following for assis- Hoefler, C. D. 2007. Male mate choice and size-assortative tance in the Þeld: Kevin Gallego, Ted Maxwell, Jamie Mer- pairing in a jumping spider, Phidippus clarus. Anim. Be- edith, Scott Seyfried, Nate Seyfried, and Beth Sullivan. We hav. 73: 943Ð954. additionally thank the staff of the Owens Valley Laboratory Holling, C. S. 1964. The analysis of complex population pro- (University of California, White Mountain Research Station) cesses. Canad. Entomol. 96: 335Ð347. for accommodations and logistical support. National Univer- Holling, C. S., R. L. Dunbrack, and L. M. Dill. 1976. Pred- sity provided funding for this research. This research con- ator size and prey size: presumed relationship in the forms to the institutional requirements of National Univer- mantid Hierodula coarctata Saussure. Canad. J. Zool. 54: sity. 1760 Ð1764. Hone, D.W.E., and M. J. Benton. 2005. The evolution of large size: how does CopeÕs Rule work? Trends Ecol. Evol. References Cited 20: 4 Ð 6. Andersson, M. 1994. Sexual selection. Princeton University Honĕk, A. 1993. IntraspeciÞc variation in body size and fe- Press, Princeton, NJ. cundity in insects: a general relationship. Oikos 66: 483Ð Barry, K. L. 2010. Inßuence of female nutritional status on 492. mating dynamics in a sexually cannibalistic praying man- Hurd, L. E. 1988. Consequences of divergent egg phenol- tid. Anim. Behav. 80: 405Ð 411. ogy to predation and coexistence in two sympatric, con- Barry, K. L., G. I. Holwell, and M. E. Herberstein. 2008. generic mantids (Orthoptera: Mantidae). Oecologia 76: Female praying mantids use sexual cannibalism as a for- 549 Ð552. aging strategy to increase fecundity. Behav. Ecol. 19: Hurd, L. E. 1999. Ecology of praying mantids, pp. 43Ð 60. In 710 Ð715. F. R. Prete, H. Wells, P. H. Wells, and L. E. Hurd (eds.), Barry, K. L., G. I. Holwell, and M. E. Herberstein. 2010. The Praying Mantids. Johns Hopkins University Press, Multimodal mate assessment by male praying mantids in Baltimore, MD. a sexually cannibalistic mating system. Anim. Behav. 79: Hurd, L. E., and R. M. Eisenberg. 1984. Experimental den- 1165Ð1172. sity manipulations of the predator Tenodera sinensis (Or- Birkhead, T. R., K. E. Lee, and P. Young. 1988. Sexual can- thoptera: Mantidae) in an old-Þeld community. I. Mor- nibalism in the praying mantis Hierodula membranacea. tality, developmental and dispersal of juvenile mantids. J. Behaviour 106: 112Ð118. Anim. Ecol. 53: 269 Ð281. Bonduriansky, R. 2001. The evolution of male mate choice Hurd, L. E., and R. M. Eisenberg. 1989a. A mid-summer in insects: a synthesis of ideas and evidence. Biol. Rev. 76: comparison of sizes and growth rate. Proc. Entomol. Soc. 305Ð339. Wash. 91: 51Ð54. Charnov, E. L. 1976. Optimal foraging: attack strategy of a Hurd, L. E., and R. M. Eisenberg. 1989b. Temporal distri- mantid. Am. Nat. 110: 141Ð151. bution of hatching times in three sympatric mantids Chown, S. L., and Gaston, K. J. 2010. Body size variation in (Mantodea: Mantidae) with implications for niche sep- insects: a macroecological perspective. Biol. Rev. 85: 139 Ð aration and coexistence. Proc. Entomol. Soc. Wash. 91: 169. 55Ð58. Claessen, D., A. M. de Roos, and L. Persson. 2000. Dwarfs Hurd, L. E., and R. M. Eisenberg. 1990. Arthropod commu- and giants: cannibalism and competition in size-struc- nity responses to manipulation of a bitrophic predator tured populations. Am. Nat. 155: 219 Ð237. guild. Ecology 76: 2017Ð2114. Claessen, D., C. van Oss, A. M. de Roos, and L. Persson. 2002. Hurd, L. E., R. M. Eisenberg, W. F. Fagan, K. J. Tilmon, W. E. The impact of size-dependent predation on population Snyder, K. S. Vandersall, S. G. Datz, and J. D. Welch. dynamics and individual life history. Ecology 83: 1660 Ð 1994. Cannibalism reverses male-biased sex ratio in adult 1675. mantids: female strategy against food limitation? Oikos Cohen, J. E., S. L. Pimm, P. Yodzis, and J. Saldaña. 1993. 69: 193Ð198. Body sizes of animal predators and animal prey in food Hurd, L. E., and I. H. Rathet. 1986. Functional response and webs. J. Anim. Ecol. 62: 67Ð78. success in juvenile mantids. Ecology 67: 163Ð167. Costa, G. C. 2009. Predator size, prey size, and dietary niche Iwasaki, T. 1991. Predatory behavior of the praying mantis, breadth relationships in marine predators. Ecology 90: Tenodera aridifolia. II. Combined effect of prey size and 2014 Ð2019. predator size on the prey recognition. J. Ethol. 9: 77Ð 81.
February 2014 MAXWELL AND FRINCHABOY: INTRASPECIFIC SIZE VARIATION IN S. limbata 101 Jakob, E. M., S. D. Marshall, and G. W. Uetz. 1996. Estimat- Petrie, M. 1983. Female moorhens compete for small fat ing Þtness: a comparison of body condition indices. Oikos males. Science 220: 413Ð 415. 77: 61Ð 67. Polis, G. A. 1981. The evolution and dynamics of intraspe- Lawrence, S. E. 1992. Sexual cannibalism in the praying ciÞc predation. Annu. Rev. Ecol. Syst. 12: 225Ð251. mantid, Mantis religiosa: a Þeld study. Anim. Behav. 43: Polis, G. A. 1988. Exploitation competition and the evolution of 569 Ð583. interference, cannibalism, and intraguild predation in age/ Lelito, J. P., and W. D. Brown. 2006. Complicity or conßict size-structured populations, pp. 185Ð202. In B. Ebenman and over sexual cannibalism? Male risk taking in the praying L. Persson (eds.), Size-Structured Populations: Ecology and mantis Tenodera aridifolia sinensis. Am. Nat. 168: 263Ð269. Evolution. Springer, Berlin, Germany. Lelito, J. P., and W. D. Brown. 2008. Mate attraction by Prenter, J., R. W. Elwood, and W. I. Montgomery. 1994. females in a sexually cannibalistic praying mantis. Behav. Assessments and decisions in Metellina segmentata (Ara- Ecol. Sociobiol. 63: 313Ð320. neae: Metidae): evidence of a pheromone involved in Matsura, T., and K. Morooka. 1983. Inßuences of prey den- mate guarding. Behav. Ecol. Sociobiol. 35: 39 Ð 43. sity on fecundity in a mantis, Paratenodera angustipennis Prete, F. R., S. Dominguez, J. L. Komito, R. Theis, J. J. (S.). Oecologia 56: 306 Ð312. Dominguez, L. E. Hurd, and G. J. Svenson. 2013. Ap- Downloaded from https://academic.oup.com/ee/article/43/1/91/387618 by guest on 08 February 2022 Matsura, T., T. Inoue, and Y. Hosomi. 1975. Ecological stud- ies of a mantid, Paratenodera angustipennis de Saussure. I. petitive responses to computer-generated visual stimuli Evaluation of the feeding condition in natural habitats. by female Rhombera basalis, Deroplatys lobata, Hierodula Res. Popul. Ecol. 17: 64 Ð76. membranacea, and Miomantis sp. (Insecta: Mantodea). Maxwell, M. R. 1998. Lifetime mating opportunities and J. Insect Behav. 26: 261Ð282. male mating behaviour in sexually cannibalistic praying Prete, F. R., J. L. Komito, S. Dominguez, G. Svenson, L. Y. mantids. Anim. Behav. 55: 1011Ð1028. López, A. Guillen, and N. Bogdanivich. 2011. Visual stimuli Maxwell, M. R. 1999. Mating behavior, pp. 69 Ð 89. In F. R. that elicit appetitive behaviors in three morphologically Prete, H. Wells, P. H. Wells, and L. E. Hurd (eds.), The distinct species of praying mantis. J. Comp. Physiol. A Neu- Praying Mantids. Johns Hopkins University Press, Balti- roethol. Sens. Neural Behav. Physiol.197: 877Ð894. more, MD. Preziosi, R. F., D. J. Fairbairn, D. A. Roff, and J. M. Brennan. Maxwell, M. R., and O. Eitan. 1998. Range expansion of an 1996. Body size and fecundity in the waterstrider Aquar- introduced mantid Iris oratoria and niche overlap with a ius remigis: a test of DarwinÕs fecundity advantage hy- native mantid Stagmomantis limbata (Mantodea: Manti- pothesis. Oecologia 108: 424 Ð 431. dae). Ann. Entomol. Soc. Am. 91: 422Ð 429. Rice, W. R. 1989. Analyzing tables of statistical tests. Evo- Maxwell, M. R., K. L. Barry, and P. M. Johns. 2010a. Exam- lution 43: 223Ð225. inations of female pheromone use in two praying mantids, Roberts, R. A. 1937. Biology of the bordered mantid, Stag- Stagmomantis limbata and Tenodera aridifolia sinensis momantis limbata Hahn (Orthoptera, Mantidae). Ann. (Mantodea: Mantidae). Ann. Entomol. Soc. Am. 103: 120 Ð Entomol. Soc. Am. 30: 96 Ð109. 127. Roff, D. A. 1992. The evolution of life histories. Chapman Maxwell, M. R., K. M. Gallego, and K. L. Barry. 2010b. Ef- and Hall, New York, NY. fects of female feeding regime in a sexually cannibalistic Schoener, T. W. 1971. Theory of feeding strategies. Annu. mantid: fecundity, cannibalism, and male response in Rev. Ecol. Syst. 2: 369 Ð 404. Stagmomantis limbata (Mantodea). Ecol. Entomol. 35: Stearns, S. C. 1992. The evolution of life histories. Oxford 775Ð787. University Press, Oxford, United Kingdom. McCoy, M. W., B. M. Bolker, K. M. Warkentin, and J. R. Whitman, D. W. 2008. The signiÞcance of body size in the Vonesh. 2011. Predicting predation through prey ontog- eny using size-dependent functional response models. Orthoptera: a review. J. Orthop. Res. 17: 117Ð134. Am. Nat. 177: 752Ð766. Whitman, D. W., and S. Vincent. 2008. Large size as an Moran, M. D., and L. E. Hurd. 1997. Relieving food limita- antipredator defense in an insect. J. Orthop. Res. 17: tion reduces survivorship of a generalist predator. Ecol- 353Ð371. ogy 78: 1266 Ð1270. Wickman, P. O., and B. Karlsson. 1989. Abdomen size, body Moran, M. D., T. P. Rooney, and L. E. Hurd. 1996. Top- size and the reproductive effort of insects. Oikos 56: down cascade from a bitrophic predator in an old-Þeld 209 Ð214. community. Ecology 77: 2219 Ð2227. Wilder, S. M., and A. L. Rypstra. 2008. Sexual size dimor- Moya-Laraño, J., R. Macı́as-Ordóñez, W. U. Blanckenhorn, phism predicts the frequency of sexual cannibalism and C. Fernández-Montraveta. 2008. Analysing body within and among species of spiders. Am. Nat. 172: 431Ð condition: mass, volume or density? J. Anim. Ecol. 77: 440. 1099 Ð1108. Wilder, S. M., A. L. Rypstra, and M. A. Elgar. 2009. The Moya-Laraño, J., J. Pascual, and D. H. Wise. 2003. Mating importance of ecological and phylogenetic conditions for patterns in late-maturing female Mediterranean tarantu- the occurrence and frequency of sexual cannibalism. las may reßect the costs and beneÞts of sexual cannibal- Annu. Rev. Ecol. Evol. Syst. 40: 21Ð39. ism. Anim. Behav. 66: 469 Ð 476. Wilson, D. S. 1975. The adequacy of body size as a niche Paine, R. T. 1976. Size-limited predation: an observational difference. Am. Nat. 109: 769 Ð784. and experimental approach with the Mytilus-Pisaster in- Zar, J. H. 2010. Biostatistical analysis, 5th ed. Pearson Pren- teraction. Ecology 57: 858 Ð 873. tice Hall, Upper Saddle River, NJ. Peters, R. H. 1983. The ecological implications of body size. Cambridge University Press, Cambridge, United King- dom. Received 28 October 2012; accepted 20 October 2013.
You can also read