Appl. Entomol. Zool. 38 (3): 313–319 (2003) Allometry of male genitalia in a lepidopteran species, Ostrinia latipennis (Lepidoptera: Crambidae) Suguru OHNO,1,2,*,† Sugihiko HOSHIZAKI,2 Yukio ISHIKAWA,2 Sadahiro TATSUKI2 and Shin-ichi AKIMOTO1 1 Laboratory of Systematic Entomology, Graduate School of Agriculture, Hokkaido University; Sapporo 060–8589, Japan 2 Laboratory of Applied Entomology, Graduate School of Agricultural and Life Sciences, The University of Tokyo; Tokyo 113–8657, Japan (Received 6 December 2002; Accepted 31 March 2003) Abstract In species of several insect orders and spiders, it has been shown that the size of male genitalia relative to body size decreases as the body becomes larger (negative allometry), while the relative size of other morphological traits tends to be constant. Such a contrast between genital and somatic traits suggests stabilizing sexual selection on male geni- talia: males with small or large genitalia are prone to fail to inseminate females due to incompatibility of their geni- talia. In the present study, we tested the contrast between genital and somatic traits for males of a lepidopteran insect, Ostrinia latipennis. We examined allometry of five genital and 11 somatic traits for each of three local populations of O. latipennis. Of the 15 allometric slopes for genital traits, 14 showed significantly negative allometry, whereas none of the 33 slopes for somatic traits represented negative allometry. These results showed that the size of male genitalia in O. latipennis is more stable than the size of somatic traits against changes in body size. This study supports Eber- hard et al.’s (1998) hypothesis which states that the low genital allometry in insects and spiders is caused by sexual se- lection. Based on currently available information on genital morphology and sexual communication in O. latipennis, it seems unlikely that the lock-and-key hypothesis is responsible for the stable genital size in this species. Key words: Allometry; male genitalia; stabilizing selection; Lepidoptera; Ostrinia latipennis Eberhard et al. (1998) examined the allometry of INTRODUCTION the sizes of male genitalia and other body parts in Allometry, i.e., the scaling relationship between 20 species of insects (including those belonging to the size of a given morphological trait and body Coleoptera, Diptera, Hymenoptera, Hemiptera, size at a particular developmental stage of an or- Dermaptera and Odonata) and spiders. They found ganism, has been focused on in studies of morpho- that allometric slopes for male genitalia were con- logical evolution. The observed allometric slope (a sistently lower than 1.0 and lower than the slopes slope in log-log regression on an indicator of body for the other body parts. Based on these findings, size) is presumably a consequence of selection that they suggested that within a species sexual selec- favors one slope over the others (Eberhard et al., tion favors males with genitalia of standard sizes. 1998). In male animals, secondary sexual traits Several studies on beetles, dipterans and a spider which are probably under sexual selection tend to have similarly detected negative allometry for male become relatively larger in individuals with a larger genital size (Wheeler et al., 1993; Palestrini et al., body, resulting in positive allometry (allometric 2000; Schmitz et al., 2000; Uhl and Vollrath, 2000; slopes being larger than 1.0) (e.g., Alatalo et al., Iwahashi, 2001; Iwahashi and Routhier, 2001; Ta- 1988; Petrie, 1988, 1992; Green, 1992; Emlen and tsuta et al., 2001). The lower allometry for male Nijhout, 2000). In contrast, ordinary morphologi- genital size appears to be common among insects cal traits probably not under sexual selection show and spiders. To date, however, studies on genital al- isometry (allometric slopes around 1.0) (Harvey lometry have been concentrated on beetles, dipter- and Pagel, 1991). ans and spiders (27 out of 31 species examined * To whom correspondence should be addressed at: E-mail: [email protected] † Present address: Fruit-fly Eradication Project Office, Okinawa Prefectural Government; 123 Maji, Naha, Okinawa 902–0072, Japan 313 314 S. OHNO et al. belong to these three groups). It is thus premature to conclude that the negative allometry of male genitalia is common in insects, without examining other major taxa. We examined the allometry of genital and non- genital morphological traits of a crambid moth, the Far Eastern knotweed borer, Ostrinia latipennis (Warren). This is the first study examining allome- try of genital size in Lepidoptera. We discuss the significance of the genital allometry in O. latipen- nis along with hypotheses on the evolution of ani- mal genitalia. MATERIALS AND METHODS Insects and morphological measurements. Os- trinia latipennis (Warren) (Crambidae: Pyrausti- nae) is distributed in northern Far East Asia includ- ing northeastern Japan (Mutuura and Munroe, 1970; Ohno, 1998). Larvae of O. latipennis feed mainly on knotweeds, Reynoutria sachalinensis and R. japonica (Polygonaceae) (Ishikawa et al., 1999; Ohno, 2000). To examine the geographic variation in male morphology among seven local populations of O. latipennis, Ohno (1998) con- ducted multivariate analyses on 16 morphological traits (11 somatic and five genital). Of these seven Fig. 1. Measurements of somatic and genital morphology populations, three (Otoineppu-Vil., Hokkaido Is., of Ostrinia latipennis (redrawn from Ohno, 1998). A, non- 44.7°N 142.3°E; Tazawako-Town, Akita-Pref., genital traits. FWL, forewing length; HWL, hindwing length; FCL, fore coxa length; FFL, fore femur length; FTL, fore tibia Honshu Is., 39.8°N 140.8°E; Unazuki-Town, length; MFL, mid femur length; MTL, mid tibia length; EPL, Toyama-Pref., Honshu Is., 36.9°N 137.6°E), with epiphysis length; OSPL, outer mid-tibial spur length; ISPL, sufficient sample size (over 30 for each popula- inner mid-tibial spur length; LP2L, labial palpus 2nd segment tion), were used for the present analyses. length. B, genital traits. VL, valva length; SL, sacculus length; The 11 somatic traits analyzed in the present HL, harpe length; CW, clasper width; AL, aedeagus length. study were the lengths of the following body parts: The overall body size was represented by the first princi- pal component (PC1) score based on the 16 traits (see Mate- fore wing, hind wing, fore coxa, fore femur, fore rials and Methods) where PC150.095FWL10.074HWL1 tibia, mid femur, mid tibia, epiphysis (leaf-shaped 0.262FCL10.333FFL10.284FTL10.379MFL10.411MTL1 projection on the fore tibia), outer mid-tibial spur, 0.207EPL10.225OSPL10.419ISPL10.315LP2L10.131VL1 inner mid-tibial spur and the 2nd segment of the 0.097SL10.090HL10.040CW10.094AL. labial palpus (Fig. 1). The five genital traits ana- lyzed included four parts of the grasping organ, body-size indicator may have a closer ontogenetic i.e., valva length, sacculus length, harpe length and and functional relationship with other somatic clasper width, and one part of the intromittent traits than with genital traits, and the use of a dif- organ, aedeagus length (Fig. 1). The sizes of the ferent body-size indicator may give rise to different traits (in mm) are shown in Table 1. See Ohno results (Green, 1999). Thus, we used the principal (1998) for further details on insect collection and component 1 (PC1) as an overall measure for body measurement acquisition. size. The PC1 score was calculated from the vari- Several previous studies on genital allometry ance-covariance matrix of the raw measurements used a single somatic trait as the indicator of body on the 16 body parts. In calculations of the PC1 size (e.g., width of pronotum or thorax). Such a scores, the constant was set at zero to prevent Table 1. Allometric slopes and other features of phenotypic variation for 16 morphometric traits of Ostrinia latipennis males from three localities (Otoineppu, Tazawako and Unazuki) in Japana Otoineppu (n535) Tazawako (n538) Unazuki (n534) Trait Mean category (mm) Slope CV r CV9 Slope CV r CV9 Slope CV r CV9 Somatic Fore wing length 14.22 0.94 ns 7.40 0.91* 3.09 1.13 ns 6.64 0.97* 1.63 1.13 ns 5.46 0.95* 1.78 Allometry ofMaleGenitaliain Hind wing length 10.68 0.95 ns 7.64 0.90* 3.33 1.15 ns 6.84 0.95* 2.08 1.27* 6.17 0.94* 2.13 Fore coxa length 2.11 1.06 ns 7.90 0.96* 2.24 1.12 ns 6.70 0.95* 2.13 1.27 ns 6.46 0.90* 2.78 Fore femur length 2.84 0.99 ns 7.44 0.95* 2.22 1.10 ns 6.43 0.97* 1.51 1.10 ns 5.39 0.93* 1.96 Fore tibia length 1.35 1.04 ns 8.41 0.89* 3.84 1.33* 8.09 0.94* 2.86 1.21 ns 6.19 0.89* 2.87 Mid femur length 3.33 0.97 ns 7.23 0.96* 2.12 1.07 ns 6.20 0.98* 1.30 1.12 ns 5.36 0.96* 1.54 Mid tibia length 3.50 1.02 ns 7.86 0.93* 2.90 1.15 ns 6.92 0.94* 2.30 1.17 ns 5.93 0.91* 2.52 Epiphysis length 0.64 0.88 ns 7.76 0.85* 4.06 1.04 ns 7.47 0.81* 4.43 1.01 ns 6.53 0.72* 4.53 Outer mid-tibial spur length 0.75 1.46 ns 14.73 0.69* 10.73 1.20 ns 9.00 0.74* 6.02 0.41 ns 8.40 0.25 ns — Inner mid-tibial spur length 1.26 1.22 ns 10.34 0.83* 5.73 0.83 ns 6.25 0.74* 4.18 1.10 ns 7.03 0.73* 4.83 Labial palpus 2nd seg.