Is the Sex Communication of Two Pyralid Moths,Plodia interpunctella and Ephestia kuehniella,under Circadian Clock Regulation?

Abstract

Females of the Indian meal moth, Plodia interpunctella, and females of the Mediterranean flour month, Ephestia kuehniella (both Lepidoptera: Pyralidae), exhibit daily rhythms in calling behavior. The peak in P. interpunctella calling occurs at dusk, whereas E. kuehniella calls preferentially at dawn. This behavior turned arrhythmic in P. interpunctella females in constant darkness (DD) and remained arrhythmic in constant light (LL), whereas E. kuehniella females showed a persistent rhythm in DD and suppression of the behavior in LL, indicating regulation by a circadian clock mechanism. The rhythm of male locomotor activity corresponded well with the sexual activity of females, reaching the peak at dusk in P. interpunctella and at dawn in E. kuehniella. An immunohistochemical study of the pheromone biosynthesis activating neuropeptide, corazonin, and pigment dispersing factor revealed distinct sets of neurons in the brain-subesophageal complex and in the neurohemal organs of the 2 species.

Daily rhythms in sexually relevant behaviors are commonly regulated by circadian clocks in insects. In most moth species, females rely on sex pheromones to attract males before mating, and several mechanisms have been proposed to prevent costly interspecific attraction. Several studies have focused on temporal isolation of sexual activity in closely related insect species. A prime example is the Queensland fruit flies Bactrocera tryoni and Bactrocera neohumeralis, which occur as sympatric species even though laboratory hybrids, including all backcrosses, are fully fertile. Whereas B. neohumeralis mates at high light intensity and during the midday, B. tryoni prefers mating at low light intensities and at dusk. This difference in behavior causes practically complete reproductive isolation (Smith, 1979).

Another example is the sympatric noctuid moths Spodoptera latifascia and Spodoptera descoinsi; they share their main sex pheromone component but the latter species is sexually active during the second half of the scotophase and the former species is active during the early scotophase (Monti et al., 1995). Similar differences have been described in the closely related Lymantria dispar and Lymantria monacha (Lepidoptera: Lymantriidae). These species also share the same main pheromone component, but the daily peak of males trapped with pheromone baits shows that mate search occurs at different hours of the day (Giebultowicz and Zdarek, 1996). These distinct differences between closely related species suggest an adaptive pressure on the temporal phase of sexually relevant daily rhythms.

This study focuses on 2 species of pyralid moths, the Mediterranean flour moth, Ephestia kuehniella, and the Indian meal moth, Plodia interpunctella (both Lepidoptera: Pyralidae), which are cosmopolitan pests that feed on a wide variety of stored products. These species are almost exclusively found indoors and have most likely been dispersed from their original habitats by transportation of food products (Palm, 1986). Like the majority of moth species, E. kuehniella and P. interpunctella females rely on sex pheromones to attract males before mating, and several sex pheromone components have been identified in both species (e.g., Zhu et al., 1999; Ryne, 2001). Like the mentioned Lymantria and Spodoptera species, E. kuehniella and P. interpunctella share the same main pheromone component, (Z,E)-9,12-tetradecadienyl acetate (Z,E-9,12-14:OAc), which is also used by several other closely related species, such as Ephestia elutella and Cadra cautella (Kuwahara and Casida, 1973). As this compound alone is sufficient to attract males (Zhu et al., 1999; Kuwahara and Hara, 1971) and several species may occur in sympathy, mechanisms to avoid interspecies attraction have likely evolved even though the 2 species never crossbreed. The primary aim of this study was therefore to compare circadian rhythms of sexually relevant behavior in E. kuehniella and P. interpunctella to find indications of divergent phase shifts in behavioral activity.

Immunohistochemical investigation was focused on neuropeptide pheromone biosynthesis activating neuropeptide (PBAN), corazonin (Crz), and pigment dispersing factor (PDF), supposedly involved in the circadian signal output regulating the daily activities. PBAN is a 33 or 34 amino acid peptide stimulating sex pheromone production in female moths and plays a key role in temporal regulation of pheromone communication (Raina et al., 1989). The undecapeptide corazonin was discovered by Veenstra (1989) as a cardioaccelerator in cockroaches and subsequently was found in other insects including moth species Bombyx mori (Hua et al., 2000), Galleria mellonella (Hansen et al., 2001), and Manduca sexta (Wise et al., 2002). Crz-immunoreactivity (Crz-ir) in neurons that express clock proteins (Wise et al., 2002; Qi-Mao et al., 2003) links it to circadian output pathway. PDF is an insect homologue of the pigment dispersing hormone that was isolated as a regulator of pigmentation in crustaceans (Rao and Riehm, 1988) and functions as a key mediator in the circadian clock of Drosophila melanogaster (see Taghert and Shafer, 2006).

We studied daily rhythms in female calling behavior (pheromone release) of both species under a light-dark cycle (LD), continuous darkness (DD), and continuous light (LL). The results confirm not only temporally separate peaks in sexual activity but also major differences in their responses to constant light conditions. We also present comparative data of PBAN-immunoreactivity (PBAN-ir), Crz-ir, and PDF-immunoreactivity (PDF-ir) in the brain-subesophageal complex and in the neurohemal organs of the 2 moth species.

Material and Methods

Insects

For behavioral experiments, we used laboratory cultures of E. kuehniella and P. interpunctella maintained at the Department of Ecology, Lund University (for their origin, diet, and rearing conditions, see Zhu et al., 1999; Olsson et al., 2006). For immunocytological experiments, we used laboratory cultures of both species maintained at the Institute of Entomology BC ASCR in Ceske Budejovice (see Vítková et al., 2007, and references therein).

Calling Behavior in LD, DD, and LL

Prior to emergence, female pupae were transferred to individual transparent plastic cups (height: 8 cm, diameter: 6.5 cm) and kept in separate climate cabinets in a 17-h:7-h LD cycle (23 °C, 60% relative humidity). Adult females were transferred to continuous darkness or light just after emergence for observation in DD or LL. Observations of individual calling behavior began the day after emergence, every hour for 7 consecutive days. Female adults were regarded as calling when their abdomens were elevated, accompanied by extrusion of the ovipositor. In darkness, observations were performed with a red flashlight.

Locomotor Activity

Locomotor activity of individual adult males placed in a 9-cm Petri dish was recorded under LD, DD, or LL at 25 °C by an automated system previously used to record behavior of firebugs (Hodková et al., 2003) and house flies (Codd et al., 2007). The activity record (15- or 30-min bins) was analyzed using Flytoolbox (Levine et al., 2002) in Matlab software (Mathworks, Inc., Natick, MA). The activity of 25-30 males recorded for 7 days under different LD regimes was averaged and plotted in Excel (Microsoft, Redmond, WA).

Immunocytochemistry

Heads of adults were collected at ZT 4 (the mid-photophase) and ZT 16 (the mid-scotophase). At least 4 females and 2 males at each time point were tested. Dim red light (660-670 nm wavelengths) was used for dissecting brains during scotophase. Immunocytochemistry was performed on cephalic ganglia of adult males and females of P. interpunctella and E. kuehniella as described elsewhere (Závodská et al., 2009) and outlined in the supplementary online material.

Antibodies

The rabbit PBAN antibodies we used were produced targeting Helicoverpa zea PBAN linked to keyhole limpet hemocyanin by a cysteine residue and N-succinimidyl-4-(maleimidomethyl)-cyclohexane carboxylate (Marco et al., 1995). Rabbit polyclonal antiserum against [His⁷]-corazonin (pETFQYSRGWTNa) was prepared by Wako Co. (Nagano, Japan) and made available to us by Dr. Seiji Tanaka. Antiserum reactivity to [Arg⁷]-Crz was demonstrated by Roller et al. (2003). Antiserum against synthetic PDH of crustaceans (NSELINSILGLPKVMNDAa) was raised in rabbits and characterized by Dr. H. Dircksen (Dircksen et al. 1987), from whom it was obtained for our study. Helfrich-Förster and Homberg (1993) demonstrated that the antibody recognizes [Leu⁸, Leu¹⁵]-PDF.

Results

Calling Behavior

The calling activity of E. kuehniella females showed a distinct diel rhythm and did not change appreciably with age in an LD cycle. Females began to call during the first scotophase after emergence, and practically all individuals were observed calling at the onset of each photophase. The proportion of calling females decreased gradually during each photophase, and the behavior was almost totally suppressed around the onset of each scotophase (Fig. 1A). The daily rhythm in calling was persistent for several days in continuous darkness, but the proportion of calling females gradually decreased with age. Calling was not entirely suppressed at the beginning of the expected scotophases, resulting in a gradually more vague rhythm after a few days in darkness (Fig. 1B). Continuous light suppressed calling behavior, which resulted in an abolished rhythm and a severely reduced number of calling females at practically all hours of the day (Fig. 1C).

Figure 1.

Calling behavior and locomotor activity in E. kuehniella (A-C, G) and P. interpunctella (D-F, H) during 1 week in different light-dark (LD) cycles. E. kuehniella females call around lights-on in a 17-h:7-h LD cycle (A), and the rhythm is persistent in constant darkness (B). The behavior is practically suppressed in constant light (C). P. interpunctella females call around lights-off in LD cycle (D) but turn practically arrhythmic in constant darkness (E) and constant light (F). Average activity profiles of 30 E. kuehniella (G) and P. interpunctella (H) males recorded under short-day, equinox, and long-day conditions, respectively. The gray bars in A and D indicate the scotophases, and the graphs in B, C and E, F are superimposed on the LD data from A and D. N = 30 individuals.

The calling activity of P. interpunctella also showed a diel rhythm, and females started to call during the latter half of the first photophase after emergence. Calling activity reached its peak at the onset of each scotophase, when practically all observed females were calling. In contrast to E. kuehniella females, calling was never entirely suppressed at any time of day (Fig. 1D). The rhythmic calling pattern of P. interpunctella was apparently lost in continuous darkness, although the highest proportions of calling activity were persistently obtained at the hours around lights-off (Fig. 1E). In continuous light, the calling behavior became arrhythmic but not suppressed (Fig. 1F).

Locomotor Activity

Both E. kuehniella and P. interpunctella males showed persistent rhythms in constant darkness (Suppl. Fig. S1), suggesting that locomotor activity is regulated by a circadian clock mechanism in both species. As can be seen in Figure 1G, the peak in E. kuehniella activity occurred at the end of night in LD regimes, whereas the activity peak of P. interpunctella occurred at light-off and the beginning of night (Fig. 1H). This trend was most profound under short-day conditions. Under long-day conditions, both species displayed bimodal activity pattern (Fig. 1G, H). In the LL regime, locomotor activity of P. interpunctella was suppressed and E. kuehniella activity stopped completely (Suppl. Fig. S2).

PBAN-ir

We observed a similar pattern in distribution of PBAN-ir in the brain-subesophageal ganglion (SOG) complexes and neurohemal organs in both species of moths, P. interpunctella and E. kuehniella (Fig. 2). Clear cytoplasmic staining in 6 and 8 cells occurred in the pars intercerebralis (PI) in each hemisphere of P. interpunctella and E. kuehniella, respectively (Fig. 2A, B, K, L). Neuronal fibers of the PI cells passed across medial protocerebrum to the base of the inferior protocerebrum, dispatching branches to ipsilateral cardiacal nerves 1+2 (NCC 1+2). The PBAN-ir pattern of P. interpunctella included 8 perikarya located in the superior lateral protocerebrum (SLP; Fig. 2A, B), and 2 cells resided in the area of the lateral horn at the superior part of the protocerebrum (Fig. 2A, C) in each hemisphere. The SLP neurons are divided into the medial and the lateral clusters, each consisting of 4 cells. Fourteen immunopositive cells were detected in the SLP of E. kuehniella (Fig. 2K, L, M). Two medial groups harbored 5 small cells each while the lateral group was formed by 4 large neurons (Fig. 2K, M). The SLP horn possessed 4 perikarya with prominent granulose staining (Fig. 2K, L). Axonal projections of the immunopositive cells located in SLP of the protocerebrum formed a meshwork above the calyx and peduncles of the mushroom body and ramified in the regions of the protocerebral bridge and the central body (Fig. 2D).

Figure 2.

PBAN-ir in the brain, the SOG, and neurohemal organs of the moth females P. interpunctella (A-J) and E. kuehniella (K-V). (A) The schematic diagram illustrating the topography of PBAN-ir cells and axons in P. interpunctella. (B) Immunopositive cells in the PI and the SLP. (C) A pair of stained cells in the horn of the protocerebrum. (D) Axonal projections in the region of the PB and the MB in the frontal protocerebrum. (E) Small cells and axonal arborization (arrow) in the Tr. (F) Strongly stained cells in the region of Mdb and Mxl in the SOG. (G) Immunopositive cells in the labial neuromere of the SOG. (H) Heavily stained fibers in the middle-ventral part of the SOG. (I) PBAN-ir in the frontal ganglion. (J) Extensive fiber arborization in the CC, NCC 1+2, and NCA. (K) The topology of PBAN-ir neurons and their axonal pathways in E. kuehniella. (L) Strongly stained neurons in the PI and the superior lateral protocerebrum. (M) Positive cells in the SLP. (N) Axonal arborization (arrow) and a group of small cells (arrowhead) in the Tr. (O) PBAN-ir cells located in the Mdb and Mxl in the SOG. (P) The LB cluster of positive cells. (Q) The central maxillary cluster of cells. (R) A lateral cell in the Mxl neuromere. (S) Prominently stained fibers in the middle part of the SOG. (T) Fibers extending from the Mdb neuromere. (U) Immunopositive cells in the FG. (V) PBAN-ir in the CC. CA = corpora allata; CB = central body; CC = corpora cardiaca; FG = frontal ganglion; la = lamina; Lb = labial neuromere; lo = lobula; MB = mushroom body; Mdb = mandibular neuromere; me = medulla; Mxl = maxillary neuromere; NCA = nervi corporis allati; NCC 1+2 = nervi corporis cardiaci; oes for = esophageal foramen; OL = optic lobe; PB = protocerebral bridge; PI = pars intercerebralis; SLP = superior lateral protocerebrum; SOG = subesophageal ganglion; Tr = tritocerebrum; VLP = ventrolateral protocerebrum. Scale bar 50 µm.

Eight to 10 small but clearly stained cells were detected in the caudal part of the tritocerebrum of both species (Fig. 2A, E, K, N). Their axons ran to the base of the protocerebrum via the deutocerebrum, ramifying in P. interpunctella into a distinct network. Other axons passed through the circumesophageal connective to the SOG, which exhibits prominent PBAN-ir. Three clusters of large positive cells were detected along the ventral midline of SOG in both P. interpunctella and E. kuehniella (Fig. 2A, K). The cluster located in the mandibular neuromere (Mdb) consisted of 4 cells (Fig. 2F, O), the maxillary neuromere (Mxl) cluster consisted of 6 centrally and 2-4 laterally located cells (Fig. 2F, O, Q, R), and the labial neuromere (Lb) cluster harbored 4-6 cells (Fig. 2G, P). The axons extending from Mdb cluster formed parallel bundles headed to the superior part of the SOG (Fig. 2H), continued through circumesophageal connectives, and arborized in the tritocerebrum. Collateral fibers created a prominent bundle projecting into the lateral region of SOG in P. interpunctella; in E. kuehniella, 2 additional parallel fiber bundles extending from Mdb cluster were found (Fig. 2T).

In both species, strong PBAN-ir was revealed in the frontal ganglion (FG). Six well-stained cells in P. interpunctella (Fig. 2A, I) and 4 cells in E. kuehniella (Fig. 2K, U) were found in FG. Thin fibers going through the frontal connectives connected immunoreactive material in the FG with the tritocerebral PBAN-ir system.

The corpora cardiaca (CC) showed very strong immunoreactivity in both pyralid moth species (Fig. 2A, J, K, V). A dense network of stained axons was found in nervi corporis cardiaci (NCC 1+2, NCC 3). Neurites reached the dorsal surface of the corpora allata via nervi corporis allati (NCA), but no PBAN-ir was detected in corpora allata themselves (Fig. 2, J, V).

Crz-ir Immunoreactivity

The brain of P. interpunctella harbored 2 triads of Crz-ir cells; 1 group is located in the medial (Fig. 3A, B) and the other in the lateral (Fig. 3A, C) part of the SLP in each hemisphere. Neuronal fibers of the cells conjoined in a trajectory that ramified into a gentle bundle around the calyx of the mushroom body and passed through a medial part of the peduncle to the base of the inferior protocerebrum extending projections to ipsilateral cardiacal nerves (NCC 1+2). Several thin fibers arising from the trajectory provided a connection to the opposite hemisphere through the central body. Three smaller additional neurons were located in the SLP horn near the optic lobe. No axonal projections of these cells were stained (Fig. 3 A, D, E).

Figure 3.

Crz-ir in the cephalic ganglia of P. interpunctella (A-H) and E. kuehniella (I-Q). (A) Representation of the immunostaining pattern in P. interpunctella. (B) Prominently stained cells in the medial (arrow) and the lateral (arrowhead) part of the SLP. (C) A lateral SLP triad of immunopositive cells. (D) Two small strongly stained cells located in the protocerebral horn. (E) A ventral cell of the horn triple of neurons located nearly the OL. (F) A pair of stained perikarya in the caudal region of the VLP. (G) A group of small cells in the tritocerebrum. (H) Heavily stained fibers and perikarya in the CC. (I) The diagram of the immunostaining pattern in E. kuehniella. (J) A pair of Crz-ir neurons and their axons forming a single track (arrowhead) in the region of the SLP. (K) Prominently stained cells in the SLP. (L) Axonal projections in the region of the CB. (M) A pair of stained cells in the region of the SLP horn. (N) Fibers forming a fanlike arrangement over the medulla in the OL. (O) Crz-ir in the OL and photoreceptors (arrowhead). (P) Positive cells in the SOG. (Q) Strongly stained perikarya (arrowhead) and fiber arborization within the CC. AL = antennal lobe; CA = corpora allata; CB = central body; CC = corpora cardiaca; la = lamina; me = medulla; NCC = nervi corporis cardiaci; oes for = esophageal foramen; OL = optic lobe; PI = pars intercerebralis; Pr = protocerebrum; SLP = superior lateral protocerebrum; SOG = subesophageal ganglion; Tr = tritocerebrum; VLP = ventrolateral protocerebrum. Scale bar 50 µm.

Two stained cells and immunopositive axons were detected in the caudal region of the ventrolateral protocerebrum at the border with the deutocerebrum (Fig. 3A, F). Four small Crz-ir perikarya were located in the caudal part of the tritocerebrum (Fig. 3A, G). A thick network of Crz-ir fibers and 4 strongly stained cells were revealed in CC (Fig. 3A, H).

Two pairs of Crz-ir neurons were located in the SLP in each hemisphere of E. kuehniella (Fig. 3I, J, K). Neurites of the cells joined into a single track (Fig. 3I, J) heading above the mushroom body to PI were arborized slightly in the regions of the protocerebral bridge and the central body (Fig. 3I, L). One or 2 fibers separated from this tract, forming a trajectory that innervated the ipsilateral CC. Two additional cells smaller and less intensively stained were located in the part of the SLP horn (Fig. 3I, M). Axonal projections of these paired cells followed 2 circuits. The short one joined the transverse track and the long one led to medulla neuromere in the optic lobe (OL).

In E. kuehniella, immunoreactivity with antiserum against [His 7]-corazonin revealed prominent staining in the OL and eyes (Fig. 3I, N, O). Cytoplasmic staining occurred in each ommatidium both in 8 photoreceptors placed along the rhabdom and in a basal photoreceptor (Fig. 3O). Immunopositive projections ran over the frontal surface of the lamina neuromere, crossed in the region of the outer chiasma, and formed a fanlike arrangement over the frontal surface of medulla neuromere (Fig. 3I, N, O).

A single positive cell in each half of the SOG was located laterally in the Mxl region (Fig. 3I, P). A projection extending from this cell was stained slightly but seemed to have run through the SOG and circumesophageal connected to the tritocerebrum.

Strong staining occurred in 4 perikarya in the CC and in fiber arborization within the CC and NCA, but no Crz-ir was detected in the corpora allata (Fig. 3I, Q).

PDF-Immunoreactivity

Both species of examined moths possessed the PDF-positive cells in the PI, the SLP, the OL, and the SOG (Fig. 4). No variances in immunostaining were observed in males in comparison to females reared under the same conditions.

Figure 4.

PDF-ir in the brain-SOG complex of P. interpunctella (A-G) and E. kuehniella (H-O). (A) The topology of PDF-ir neurons and the pathways of their projections in P. interpunctella. (B) Immunopositive cells with their axonal projection in the PI. (C) The medial group of 2 stained cells in the SLP. (D) Immunopositive cells belonging to the lateral SLP cluster. (E) The Pfv cluster of PDF-ir cells (arrow) in the OL. (F) Extensive fiber arborization in the CC and NCC 1+2. (G) PDF-ir cells in the medial Lb neuromere (arrow) and a small, weakly stained cell in the lateral part of Mxl neuromere (arrowhead) of the SOG. (H) The topology of PDF-ir neurons and their axonal projections in E. kuehniella. (I) Immunopositive cells in the PI. (J) A cluster of stained neurons in the SLP. (K) Weakly stained cells located in the IMP. (L) The Pfv cluster of positive cells (arrow) in the OL. (M) A pair of PDF-ir cells in the Mxl neuromere region of the SOG. (N) A prominently stained cell at a level of the Lb neuromere in the SOG. (O) Thick network of PDF-ir axons in the CC. CA = corpora allata; CC = corpora cardiaca; IMP = inferior medial protocerebrum; Mdb = mandibular neuromere; Mxl = maxillary neuromere; NCC = nervi corporis cardiaci; OL = optic lobe; Pfv = proximal frontoventral region; PI = pars intercerebralis; SLP = superior lateral protocerebrum; SOG = subesophageal ganglion. Scale bar 50 µm.

Six immunopositive cells were located in the PI in each hemisphere in P. interpunctella (Fig. 4A, B). Their axons joined together and ran across medial protocerebrum to the base of the inferior protocerebrum, innervating the ipsilateral CC. In E. kuehniella, 5 PI cells (Fig. 4H, I) sent their axonal fibers to the contralateral CC through the NCC 1+2. Five prominently stained neurons that divided into a medial cluster consisting of 2 cells (Fig. 4A, C), and a lateral cluster of 3 cells (Fig. 4A, D) were detected in the SLP in each hemisphere in P. interpunctella, whereas 4 distinct neurons occurred in the SLP in E. kuehniella (Fig. 4H, J). Thin fibers emanating from the cells traversed above the calyx, joining the longitudinal track in the area of the protocerebral bridge. E. kuehniella possessed 3 additional weakly stained cells in the caudal region of the inferior medial protocerebrum (Fig. 4H, K).

In P. interpunctella, gentle fibers arising from the lateral SLP group of cells headed to the optic lobe, reaching the proximal frontoventral (Pfv) region, where 4 small cells were located (Fig. 4A, E). Neurites of the Pfv cluster of cells reached both the ventral border between the OL and protocerebrum and the space between lobula and medulla neuromeres. No axons around the frontal medulla surface were detected in the OL of P. interpunctella. In E. kuehniella, a cluster of 4 PDF-ir perikarya located in the Pfv area of the OL (Fig. 4H, L) sent fibers in a fan-like shaped network spreading over the frontal surface of the medulla.

The SOG harbored, in both moth species, 2 groups of PDF-ir cells in the area of the Mxl and Lb neuromeres (Fig. 4A, H). Two cells were located laterally in the frontal part of Mxl neuromere in P. interpunctella (Fig. 4A, G), whereas in E. kuehniella a pair of cells was placed medially (Fig. 4H, M). The ventromedial Lb clusters consisted of 4 and 2 perikarya in P. interpunctella and E. kuehniella, respectively (Fig. 4A, G, H, N). Cells of both Lb and Mxl clusters sent their projections into a track that passed along the SOG midline into circumesophageal connectives and continued toward the tritocerebrum. PDF-ir fibers permeated the NCC 3 and formed a thick network in the CC (Fig. 4A, F, H, O). In E. kuehniella, stained fibers were detected in the FG.

Discussion

We conclude that the daily temporal allocation of female calling behavior differs in the 2 pyralid moth species E. kuehniella and P. interpunctella. Females of the 2 species release the sex pheromone to attract males at different times of the night, dawn and dusk, respectively, and the calling behavior is affected differently by constant light conditions. The calling rhythm in E. kuehniella shows characteristics of circadian regulation as the rhythm is persistent in constant darkness and the behavior is suppressed by constant light. On the contrary, females of P. interpunctella turn arrhythmic in DD but continue to call in both constant dark and light. Males of both species show persistent daily rhythms in locomotor activity in constant darkness.

The original habitats of our studied species are not known, and they are almost exclusively found indoors, that is, in artificial and—from an evolutionary perspective—recent environments. It is therefore hard to speculate how the asynchrony in daily sexual behaviors has evolved. However, several of the closely related pyralid species share the same main pheromone component, and indirect selection on males occurs already during the mate search and localization of females by sex pheromone cues (see Wiley and Poston, 1996, for a general discussion of indirect female choice). In a competitive environment with more than 1 species present, it is therefore plausible that a shift of the activity peak may have been beneficial to avoid costly and time-consuming interspecific attractions.

A closer examination of P. interpunctella calling in constant darkness reveals that even though females never ceased calling at any time of day, the highest numbers of calling females were persistently obtained at hours corresponding to lights-off. This suggests that the onset of calling behavior may nevertheless be regulated by a clock mechanism but that the termination of calling is dependent on the lights-on cue. However this hypothesis is contradicted by high calling activity in constant light (Fig. 1F). Since P. interpunctella females showed no calling rhythm in constant darkness from the very beginning, it is unlikely that desynchronization between individuals caused the overall arrhythmicity. However, desynchronization would explain the gradual dampening of the E. kuehniella rhythm in DD.

Circadian clocks use changes in light conditions to synchronize with the environment, and if constant light suppresses calling behavior in E. kuehniella, why do P. interpunctella females continue to call in constant light? One explanation is that the suppression of calling behavior in E. kuehniella occurs independently of the clock mechanism. This would require, however, that the 2 species show differences in a secondary light input system, which affects the behavior independently of the clock function. We find it more likely that either the calling behavior in P. interpunctella is disconnected from the circadian clock mechanism or the clock output signal that regulates calling differs between the 2 species.

Male mate search, and subsequently reproduction, require locomotor activity, and we therefore use locomotor activity as an indicator of hours when mate search would primarily occur. E. kuehniella males showed highest locomotor activity at the end of the scotophase, which corresponds very well with the observed peaks in female calling activity. The locomotor activity rhythm in P. interpunctella was highest during hours corresponding to the early scotophase.

The performed immunostainings showed minor differences in distribution or localization of PBAN-ir, Crz-ir, and PDF-ir in the cephalic ganglia of both species (Suppl. Table S1). The neuropeptides PBAN, corazonin, and PDF have all been proposed to be essential for clock output and showed strong staining patterns in both E. kuehniella and P. interpunctella brains. No differences in distribution of PBAN-ir, Crz-ir, and PDF-ir were revealed with respect to both the sex and the time points of the LD cycle (see Suppl. Table S1, Suppl. Figs. S3, S4).

PBAN polypeptides are members of the pyrokinin family with different physiological functions in insects and are characterized by a common C-terminal sequence called FXPRLamide (Abernathy et al., 1996). The PBAN-ir does not differ substantially between the 2 species, apart from an additional group of 5 small cells in the medial SLP in E. kuehniella (Suppl. Table S1, Fig. 2G). The strongly stained cells in the SOG were found in the Mdb, Mxl, and Lb neuromere of the SOG in both species. Previous studies revealed that 3 clusters of neurosecretory cells in the midventral SOG showed strong immunoreactivity with antiserum against PBAN in several other moths as well: Helicoverpa zea (Ma et al., 1996), Agrotis ipsilon (Duportets et al., 1998), Agrotis segetum (Závodská et al., 2009), and tortrix Adoxophyes sp. (Choi et al., 2004). In both examined moths, immunopositive fibers from the pars intercerebralis led into the nervi corporis cardiaci (NCC1+2) and ramified in the CC (Fig. 2A, K), where immunostained cells in the tritocerebrum likely sent their processes via the NCC 3 (Fig. 2J). These results suggest that PBAN-like peptide synthesized in both E. kuehniella and P. interpunctella brain neurons is transported to the CC before its release into the hemolymph and transfer to the pheromone gland.

Nässel (2002) suggested that corazonin, an 11 amino acid neuropeptide, functions as both a humoral factor and a neuromodulator, which are both plausible roles in output pathways of circadian clock. Its linkage to circadian clock pathways is based on its co-expression in cells expressing clock protein Period in the Antheraea pernyi brain (Sauman and Reppert, 1996; Sauman et al., 2005) as well as in the M. sexta brain (Wise et al., 2002).

Interestingly, the distributions of Crz-ir in P. interpunctella and E. kuehniella do differ (Suppl. Table S1). Whereas E. kuehniella shows 2 stained cells in the SOG with no equivalent in P. interpunctella, the latter species shows 2 cells in caudal part of ventrolateral protocerebrum and 4 small cells in the tritocerebrum (Fig. 3).

In E. kuehniella we detected prominent Crz-like staining in photoreceptors and fibers running over frontal surface of lamina and medulla neuromere in the OL (Fig. 3N, O), which is a unique Crz-ir distribution not only in moths but even in other insects tested previously (e.g., Sehadová et al., 2007; Závodská et al., 2008). We observed positive immunostaining with antiserum against [His 7]-corazonin in photoreceptors of all 12 examined E. kuehniella heads but never saw an immunopositive signal in the photoreceptors of P. interpunctella, although samples of both moth species were treated together under the same conditions. In addition, we observed clear staining in 4 perikarya in CC of E. kuehniella. Although it cannot be entirely excluded that an artifact in the eye and CC occurred during the immunohistochemical process, which gave an unspecific staining, we are convinced that the positive reaction in photoreceptors and cells in the CC was specific considering the reliability of antiserum against [His⁷]-corazonin (Roller at al., 2003). We would not venture to claim that the Crz-like substance recognized by the antibody in the photoreceptors is also produced in these cells or attempt to assign any function to this compound in the photoreceptors, yet it is remarkable that such differences in the staining pattern were observed in the 2 species. The differences in expression patterns between these closely related species at least suggest that corazonin was subjected to a selective pressure over time.

PDF is an 18 amino acid neuropeptide that is essential for maintained locomotor activity rhythms in Drosophila (Renn et al., 1999; Helfrich-Förster, 2005). PDF homologues (which usually differ only by 2 amino acid substitutions) have been identified in diverse insect species (reviewed by Hamasaka et al., 2005). Both examined species of moths possess 4 small but clearly stained cells in the Pfv region of the OL (Fig. 4), which is a typical location of PDF-ir cells in Drosophila and other insect species (Sehadová et al., 2003; Závodská et al., 2003). Expression patterns of PDF-ir vary only slightly in the number of SLP, PI, and SOG cells in moth brains, but no equivalent cells in the interior medial protocerebrum were detected in P. interpunctella (Suppl. Table S1).

To summarize, the circadian regulation of the sex pheromone communication in P. interpunctella and E. kuehniella shows remarkable differences. The restriction of their sexual activity to different hours of the day suggests that the differences may be due to temporal divergence in the sex communication, between either these or other species relying on the same sex pheromone components. Since effective reproduction requires synchronized behaviors in both sexes, this temporal shift in activity has affected both males and females, as would be expected. The variances in the expression patterns of especially corazonin indicate that differences between the 2 species occur also on a molecular level and in the pathways by which the circadian rhythms are regulated.