Sexual differentiation in Hierodula patellifera

The intersexuality in H. formosana induced by C. formosanus makes the sexually dimorphic structures on the infected male to resemble that on the female

(Chiu et al., 2015). Most of the sexually dimorphic structures are evolved under the

sexual selection which are developed for reproduction (Shine, 1989; but see

González-Solís et al., 2000). Since the sexually dimorphic structures is often related

to host reproduction, many of these characters are identical in the early stage and differentiate during the development (Negri and Pellecchia, 2012).

In our examination of the developmental trajectories of the sexually dimorphic characteristics, the alternative Hierodula host, H. patellifera, were used since its smaller size and relatively shorter development time (there are generally 7–9 instars before the last molting in H. patellifera and 14–15 instars in H. formosana). To examine the morphological change in each instar, three sexually dimorphic characteristics were chosen since they can be examined the morphology without killing the mantids (except the early instars which are too small to be examined alive).

The three sexually dimorphic characteristics are: 1) antennal sensilla, 2) red pigments on 5^th–7^th tergum of the female, and 3) morphology of the last three sternum (6^th–8^th) forming the genitalia. It is interesting that they displayed the different parasitic effects under the infection of C. formosanus.

The morphology of the 6^th–8^th sternum is the main characteristic frequently applied in distinguishing the sex of mantids (Prete et al., 1999) and is also applied in the horsehair worm-infected mantids (Chiu et al., 2015). There are six visible segments of sternum in the female adult and eight segments in the male adult (Prete et

al., 1999), but in the ten 1

^st and 2^nd instar nymphal H. patellifera examined

respectively, all the individuals are showed the eight-segment abdomen (Fig. 17A, B).

The differentiation in the morphology begins to be detected from the 3^rd instar. In the 3^rd instar female, posterior edge of the 6^th sternum slightly depress (Fig. 17C, E) and a pair of appendages respectively appear at posterior edge of the 7^th (Fig. 17C, F) and 8^th (Fig. 17C, D) sternum. In the 4^th instar, both of the depression and the appendages become distinct and color of the appendages turns to red (Fig. 18A). In the 5^th instar, the 6^th sternum with depressed edge becomes wide and almost covers the 7^th and 8^th sternum with the slender appendages (Fig. 18B). The 6^th sternum finally covers the 7^th and 8^th sternum and become the last visible sternum and keep this morphology until the last molting (Fig. 18C). The appendages on the 7^th and 8^th sternum are finally sclerotized and become the valvula of the ovipositor (Fig. 19). By contrast, the last three sterna of the male are almost not changed in the shape in all the development

(Fig. 17G, 18D–E), only the last sternum (8

^th sternum) become moderately elongated in the last instar (Fig. 18F). Development of the sternum seems to be not influenced by the infection since the H. patellifera artificially infected by Chordodes shows the characteristic described above, only the shape of the last sternum is slighter smaller in some infected male (Fig. 21D). The character applied in distinguishing the insect sex has ever been found to be altered by parasites and causes the morphological sex reversal

(Vance, 1996). The consistent development which free from the parasitic

effect supports the correctness of the sex identification in Chiu et al. (2015), which, honestly, was ignored during the original field survey.

Fig. 17. Ventral view of sterna on the 1^st to 3^rd instar nymphs of Hierodula patellifera by SEM. (A–B) Sterna of the 1^st and 2^nd instar nymph. The sex is unable to be identified. (C–F) Sterna of the 3^rd instar female nymph. The posterior edge is slightly depressed on the posterior edge of 6^th (E), and the paired appendages respectively appear at posterior edge of the 7^th (F) and 8^th (D) sternum. (G) Sterna of the 3^rd instar male nymph. The morphology is generally similar to that of the the 1^st and 2^nd instar nymph.

Fig. 18. Ventral view of sterna on the 4^th, 5^th, and 9^th (the last) instar nymphs of

Hierodula patellifera. (A–C) Sterna of the female instar nymph. The paired

appendages become distinct and red in color from the 4^th instar (A) and are covered by the 6^th sternum from the 5^th (B) to the last instar (C). (D–F) Sterna of the male instar nymph. The morphology is not obviously changed in the 4^th (D) and 5^th (E) instars. Only the last sternum (8^th sternum) become moderately elongated in the last instar (F).

Fig. 19. Detailed comparison of the 6^th-8^th sternum in the female Hierodula patellifera of the last instar nymph and the adult. (A–E) The sterna of the last instar nymph with the ventral view of abdominal tip (A), the lateral view of the 7^th-8^th sternum (B), the ventral view of 6^th (C) 7^th (D), and 8^th (E) sternum.

(F–G) The sterna of the adult with the ventral view of abdominal tip (F), and ventral view of the 7^th and 8^th sternum (G).

The red pigments on 5^th-7^th tergum of the adult female (Fig. 20A, B) is another characteristic differentiated from 3^rd instar

(Fig. 20C). It is believed that the gland

which produces the sex pheromone is opened at these three segments of female mantids since they are usually covered by the wings and showed when the female mantid bends the abdomen to release the sex pheromone

(Robinson and Robinson,

1979). We now have no evidence to suggest any relationship of the red pigments on

5^th-7^th tergum and the gland open, but know this characteristic is only appeared in the female H. patellifera. The examination of the ten 1^st and 2^nd instar nymphs suggests the red pigments are not displayed in the both sexes and can be slightly seen under the stereomicroscope from the 3^rd insatr

(Fig. 20C). The red pigments become clear and

visible from the 4^th instar (Fig. 20D) to the adult. The red pigment was found in some infected male H. patellifera artificially infected by C. formosanus

(Fig. 21A) or

Chordodes sp. (Fig. 21B, C) even all the mantids were fed with the horsehair worm

cysts after 3^rd instar. Although not all the infected males showed this characteristic, this result suggests the appearance of the red pigment can be induced by the parasite after the time of sexual differentiation. The different responses of the sexually dimorphic characteristics to the horsehair worm infection has been previously found in the field-collected H. formosana since the number of antennal segment is higher in the male than that in the female, but there was not significantly different in the non-infected males and the infected males (Chiu et al., 2015). These characteristics not affected by the infection might be caused by the insensitivity of the factor which the horsehair worm applied to regulate the host development, or due to development of the characteristic is passed and determined (Rempel, 1940). The adult midges infected by the nematodes show the different level of morphological alterations with the different timing of infection, since the nematode is not able to manipulate the

characteristics which are already differentiated during the ontogeny

(Rempel, 1940).

It might be one possible reason to explain development of the sternum since all the mantids were infected after 3^rd instar, when the sexual differentiation has already proceed. However, the hypothesis of "ontogeny steps" might not explain the parasitic effect on the red pigment since this characteristic already pass the process of sexual differentiation but is still revisable by the parasitic effect.

Fig. 20. Red pigments on 5^th–7^th tergum of the mantid, Hierodula patellifera. (A) Live female adult releasing sex pheromone and showing the red pigments on the abdomen (arrow). (B) The red pigments appearing in the female adult (arrows) but not in the male adult. (C–D) The red pigments are visible in the 3^rd instar female (C) and distinct in the 4^th instar female (D). (E–F) The red pigments are not found in the nymphal male of the 3^rd (E) and the 4^th (F) instar nymph.

Fig. 21. Nymphal male mantids, Hierodula patellifera, infected by Chordodes

formosanus and Chordodes sp. (A–C) Nymphal males infected by C.

formosanus (A) and Chordodes sp. (B–C) displaying the red pigments which

are only appeared in the female. (D) Ventral view of sterna of the nymphal male infected by C. formosanus (same individual with (A)).

Development of the red pigments is not applied in the further experiment since its parasitic effect was not happened to all the individuals and the method to examine is not yet stable enough. The advanced experiment in the mechanism of parasitic intersexuality are conducted by the examination of the differentiation of sensillum distribution on antennae. As that happened on H. formosana, the sensilla amount is also significantly higher in the male adult of H. patellifera (Table 1, 2; Fig. 22A).

Among the three types of the sensilla identified by their morphology (large trichoid sensilla, small trichoid sensilla, and groove basiconic sensilla) (Fig. 22), number of both large trichoid sensilla and the groove basiconic sensilla are significantly larger in the adult of the male than the female, especially the number of groove basiconic sensilla which is nearly three-folds larger in the male. Whereas there is no significant difference in the small trichoid sensilla between male and female. Since the density of seneilla on the insect cuticle cannot been unlimitedly increased (Wigglesworth, 1940), the male has wider surface to bear the sensilla by three means: higher number of the flagellum segment

(Table 1, 2), larger antennal segment with wider surface area (Table 1, 2; Fig. 23A), and distribution of the groove basiconic sensilla which is

anterior to the base of the antenna (the first flagellum segment bearing the grooved basiconic sensilla in

Chiu et al., 2015) (Table 1, 2; Fig. 24). These three sexually

dimorphic characteristics of antennae respectively contribute to 5.02%, 54.42%, and 40.56% of the different amount of the grooved basiconic sensilla in the adult H.

patellifera (Table 2). The time of sexual differentiation is different among these three

characteristics. The total segment is that first differentiate from the penultimate molting at the last instar (Table 1). However, the sensilla number does not significantly differentiate at this time since the flagellum segment bearing the groove basiconic sensilla is similar between the male and the female (Table 1, Fig. 22B).

Both of the antennal surface area and distribution of the groove basiconic sensilla, which mainly cause the higher among of sensilla in the male adult, are sexually differentiated from the last molting. It is worth to note that the distribution of the groove basiconic sensilla is sexually differentiated in both adults of H. patellifera and

H. formosana which the male has the wider distribution of sensilla toward the anterior

of the antenna. However, in H. patellifera, the distribution of female does not maintain the nymphal type of the last instar. In both the male and the female, the distribution of the groove basiconic sensilla spreads toward the antennal base at the last molting. Thus, the sexual dimorphism in the sensillum distribution is not caused by the maintaining of the nymphal character in the female but by the different degree of the distribution extending (Table 1).

Other than the sexual differentiation during the last molting, the examination of the antenna in each instar nymph reveals the change of sensillum distribution in every instars. The first flagellum segment bearing the grooved basiconic sensilla changes toward the posterior end of the antenna in each nymphal instar from around 25^th segment to posterior than 50^th segment in the last instar (Table 1, Fig. 24). This development is almost the same in both males and females until the 9^th instar. At the 9^th instar, the first flagellum segment bearing the grooved basiconic sensilla is significantly posterior in the male nymph than that of the female (Table 1). However, this difference is likely to be caused by the sexual differentiation of the antennal segment number while the male nymph has higher number of the segment in the 9^th instar. In comparing the segment number bearing the groove basiconic sensilla, it is similar in the both sexes in all nymphal instars and diverged in only adult stage (Table

1). In it now known the distribution of the groove basiconic sensilla suffers the

parasitic effect in the male adult (Chiu et al., 2015). Since the female adult of H.

formosana maintains the antennal characteristics as the last instar, it is believed that

the intersexuality induced by the horsehair worm might be the result of juvenilization

(Chiu et al., 2015). The "juvenilization" and "interference in insect sexual

differentiation" are two main hypothetical mechanisms in explaining the intersexuality.

The development of the sensillum distribution includes both change in every instars and sexual differentiation in the last molting provide a suitable indicator to test these two exclusive hypotheses

Table 1. Comparisons of sansillum amounts and antennal characteristics between male and female mantids, Hierodula patellifera

1. The P values conducted by Student T test indicate significant differences between sexes.

2. LTS: Total number of large trichoid sensilla, STS: Total number of small trichoid sensilla, GBS: Total number of groove basiconic sensilla, TFS: Total number of flagellum segment, ASA: Antennal surface area of the flagellum segment bearing the grooved basiconic sensilla, FFG: the first flagellum segment bearing the grooved basiconic sensilla, HFS: Total number of flagellum segment bearing the grooved basiconic sensilla.

1^st instar 8^th instar 9^th instar

Table 2. Difference of the grooved basiconic sensilla in the three sexually dimorphic zones of adult Hierodula patellifera Segment

Zone1 Zone2 Zone3

1^st-43^th 44^th-89^th 90^th to the antennal tip

Male adult 1209.33 ± 635.80 2314.00 ± 1085.05 175.17 ± 89.84 Female adult 210.20 ± 143.13 973.40 ± 477.97 51.60 ± 56.97

P-value by T-test P < 0.001 P < 0.001 P = 0.008

Average difference 999.13 1340.60 123.57

Percentage to total difference 40.56% 54.42% 5.07%

1. The P values conducted by Student T test indicate significant differences between sexes.

2. Zone1: antennal segment anterior the female segment first bearing the grooved basiconic sensilla. Zone2: antennal segment between the female segment first bearing the grooved basiconic sensilla and the female total segment number. Zone3: antennal segment posterior the female total segment number. See Fig. 23A.

Fig. 22. Comparison of sensillum amounts between the male and the female

Hierodula patellifera in the adult (A), the last (9

^th) instar (B), and the first instar (C). The P values conducted by Student T test indicate significant differences between sexes.

Fig. 23. Comparison of antennal size (surface area) between the male and the female

Hierodula patellifera in the adult (A), the first instar (B), and the last (9

^th) instar (C). Zone1-3 in (A) are separated by "the antennal segment of the female segment first bearing the grooved basiconic sensilla" and "the antennal segment of the female total segment number". See Table 2.

Fig. 24. Developmental trajectories of the sensillum distribution (the antennal segment first bearing the grooved basiconic sensilla) on mantid antennae,

Hierodula petallifera. The 10

^th instar is the adult stage. Numbers in the brackets are numbers of samples examined.

在文檔中 臺灣鐵線蟲生物多樣性及寄生對於寄主形態發育的操控 (頁 100-117)