Reproductive experiments

In the present study, we collected individuals in the last instar nymph or sub-imago stage, virgin females, and copulated females of D. tipuliformis from a stream in a field (Fig. 1A). In the laboratory, the nymphs or sub-imagos were separated from each other into individual cases and were reared up to the adult stage. Nymphs were reared in plastic cases (78 mm × 114 mm × height 42 mm; Inomata, Osaka), and sub-imagos were reared in 30 mL cylindrical container named “Push Vials” (diameter 30 mm × height 59.2 mm; Nikko Hansen, Tokyo). Since nymphs mainly eat detritus, we raised them in stream water containing detritus. Almost all the last instar nymphs emerged within one week from the start of rearing. The emerged sub-imago once again molt and become imagos, even when combining these sub-imago and imago stages, that period amounts to at most 2–3 days in nature (Takenaka and Tojo, 2014). However, we rearing the sub-imagos following emergence under dark and low temperature conditions where able to extend the combined sub-imago and imago stage to about five days. In this manner, we tried to spread the timing at which the artificially bred specimens became available for hand-pairing by adjusting the various times after their emergence that they took to reach the adult stage.

We conducted three types of reproductive experiments, as follows. We investigated: (1) the rate of embryos developing to the final embryonic stage and hatching rates of egg batches that were oviposited by the artificially hand-paired females, (2) the rate of embryos developing to the final embryonic stage and hatching rates of egg batches that were oviposited by copulated females collected in the field, and (3) the egg batches oviposited by virgin females (not hand-paired females) as a control check for the potential of parthenogenesis. After 30 days of incubation at 20–25°C, the rate of embryos developing to the final embryonic stage and percentage of hatching success were measured under a microscope (SMZ1500, Nikon, Tokyo, Japan) for each egg batch. In this study, the developmental incidence was calculated based on the ratio of the total number of observed eggs and the number of eggs for which embryogenesis proceeded. When the embryogenesis reached “stage 13 (cf. Tojo and Machida, 1997, 1998)”, it was counted as a developed egg. The hatching rate was calculated from the percentage of eggshells left after hatching.

Fig. 1.

Male of Dipteromimus tipuliformis (A); Natural mating in a field of D. tipuliformis (B); Hand-pairing method shown by grasping the wings of the virgin female and the male and then bringing the individuals together (C): male tightly grasped female abdomen between 8th and 9th abdominal segments, when male forceps closed onto the female abdomen ((D); (E) is enlarged); male genitalia inserted to female genitalia from subgenital plate between 7th and 8th abdominal segments (E); female started to oviposit when female terminal abdominal segments made contact with water surface (F). these processes were corresponding Table 1.

Statistical analysis

To compare the rate of embryos developing to the final embryonic stage and hatching rates of the egg batches by females that the artificially hand-paired and females that had copulated in the field and non-mating females, we performed nonparametric multiple comparison by using the Steel–Dwass test of the ‘Steel-Dwass’ function ( http://aoki2.si.gunma-u.ac.jp/R/Steel-Dwass.html) in R version 3.1.1 (R Core Team, 2014).

DNA sequencing

We conducted DNA analysis for each of the males and females that were used in hand-pairing experiments, and also that of their corresponding offspring (i.e., hatched nymphs) from these experiments. The extraction of total genomic DNA was conducted according to the same methods as used in our previous studies (Sekiné et al., 2013; Suzuki et al., 2014; Saito and Tojo, 2016a, b). Each total genomic DNA was used to amplify DNA fragments [the nuclear DNA (nDNA) PEPCK regions] by polymerase chain reaction (PCR) used to sets of primers: 5′-AGTGATGGCGGAGTCTTCTG-3′ and 5′-ATCACCTTGCCCTGTGTAATTC-3′ (own design). The purified PCR products, sequencing, and alignment was conducted according to the same methods as previously mentioned. Sequence data of the nuclear DNA PEPCK regions have been submitted to the GenBank (accession numbers: LC441034–LC441038).

Establishment of a hand-pairing method

In all of the examinations, only virgin Dipteromimus tipuliformis females and males were used (Fig. 1A). In addition, as a control experiment, individuals which had completed copulation under natural outdoor conditions were also used (Fig. 1B).

Table 1.

Hand-pairing procedure for the mayfly, Dipteromimus tipuliformis in laboratory.

The detailed hand-pairing process is shown in Table 1. In this hand-pairing method, all of the examined males tightly grasped their paired female's abdomens, and penetrated the female genitalia from her subgenital plate (i.e., between 7th and 8th abdominal segments). This mating style (copulation style) was the same as the style under natural conditions observed in the field (Takenaka and Tojo, 2014). We established a reliable technique for artificial mating experiments, using hand-pairing. In these experiments, it was confirmed that the examined male indeed penetrated the female's genitalia (Please refer to the video documentation, as a supplementary data).

Developmental and hatching rates of egg batches

The rate of embryos developing to the final embryonic stage and hatching rates of egg batches that were oviposited from the seven females that had copulated in the field were indicated to be at particularly high rates (i.e., the rate of embryos developing to the final embryonic stage was 95.7 ± 7.5% and the hatching rate was 91.5 ± 13.2%; Table 2). The rate of embryos developing to the final embryonic stage and hatching rates of egg batches that were oviposited by the artificially hand-paired females were also indicated to be particularly high (i.e., the rate of embryos developing to the final embryonic stage was 94.4 ± 4.8% and the hatching rate was 91.5 ± 6.0%; Table 2). On the other hand, the egg batches that were oviposited by virgin females (not hand-paired females) that served as a control for the potential for parthenogenesis did not develop nor hatch (i.e., both the rate of embryos developing to the final embryonic stage and hatching rate were 0.0%; Table 2). The rate of embryos developing to the final embryonic stage and hatching rates of egg batches by the artificially hand-paired females were significantly high compared with the egg batches that were oviposited by virgin females (Steel–Dwass test, P < 0.01). On the other hand, the rate of embryos developing to the final embryonic stage and hatching rates of egg batches by the artificially hand-paired females were not significantly different compared with the egg batches by females that had copulated in the field (Steel–Dwass test, P = 0.83).

Table 2.

The rate of embryos developing to the final embryonic stage of each egg batch that oviposited by the artificially hand-paired females, natural mated females (copulated females collected in the field) and virgin females (not hand-paired females).

Fig. 2.

Comparison of the nucleotide sequences (i.e., the genotype of the nuclear DNA PEPCK region, 563-bp) of the male and female pair used in the hand-pairing experiment, and the genotype of the randomly selected offspring (hatched nymphs) from the hand-paired females. Only the first and last nucleotides, and the sites where the polymorphism between targeted specimens was observed, are indicated. Three nymphs share the same genetic sequences.

The genotypes of offspring (i.e., hatched nymphs)

After the hand-pairing experiments, we determined the genotypes of the male and female that were used in each experiment. We also analyzed the genotype of the offspring (i.e., hatched nymphs) from each of the hand-paired females. For this analysis, we utilized the nuclear DNA PEPCK region, which is observed to be a comparatively genetic polymorphism rich region (Fig. 2).

It was revealed that the offspring (hatched nymphs) from hand-pairing method had heterozygotic genotypes derived from both the male and female used in each hand-pairing experiment (e.g., the locus 151-bp: hand-paired male had “G” homo-type, on the other hand the hand-paired female had “A” homo-type, and the hatched nymph oviposited by the artificially hand-paired female had the corresponding “A/G” hetero-type; Fig. 2). In other words, it was clearly indicated that the hatched nymph had derived its genotype from both the male and female that were used in each experiment (Fig. 2). We conducted the verification of such hand-pairing for multiple pairs, but we have shown only the result of one pair of these because all results of genetic analyses indicated the same outcome (Fig. 2).