All nymphs of the final instar were collected by digging beneath the woods, i.e., Cryptotympana atrata beneath Populus tomentosa Carr., Platypleura kaempferi beneath Metasequoia glyptostroboides Hu & Cheng, and Meimuna mongolica beneath Pyrus xerophila Yü on the campus of Northwest A&F University, Yangling, Shaanxi Province, China, from October to December, 2013. All captured nymphs were transferred alive to a beaker and anesthetized by chilling in a 4 °C refrigerator for morphological investigation. Exuviae and adult cicadas of the above three species were also collected on their host plants from June to July, 2013, respectively, aiming to confirm the identification of the final instar nymphs of each related species based on morphology. In addition, the nymphs of the final instar, exuviae and adult cicadas of Subpsaltria yangi Chen (belonging to the subfamily Tettigadinae) and Karenia caelatata Distant (belonging to the subfamily Cicadettinae) were also collected in the same way in Mts Helan, Ningxia Hui Autonomous Region, China, in June, 2012, and at Ningshan County in Mts Qinling, Shaanxi Province, China, in July and August, 2012, respectively. All the above mentioned materials and the exuviae of Cicadetta shansiensis (Esaki & Ishihara) deposited in the Entomological Museum of Northwest A&F University, China were examined, aiming to make a comparative morphological study on the hind tibial spines among these species and also other related species which have been investigated by some authors (Hayashi 1999; Maccagnan and Martinelli 2004, 2011; Logan and Connolly 2005).

Nymphs were classified to sex by the developing genitalia at the apex of abdomen. For males, the several terminal abdominal segments of part materials were slightly extracted to show the 9^th abdominal sternite if necessary, which was partly concealed by the 10^th abdominal sternite. Observations of the morphological features were carried out using a Motic SMZ168 Stereoscopic Zoom Microscope. Photographs were taken with a scientific digital micrography system equipped with an Auto-montage imaging system and a Qimaging Retiga 2000R digital camera (CCD). Drawings were made with the aid of a camera lucida attached to the microscope.

Twenty individuals (10 males and 10 females, respectively) of each species were measured. The measurements are as follows: (1) crown length (CL) measured on dorsal view along its median line from frontoclypeal suture to posterior margin of head (Fig. 1A); (2) pronotum length (PL) measured on dorsal view along its median line (Fig. 1A); (3) pro-mesonotum length (PML) measured on dorsal view from pronotum to mesonotum along its median line (Fig. 1A); (4) head width (HW) measured on dorsal view from the outside of one compound eye to the other (Fig. 1B); (5) pronotum width (PW) measured on dorsal view at the posterior margin (Fig. 1B); (6) body length (BL) measured on lateral view from the apex of postclypeus to the distal margin of abdomen (Fig. 1C); (7) wing length (WL) measured on lateral view from rim of pronotum to apex of wing pocket (Fig. 1C); (8) postclypeus length (PCL) measured on front view from its suture with the anteclypeus to the frontoclypeal suture (Fig. 1D); (9) postclypeus width (PCW) measured on front view from one side of the outermost edge of frontoclypeal to the other (Fig. 1D); (10) fore femur length (FL) measured along the median line of its external side (Fig. 1E); (11) fore tibiae length (TL) measured along the median line of its external side (Fig. 1E); (12) femoral tooth angle (FA) measured between the longitudinal axis of femur and its posterior tooth (Fig. 1E).

Among these measurements, FA was measured using the Image Lab version 2.2.4.0 software (MCM Design, Hillerød, Denmark). The remaining measurements were taken using a vernier caliper with the accuracy of 0.02 mm.

Multivariate and univariate general linear model (GLM) analyses were conducted to determine whether morphological characters differed by species or sex. Statistical analyses were performed using SPSS 17.0.

The subfamily and tribal classification follows that of Moulds (2005). The terminology adopted to describe the structures of the foreleg was based on Duffels and Ewart (1988), and the femoral formula used to indicate the number and sequence of the teeth of fore femur was based on Maccagnan and Martinelli (2004).

The results showed that the species (Wilks’λ = 0, F = 817.078, hypothesis df = 24, error df = 90.000, P = 0) was a significant factor for all morphological characters, and that sex (Wilks’λ = 0.469, F = 4.241, hypothesis df =12, error df = 45.000, P = 0) was also significant for all characters, except for BL, FL and FA (Table 1).

There are great similarities in the gross morphology of antennae among the final instar nymphs of these three cicadas, i.e., the scape inserts in an antennal fovea of the cranium at the side of the postclypeus near an anterior tentorial pit, which is partially concealed in the antennal fovea and dorsally covered by the overhanging ridge of the vertex. However, differences also exist and are mainly shown in two aspects of these species: i) the shape of antennae (tapering apically in Cryptotympana atrata and Platypleura kaempferi (Fig. 3A, C), but apical segment of flagellum in Meimuna mongolica with full length in similar diameter (Fig. 3B)); and ii) the number of flagellar segments (seven in Meimuna mongolica (Fig. 3B), eight in Cryptotympana atrata (Fig. 3A) and nine in Platypleura kaempferi (Fig. 3C)).

Similarly, though there are many similarities in the morphology of forelegs among the final instar nymphs of these three cicadas, differences also exist and are mainly shown in four aspects: i) the shape of the base of posterior tooth on femur (extraordinarily broadened in Meimuna mongolica (Fig. 5C), but moderately broadened in Cryptotympana atrata and Platypleura kaempferi (Fig. 5B, D)), ii) the shape of the base of intermediate tooth (extraordinarily broadened in Cryptotympana atrata and Platypleura kaempferi (Fig. 5B, D), but moderately broadened in Meimuna mongolica (Fig. 5C)); iii) the shape of blade of tibia (with a large and long tooth-like point of blade of tibia in Cryptotympana atrata and Platypleura kaempferi (Fig. 5B, D), but with a very small, reduced apical toothlet-like point of blade of tibia in Meimuna mongolica (Fig. 5C)); and iv) the femoral tooth angle (about 61°, 71° and 77° in Meimuna mongolica, Cryptotympana atrata and Platypleura kaempferi, respectively).

In Tettigarctidae, three hind tibial spines were found in Tettigarcta crinita Distant. In Cicadidae, the numbers of hind tibial spines of the final instar nymphs of different species are usually the same within a subfamily, but vary among different subfamilies (Table 2). In Tettigadinae, three hind tibial spines were observed in Subpsaltria yangi. In Cicadettinae, except for Cicadetta shansiensis with three and Karenia caelatata with five hind tibial spines, four hind tibial spines were observed in all other species: Amphipsalta cingulata (Fabricius), Amphipsalta zelandica (Boisduval), Kikihia ochrina (Walker), Kikihia scutellaris (Walker), Notopsalta sericea (Walker), Rhodopsalta cruentata (Fabricius), and Carineta fasciculata (Germar). In Cicadinae, three hind tibial spines were found in the genus Mogannia Amyot & Audinet-Serville, four hind tibial spines in the genus Nipponosemia Kato, five hind tibial spines in the four investigated species (Cryptotympana atrata (Fig. 6D), Meimuna mongolica (Fig. 6E), Quesada gigas (Olivier) and Fidicina mannifera (Fabricius)), and four hind tibial spines with an additional small accessory spine internally were observed in the other four species, i.e., Dorisiana drewseni (Stål), Dorisiana viridis (Olivier), Fidicinoides pronoe (Walker) and Platypleura kaempferi (Fig. 6F).

There are some differences in the situation of an intermediate tooth on fore femur from the femoral comb, e.g., continuous from the femoral comb, or well separated from the comb. For example, the intermediate tooth is continuous from the femoral comb in Meimuna mongolica (Fig. 5C), Cicadetta shansiensis and Karenia caelatata; however, it is well separated from the comb in Cryptotympana atrata (Fig. 5B), Platypleura kaempferi (Fig. 5D) and Subpsaltria yangi.