Songs and morphology in three species of the Chorthippusbiguttulus group (Orthoptera, Acrididae, Gomphocerinae) in Russia and adjacent countries

Abstract Songs and morphology are compared between Chorthippusmiramae (Vorontsovsky, 1928) that was previously named as C.porphyropterus and two other closely related species, C.brunneus (Thunberg, 1815) and C.maritimus Mistshenko, 1951. We compare them because the calling song of C.miramae was previously shown to have song elements similar to those of other two species. One morphological character, the length of stridulatory file, appeared to be the best character to distinguish between all three species. For C.maritimus and C.miramae, we present the morphological descriptions since they are absent in the literature. We also establish the synonymy C.maritimus = C.bornhalmi Harz, 1971, syn. n. = C.biguttuluseximius Mistshenko, 1951, syn. n. In the song analysis, we analyse not only the sound but also the leg-movement pattern, which is very helpful to find a homology between various song elements. We show that the calling song of C.miramae usually contains two elements, one element being similar to the C.brunneus calling song, and another – to the C.maritimus calling song. Despite some similarities, the calling song elements in C.miramae have some peculiarities. The courtship song of C.miramae is similar to the C.brunneus song, whereas the rivalry songs of C.miramae comprise both the maritimus-like elements and the unique ones. C.miramae generally demonstrates a richer song repertoire than the other two species.


Introduction
In singing Orthoptera, the song is an important component of reproductive isolation. Acoustic signals are often used in taxonomy, when sibling species are similar in mor phology, but different in songs. In grasshoppers of subfamily Gomphocerinae, the song is produced by stroking the stridulatory file of each hind femur across a raised vein on the fore wing. It is noteworthy that using both hind legs, the grass hoppers have two separate soundproducing devices, which must be coordinated with one another. The stridulatory movements of the two legs often differ in amplitude and pattern, and the legs can exchange roles from time to time, which leads to an increase of song comple xity (e.g., Elsner 1974;Helversen and Elsner 1977;Helversen and Helversen 1994). To distinguish cryptic grasshopper species, not only the sound recordings but also the recordings of the leg movements are used by various authors (Helversen 1986;Gotts berger and Mayer 2007;Helversen 2003, 2009;Willemse et al. 2009;Vedenina et al. 2012;Tarasova et al. 2021).
The main subject of the current study is one species of the biguttulus group, C. porphyropterus, which we name as C. miramae (Vorontsovsky, 1928nec Ramme, 1936, 1951, and two closely related species, C. brunneus and C. maritimus, whose songs resemble song elements of C. miramae. Since in Russia and adjacent countries C. brunneus, C. maritimus and C. miramae often occur with two other species of the biguttulus group, C. biguttulus and C. mollis, we describe the main morphological differences from the latter two species as well.

Materials and methods
Localities where material was collected are shown in Fig. 1. All localities were num bered and all numbers are listed in Results, in the paragraph "Material examined". On the map, however, only localities with song recordings are numbered.

Morphological analysis
In all specimens studied, we measured the following morphological characters: the lengths of pronotum, forewing and hind femur, the width of costal and subcostal (C & Sc) areas of fore wing, the distance from the center of stigma to the tip of fore wing, the length of stridulatory file and the distance from the most distal stridulatory peg to the tip of knee (Table 1, Fig. 2). In 10 specimens of each sex and species, the body length, the width of fore wing and the number of stridulatory pegs were measured. These morphological features have been chosen on the basis of the literature (Ragge et al. 1988;Bukhvalova 1993;Benediktov 1999;Willemse et al. 2009).The length of pronotum was measured along the midline. The length of forewing was measured from the humeral plate to the tip of the wing; the widths of the C & Sc areas were measured at the point where costal area was of the greatest width (Fig. 2B). The length of hind femur was measured from the anterior margin of the upper basal lobe to the hind margin of the upper kneelobe; the length of stridulatory file was measured from the most proximal peg to the most distal peg; the distance between stridulatory file to the tip of the knee was measured from the most distal peg to the hind margin of the upper kneelobe ( Fig 2C). Morphological studies were carried out with an MBS9 light microscope at 8-56× magnification using an ocular micrometer. Material for the morphological analysis was taken from the Zoological Museum of Moscow State University (ZMMU) and the personal collections of V. Vedenina (CV).
All statistical analyses were performed using Excel 2016 and STATISTICA 12.0.0. To visualize and clarify the differences in morphology between the three species, a principal component analysis (PCA) was applied to 6 morphological cha racters (Fig. 3E).   (Vorontsovsky) from Orenburg region A fore wing with complete venation B fore wing with main veins; C hind leg. The measured mor phological characters are indicated with arrows and brackets.

Song recordings and analysis
The calling song was recorded from an isolated male; the courtship song was recorded when a male was sitting near a female; the rivalry song was recorded from males sitting near each other. Recordings of the calling and rivalry songs in the field were carried out with a MD382 microphone (upper frequency limit 12.5 kHz; before 2008), or a Spirit IM01 microphone (upper frequency limit 20 kHz), and an Elektronika3021 cassette recorder (upper frequency limit 10 kHz; before 2005), or a Sony Walkman MZNH900 minidisk recorder (sampling frequency 44.1 kHz). The signals were A/D converted with a PC card L305 (LCard Ltd., Russia). The ambient temperature near a singing male in the field was 20-40°C.
During stridulation of the males studied in laboratory, both the sound and the hind leg movements were recorded with a custombuilt optoelectronic device (Helversen and Elsner 1977;Hedwig 2000). A piece of reflecting foil was glued to the distal part of each hind leg femur of a male and two optoelectronic cameras were focused on the illuminated reflecting dots. Each camera was equipped with a positionsensitive photodiode that converted the upward and downward movements of the hind legs into voltage signals. These signals, together with the recordings of the sounds (a micro phone type 4191, ½ inch; a conditioning amplifier type 2690; Brüel & Kjaer, Naerum, Denmark), were A/Dconverted with a custombuilt PC card. The sampling rate was 1325 Hz for recording the stridulatory movements and 100 kHz for sound recordings. In the laboratory, the ambient temperature near a singing male was 30-32°C.
For the song description we used the following terminology (Figs. 4,6): pulse -the sound produced by one stroke of a hind leg (the shortest measurable unit or the first order unit); syllable -the sound produced by one complete up and down movement of the hind legs, starting when the legs leave their initial position and ending when the legs return to their original position and representing the repeated unit of a stable structure (the secondorder unit); echeme -series of consistent syllables separated by pauses (the thirdorder unit). We measured three characters in C. brunneus (echeme rate, echeme duration and pulse rate), four characters in C. maritimus (echeme rate and duration and syllable rate and duration) and seven characters in C. miramae (echeme rate, echeme duration and pulse rate for the brunneuslike echeme and echeme rate and duration and syllable rate and duration for the maritimus-like echeme). To visualize and clarify the differences in calling song between the three species, a PCA was applied to 5 song characters (Fig. 5E). We did not use echeme rate for both types of echemes because not all recorded males produced several echemes. When a character was equal to 0, we changed it to 0.01 by convention because we only used the logarithmic values for PCA.

Nomenclatural notes
The names Stauroderus mollis porphyroptera and S. miramae (both currently included in the subgenus Glyptobothrus Chopard, 1951) were described by Vorontsovsky (1928a, b) in two papers on grasshoppers from Orenburg published in the same issue. S. mollis porphyroptera was described as a new variation and designated as a var. nov.; therefore, the authorship of Vorontsovsky in this case is beyond doubt (Vorontsovsky 1928a, p. 12). Vorontsovsky attributed the authorship of the S. miramae to Ramme, with the following comment: "For the identification of this species, as well as the form, I identi fied as a variety of the species Stauroderus mollis, I take the opportunity to express here my deep gratitude to E.F. Miram, who informed me that S. miramae has just been described from Crimea by Dr. Ramme as a new species." (Vorontsovsky 1928a: 12, footnote). Actually, Ramme mentioned Chorthippus miramae for the first time only in 1939 without a description, specifying that this species "will be described in the near future" (Ramme 1939: 131). Therefore, the name C. miramae Ramme, 1939 is sug gested to be a nomen nudum. Only in 1951, Ramme described this species based on material from Ukraine, Crimea, Southern and Southeastern European Russia, Cauca sus, and Transcaucasia, with the type locality in Southern Crimea (Ramme 1951). On the other hand, Vorontsovsky (1928b) presented a short description of C. miramae. For this reason, he is the author of the taxon from Orenburg in spite of the fact that he attributed the authorship to Ramme. Further, the study of signals showed that C. miramae Vorontsovsky and C. miramae Ramme represent the different species (see below).
Summarizing the following three taxa were described in the papers mentioned above: C. mollis porphyroptera (Vorontsovsky 1928) from the type locality in Oren burg, C. miramae (Vorontsovsky, 1928) from the type locality in Orenburg, and C. miramae Ramme, 1951 from the type locality in Southern Crimea.
According to the study of Bukhvalova (1993) based on investigation of the male songs, the Chorthippus biguttulus group includes 5 species in Russia: C. biguttulus (Lin naeus, 1758), C. brunneus (Thunberg, 1815), C. mollis (Charpentier, 1825), C. miramae Ramme, 1939 andC. yersini Harz, 1975. The study of the songs of specimens from Crimea, Southern European Russia, North Caucasus, Central Asia, and the Rus sian Far East showed that C. miramae Ramme, 1939sensu Bukhvalova (1993 is a good species, which is widespread throughout the southern part of Russia and adjacent terri tories. It was described as C. biguttulus meridionalis Mistshenko, 1950 from mountains of Central Asia (Mistshenko 1950), as C. miramae Ramme, 1951 from Crimea, andas C. maritimus Mistshenko, 1951 from the Russian Far East (BeyBienko and Mist shenko 1951). However, it differs from the taxa described by Vorontsovsky (1928a, b) from Orenburg (Bukhvalova 1998). The name C. biguttulus meridionalis Mistshenko, 1950 is invalid, since it is a junior homonym of C. bicolor var. meridionalis (Fruhstor fer, 1921). The name C. miramae Ramme, 1951 is a junior homonym of C. miramae (Vorontsovsky, 1928). As a result, the valid name of this taxon should be C. maritimus Mistshenko, 1951. It should be also noted that some authors improperly considered the date of publication of the name C. miramae Ramme to be 1939 (BeyBienko andMistshenko 1951;Harz 1975;Bukhvalova 1993;Wosnessenskij 1996) and treated this taxon as a subspecies of C. brunneus (BeyBienko and Mistshenko 1951;Harz 1975).
C. bornhalmi Harz, 1971 was described from Croatia in the Balkans and has been shown to occur from Italy to Turkey (Willemse et al. 2009;Sirin et al. 2010). The range of C. maritimus extends from southern Ukraine to the Russian Far East. In the current study, we compare the morphology and songs in C. bornhalmi (from Bulgaria and Greece) and C. maritimus, and establish the synonymy C. maritimus Mistshenko, 1951= C. bornhalmi Harz, 1971, syn. n.
C. biguttulus eximius Mistshenko, 1951 was described from Sukhumi, Abkhazia (Mistshenko 1951). A study of songs from the environs of the type locality (loc. 34 in Fig. 1) showed that this subspecies also is identical to C. maritimus. Since C. maritimus (as C. biguttulus maritimus) and C. biguttulus eximius were described in the same pa per, we choose a valid name C. maritimus for this species and establish the synonymy C. maritimus = C. biguttulus eximius syn.n.
C. miramaellus Wosnessenskij, 1996 and C. sinuatus Mistshenko andWosnessen skij, 1996 proposed by Wosnessenskij (1996) to replace C. miramae Ramme, 1951 andC. biguttulus meridionalis Mistshenko, 1950 respectively, are the junior synonyms of C. maritimus (Bukhvalova 1998). We suggest that C. maritimus tsejensis Bukhvalova, 1993 from North Ossetia, North Caucasus (Bukhvalova 1993) and C. meridionalis karakalensis Sytshev and Woznesenskij, 1995 from Southwestern Turkmenistan (Sychov and Voznesensky 1995) also belong to C. maritimus; however, additional studies are needed to clarify their status. It should be noted that M.M. Sychov and A.Yu. Voznesensky transliterated their own names in different ways in different papers, both in English and Latin; here we present their original spellings from the corresponding papers. Benediktov (1999) reinvestigated material from Orenburg used by Vorontsovsky and concluded that C. mollis porphyroptera (Vorontsovsky, 1928) and C. miramae (Vo rontsovsky, 1928) are synonyms. Benediktov (1999) compared the lengths of stridu latory files (the most characteristic feature of this species) in the type specimens of Vorontsovsky and found them to be identical. He proposed C. porphyropterus as the valid name, raising its rank, and changing its gender ending. However, according to chapter 24 of the International Code of Zoological Nomenclature (1999), when syno nyms are established simultaneously, but are proposed at different ranks, the name proposed at a higher rank takes precedence. Consequently, the valid name of the taxon from Orenburg should be C. miramae (Vorontsovsky, 1928). Also, Benediktov (1999) established the synonymy C. porphyropterus = C. biguttulus forma tomensis Berezhkov, 1956, proposed the new combination C. porphyropterus euchedickei Helversen, 1989for C. biguttulus euchedickei Helversen, 1989, and pointed out that C. yersini Harz, 1975sensu Bukhvalova, 1993 is conspecific with C. miramae (Vorontsovsky, 1928). The true identity of C. biguttulus forma tomensis described known only from the bank of the Tom' River near Ust'Iskitim, ca. 85 km south of Tomsk, Western Siberia (Be rezhkov 1956), requires confirmation from song recordings from the type locality. The combination C. biguttulus euchedickei was restored by Willemse et al. (2009). The conspecificity of C. yersini sensu Bukhvalova, 1993nec Harz, 1975and C. miramae (Vorontsovsky, 1928 are absolutely correct. Later on, Benediktov (2005) established the synonymy C. porphyropterus = C. brunneus mistshenkoellus Oliger, 1974 on the basis of investigation of the types of C. brunneus mistshenkoellus Oliger, 1974 from Tolyatti, Samara region. However oscil lograms of the song of C. maritimus from Tolyatti (Benediktov and Mikhailenko 2017) cast doubt on this synonymy.
The status of C. brunneus (Thunberg, 1815) is unambiguous. In addition to the nominotypical subspecies, this species includes C. brunneus mistshenkoellus mentioned above and C. brunneus brevis Klingstedt, 1939 from Southern Finland (Klingstedt 1939), the statuses of which require further clarification.
In the current paper, we consider the following three taxa: C. brunneus (Thunberg, 1815), C. maritimus Mistshenko, 1951, and C. miramae (Vorontsovsky, 1928). C. mari timus tsejensis Bukhvalova, 1993, C. meridionalis karakalensis Sytshev et Woznesen skij, 1995, C. brunneus mistshenkoellus Oliger, 1974, and C. brunneus brevis Klingstedt, 1939 are excluded from the consideration since their statuses are unclear. Distribution. (Fig. 1). The range of this species extends from Europe to the southwestern part of European Russia. In Europe this species occurs over a wide range, excluding the central and southern part of the Iberian Peninsula and Greece (Ragge andReynolds 1988, Sirin et al. 2010). Further to the east, it occurs in the Baltic republics, Belarus, Moldova, and Ukraine. The eastern border of the range lies on the longitude of the Saratov and Kostroma regions of Russia. The species tends to be mesophilic. The range of C. brunneus overlaps with that of C. maritimus in southeastern Europe, Moldova, Ukraine, and the southeastern part of Euro pean Russia.

Gryllus brunneus
Recognition. (Table 1, Fig. 3). The males of C. brunneus can be distinguished from the males of C. miramae and C. maritimus by a short stridulatory file (Fig. 3A). This, however, is not applicable to the females (Fig. 3B). Both sexes of C. brunneus are characterized by the lowest number of stridulatory pegs (58-93 in ♂, 51-95 in ♀.).
In comparison with C. miramae and C. maritimus, both sexes of C. brunneus tend to have the shortest pronotum, the narrowest C & Sc areas of fore wing, and the stigma closest to the wing tip (Table 1). The PCA applied to 6 characters shows a substantial overlap between C. brunneus and C. maritimus (Fig. 3C,D). In PCA, however, we do not use the number of stridulatory pegs, since this value was measured for a small number of males. Meanwhile, it was previously shown that C. brunneus can be easily distinguished from all other species of the C. biguttulus group by the lowest number of stridulatory pegs, especially in nominate subspecies (Oliger 1974;Ragge and Reynolds 1988;Bukhvalova 1993;Willemse et al. 2009).
Calling song (Table 2, Figs. 4,5). The calling song of C. brunneus consists of seve ral short echemes repeated at the rate of about 0.3-2.1 /s. Each echeme lasts on ave rage 0.1-0.4 s and has a relatively stable temporal structure. It consists of short pulses, which are grouped into 4-7 syllables (Fig. 4C). The gaps between the subsequent syl lables can't be traced by the sound analysis, but they can be distinguished by the analy sis of the leg movements. The two legs are moved with a large phase shift, and some times almost alternately (Fig. 4E). Each leg generates one short pulse during a straight upstroke, whereas two short pulses are produced during a twostep downstroke. The pulse duration and the pulse rate vary in the ranges of 7-8 ms and 91-111/s, respec tively (at the temperature 29-30°C). The population from loc. 40, shows an extremely long echeme duration and low echeme and pulse rate (Table 2). Notably, the values are relatively stable within the same population.
Courtship and rivalry songs. The courtship and rivalry (Fig. 4F, G) songs of C. brunneus are similar to the calling song.

Chorthippus maritimus Mistshenko
Chorthippus miramae Ramme, 1939: 131, nomen nudum.   Chorthippus meridionalis Mistshenko, 1950    Distribution. (Fig. 1). C. maritimus is a widespread transPalearctic species. It includes C. bornhalmi from the Balkans and Anatolia (Willemse et al. 2009;Sirin et al. 2010;Skejo et al. 2018) and as C. biguttulus eximius from Sukhumi, Abkhazia (Mist shenko 1951). It also occurs in Moldova and southern Ukraine (Heller et al. 1998). In the territory of Russia, its range stretches from Krasnodarsky krai to Sakhalin along the southern border. This species also occurs in Caucasus, southern Kazakhstan, Turk menistan, very likely Uzbekistan, Kyrgyzstan, Mongolia, northerneast China, Korea and Japan (Storozhenko 2002). The ranges of C. maritimus and C. brunneus overlap in Eastern Europe, Ukraine and the southeastern part of European Russia. Moreover, C. maritimus and C. brunneus often occur syntopically. The range of C. maritimus also overlaps with the range of C. miramae in the southeastern part of European Russia and in surroundings of the Baikal Lake, however, they do not occur in the same biotopes.
Recognition. (Table 1, Fig. 3). The males of C. maritimus can be distinguished from the males of C. brunneus by the longer stridulatory file (Fig. 3A) and the higher number of stridulatory pegs (see Description). These characters are also mentioned as the dis tinguishing features between C. brunneus and C. bornhalmi by other authors (Willemse et al. 2009;Skejo and Ivcovic 2015). The length of stridulatory file in C. maritimus is intermediate between those in C. miramae and C. brunneus. Both sexes of C. maritimus also tend to have the longest fore wings and pronotum in comparison with C. miramae and C. brunneus (Table 1). C. maritimus can be also distinguished from other species of the biguttulus group by the narrower costal area of fore wing. By contrast, C. maritimus differs from C. mollis by the wider costal area of fore wing and by the lower density of stridulatory pegs (Bukhvalova 1993;Oliger 1974). C. bornhalmi and C. biguttulus eximius are not different in morphology from C. maritimus from Ukraine and Russia.
Description. (Table 1, Fig. 3). The head structure as in genus. Ratio length of verti cal diameter of eye to maximum length of foveolae 2.8-3.4 in ♂, 3.0-3.2 in ♀; ratio minimum interocular distance to length of subocular groove 0.6-0.8 in ♂, 0.7-0.9 in ♀. Antennae filiform. Prozona is slightly shorter than metazona; median carina is distinct and continuous. Lateral pronotal keels are distinctly incurved, ratio between minimum and maximum widths 2.3-2.6 in ♂, 2.3-2.9 in ♀. In western populations keels are more angled, min/max width ratio up to 3.0. Tympanal aperture slitlike, 2.3-2.8 times in ♂, 2.6-2.8 in ♀ as long as broad. Fore and hind wings well developed in both sexes, wings far surpassing the apices of the hind knee. Costal area of fore wing has maximum width in the middle part or in the last third of the wing. Subcostal area narrow, its width 0.25-0.3 mm in ♂, 0.15-0.2 mm in ♀ (measured on the line of maximal width of costal area). Ratio width of fore wing to C & Sc areas 3.1-3.5 in ♂, 4.4-4.7 in ♀. Apical constriction (distance from C and Sc confluence to the wing tip) prolonged, ratio length of apical constriction to the wing length 3.3-3.8 in ♂, 3.5-3.8 in ♀. Stigma far from the wing tip, ratio length between stigma center and the wing tip to the wing length 2.4-2.7 in ♂, 2.3-2.5 in ♀. Hind femur gracile, ratio femur length to maximum width 4.4-4.6 in ♂, 4.4-4.7 in ♀. Stridulatory file consists of one row, its length nearly equal to the distance between last peg and tip of hind knee. The number of stridulatory pegs 100-168 in ♂, 104-157 in ♀. Body coloration va ries from light straw to dark brown, sometimes with a red tone. The ventral side of the body lighter than dorsal side, and densely pubescent. Fore wings smoky, with a few dark spots in M area. Hind wings transparent at the base and slightly smoky in apical part, distal half of C area smoky or brownish. Hind femur in the inner side with black lengthwise line. Hind knees dark brown or blackish, particularly on upper lobe. Hind tibiae orange or reddish.
Measurements in mm. Body length: 15-18 in ♂, 19-26 in ♀, pronotum length: 3.1-3.4 in ♂, 4.1-4.4 in ♀, fore wing length: 14.1-15.5 in ♂, in 17.2-18.5 in ♀, fore wing width 3.1-3.4 in ♂, 3.2-3.5 in ♀, hind femur length: 9.8-10.6 in ♂, 12.8-14.1 in ♀. (Table 3, Figs. 5,6). The calling song of C. maritimus usually contains one to several echemes of median duration ranged from 1 to 4 s. In some populations (49,62,63), however, the median echeme duration is higher, ranging between 5-11.1 s ( Table  3, Fig. 5C). The echeme rate also greatly varies between different populations (0.05-0.42 / s). The number of syllables per echeme varies in the range of 15 to 40, in populations with prolonged echemes -in the range from 40 to 70. The syllable duration is relatively stable within the same population; however, its median duration can vary between the populations in the range of 86-162 ms (Fig. 5D). At the beginning of each echeme, the sound is very soft, but then it reaches maximum loudness after the first third of the echeme duration, being constant until the echeme end (Fig. 6D). The syllables are gener ated by the leg movements with a small phase shift, which comprise the straight upstroke and stepwise downstroke (Fig. 6E, F). Both upstroke and downstroke have the similar duration. The leg upstroke generates a noisy sound with unclear structure and slightly increasing amplitude; the stepwise downstroke generates 4-5 distinct pulses. The pulses,  however, can be sometimes fuzzy. The durations and rates of echeme and syllable in C. bornhalmi (from loc. 6) and in C. biguttulus eximius (from loc. 34) fall into the range of values in C. maritimus from several localities (Table 3, Fig. 5C, D). The syllable structure is also quite similar in C. bornhalmi (Fig. 6E) and C. biguttulus eximius (Fig. 6F). Courtship song. The courtship song of C. maritimus is similar to the calling song. Rivalry song (Fig. 6G, H). The rivalry song of C. maritimus contains echemes of a shorter duration than the calling song. In some males the first syllable of the rivalry echeme lasts 1.5-2 times as long as the subsequent syllables, which results from the prolonged first downstroke (Fig. 6H). The pulses produced during the first downstroke are repeated twice as slowly as the pulses of the subsequent syl lables. The subsequent 2-8 syllables are of the same structure as the syllables in the calling song.
Recognition. (Table 1, Figs. 2, 3). C. miramae can be distinguished from most species of the biguttulus group by remarkably long stridulatory file (Fig. 2C). This feature was previously shown by Benediktov (1999), who described the last distal stridulatory peg to be situated at least at a level of the second tibial spine when tibia is attached to femur. Within the biguttulus group, a similarly long file is only shown in C. biguttulus euhedickei von Helversen, 1989, that occurs in the southern Balkans and Anatolia and in C. maroccanus Nadig, 1986, that occurs in North Africa (Ragge and Reynolds 1988;Willemse et al. 2009). The latter two taxa, however, are quite different from C. miramae in other morphological characters and songs. In other species of the biguttulus group, the length of stridulatory file is noticeably shorter, and the last distal stridulatory peg is situated at least at the level of the 4 th tibial spine when the legs are bent (Benediktov 1999). Notably, in C. miramae, the number of stridulatory pegs is only slightly higher than in C. maritimus, and can't be considered as a good character. C. miramae tends to have the longest distance between stigma and the wing tip, and the broadest width of C & Sc areas in comparison to C. maritimus and C. brunneus. The PCA based on 6 morphological characters shows that C. miramae represents a separate cluster from C. maritimus and C. brunneus, but it is stronger in males than in females (Fig. 3C, D).
Description. (Table 1, Figs. 2, 3). The head structure as in genus. Ratio length of vertical diameter of eye to maximum length of foveolae 3.2-3.6 in ♂, 2.8-3.2 in ♀; ratio minimum interocular distance to length of subocular groove 0.6-0.8 in ♂, 0.7-1.0 in ♀. Antennae filiform. Median carina distinct and continuous. Prozona slightly shorter than metazona. Lateral pronotal keels distinctly incurved, ratio minimum to maximum widths 2.1-2.6 in ♂, 2.4-2.6 in ♀. Tympanal aperture 2.8-3.3 times in ♂, 2.8-3.4 in ♀ as long as broad. Fore and hind wings well developed in both sexes, wings far surpassing the apices of the hind knee. Width of costal area of fore wing reaches its maximum in the middle or in the last third part ( Fig. 2A, B). Width of subcostal area 0.3-0.35 mm in ♂, 0.2-0.23 mm in ♀ (measured along the line of maximal width of costal area). Ratio width of fore wing to width of C & Sc areas 3.0-3.2 in ♂, 4.3-4.5 in ♀. Length of apical constriction (distance from C and Sc confluence to the wing tip) is a quarter of the wing length. Ratio length between stigma center and the wing tip to the wing length 2.1-2.8 in ♂, 1.8-1.9 in ♀. Hind femur gracile, ratio femur length to maximum width 4.5-4.9 in ♂, 4.6-4.9 in ♀. Stridulatory file remarkably long in both sexes: distance between the last peg and the knee tip 2-2.7 times in ♂, 1.7-2.4 in ♀ as large as length of stridulatory file. In males, stridulatory pegs form one row and have different density along the file (Fig. 2C). Most proximal part of stridulatory file starts with several rare and dispersed pegs that are followed by more densely disposed pegs. The second part of stridulatory file more prolonged, consisting of more rare pegs with stable interpeg intervals. In the third, most distal part the peg density decreases pro portionally to the length of stridulatory file, and the pegs often do not lay in one raw. In females, stridulatory pegs arranged in one row and distributed rarer than in males. The peg density decreases from the proximal towards the distal parts. The number of stridulatory pegs 118-182 in ♂, 98-157 in ♀. Body coloration similar to coloration of C. maritimus.
Calling song. (Table 4, Figs. 5, 7). The calling song of C. miramae includes the two types of randomly alternating echemes, typical maritimuslike and optional brunneus like echemes. The first echeme type was present in the songs of all 34 males recorded, the second echeme type -in the songs of 28 males. The song usually starts with the maritimuslike echeme, which is similar to the C. maritimus calling song, but lasting shorter (the median duration varies in the range of 0.3-2.9 s). The number of syllables per echeme varies in the range of 5 to 35. Each echeme starts with the lowamplitude syllables. In short echemes, the amplitude reaches its maximum in about the echeme middle (Fig. 7F). In long echemes, the amplitude gradually increases, and keeps a constant level after about one quarter of an echeme (Fig. 7G). The syllables are about 1.5 times as short as the syllables in C. maritimus, lasting in the range of about 66-114 ms ( Table 4). The syllable duration is rather stable within one population; however, it is more variable between populations. Oscillographic analysis shows no distinct pulses within the syllables in some populations, whereas distinct pulses are visible on the oscillograms of the songs from other populations. The shift between the two legs is greater in C. miramae than in C. maritimus (Fig. 7I, J).
The brunneuslike echemes are more often produced by the males from the Sibe rian and the easteuropean Russian populations, but they are rare in the songs from northern Kazakhstan. The echeme duration in C. miramae is almost three times as high as in C. brunneus (Fig. 5A). Similarly to C. brunneus, the C. miramae echeme consists of the short pulses, the amplitude of which gradually increases, reaching maximum intensity at about half of its duration, and then gradually decreases towards the end. The pulse duration and the pulse rate in C. miramae are almost the same as in C. brunneus (9-13 ms and 77-96 /s respectively, data are given for 29-30°C). However, the leg movement patterns are different in two species. In C. miramae, the brunneuslike echeme is produced by simple up and down legmovements that vary in amplitude and duration (Fig. 7J). In C. brunneus, each leg generates a simple upstroke but a twostep downstroke (Fig. 4D). The oscillographic analysis of the C. miramae song shows that the pulses highly vary in amplitude and duration, whereas the pulses in the C. brunneus song are much more stable in these parameters. In some males of C. miramae, the pulses are tended to group into syllables; the pulse number per syllable is unstable (Fig. 7H).
The order of the two echeme types in the C. miramae song is erratic, though there are some common variants in different populations. For example, several maritimus like echemes are followed by one brunneuslike echeme (Fig. 7D). Another variant implies alternation of the two echeme types. A rarer case is when one maritimuslike echeme is followed by several echemes of the second type (Fig. 7A, E). The intervals between echemes of the same type may exceed the echeme duration 1.5-3 times for  *n/a -nonapplicable the maritimuslike echemes, and 3-5 times for the brunneuslike echemes. An interval between the maritimuslike and the subsequent brunneuslike echemes can be very short (Fig. 7F, J), or can exceed the echeme duration 3-5 times. Courtship song and female response song. (Fig. 8). The courtship song of C. miramae consists of the brunneuslike echemes. However, the courtship sound is much soft er than in the calling song. The courtship echemes are shorter than in the calling song, not reaching 1 s (the median duration is about 0.4 s). The echemes are usually repeated at the rate of about 0.2-0.6/s, and their duration varies from 0.7 to 1.0 s. Pulses are short (6-9 ms), frequent (repeated at the rate of 61-95/s), and of a low amplitude (Fig. 8F). In some cases, the leg movements do not produce any sound at all (Fig. 8A, D).
A female produces the brunneuslike song in response to the male courtship or rivalry song (Fig. 8A, B). The female alternates her response echemes with the male echemes (Fig. 8D). The duration of the female echeme is similar to that in the male courtship, or 1.5-2 times longer than in the male courtship. The leg movement pattern in the female response song is similar to that in the male courtship song, but less regu lar (Fig. 8E, F). The pulses are longer (10-21 ms) and repeated at the rate of 43-77/s, especially in the first third of the echeme (Fig. 8D, E).
Rivalry song. (Fig. 9). Several males of C. miramae sitting close to each other pro duce a diversity of echemes of different duration, structure and leg movement pattern. For example, one can find a rivalry song similar to that of C. maritimus, which starts with the prolonged first syllable, which results from the prolonged first downstroke (Fig. 9D, E). The pulses produced during the first downstroke follow twice as slowly as the pulses of the subsequent syllables. The subsequent syllables are of the same struc ture as in the maritimuslike echeme of the calling song.
Most often, the males produce single syllables similar to the first one with distinct pulses described above. These syllables are repeated at the rate of about 2-2.5 /s (Fig.  9F, G). Notably, the two legs may produce different number of the up and down strokes. Rarely, the males produce the maritimuslike echeme without the first syllable of distinct pulses (Fig. 9H).
The same male may produce echemes of different structure in the rivalry situ ations. Some females are actively responding to the male rivalry songs.

Discussion
What is the function of the long stridulatory file?
The morphological analysis conducted in the current study shows that one character, the length of stridulatory file, appears to be the most reliable character to distinguish C. miramae, C. maritimus and C. brunneus. The difference in the file length between C. maritimus and C. brunneus can be explained by the difference in the peg number. By contrast, the extremely long file in C. miramae is not due to the significant increase in the peg number, but due to the more widely spaced pegs in the distal part of the file.
The long stridulatory files are known in some other species of the biguttulus group. C. biguttulus euchedickei from the southern Balkans and northwestern Anatolia (Wil lemse et al. 2009) andC. maroccanus Nadig, 1976 from northwestern Africa (Ragge and Reynolds 1988), are also characterized by extraordinary long stridulatory files and the widely spaced distal pegs. In C. brunneus brevis Klingstedt 1939 from Southern Finland and northeast Russia (Ragge and Reynolds 1998;Benediktov 2017), the file length is much greater than in the nominate subspecies. In C. brunneus brevis, how ever, the increased length of the stridulatory file can be explained by the increase in the peg number. In one endemic of the biguttulus group in Anatolia, C. relicticus Sirin, Helversen & Ciplak, 2010, the peg number was shown to be extremely high (175-225 in male, 194-245 in female;Sirin et al. 2010). Unfortunately, the authors did not measure the length of stridulatory file in C. relicticus, but we assume that the file could be also long.
What could be a function of the long stridulatory file? The different parts of the long file can be used during stridulation to produce various song elements (Vedenina et al. 2007;Vedenina and Helversen 2009). This, however, is only evident in C. biguttulus euchedickei (Helversen 1989;Willemse et al. 2009). The calling song of this species consists of 1-3 typical loud echemes, similar to those in the nominate form, that are followed by 1-5 softer aftersongs (quiet parts of the song produced at the end of singing). Aftersongs are produced at a low position of the legs, and presumably the distal pegs are used for sound generation. In C. miramae, however, the long stridula tory file does not seem to be specifically involved in sound generation: at least, no song elements were found to be generated by distal pegs only. The leg movements in other species of the biguttulus group with a long file or high peg number have not been stu died or studied only for certain song types.
It is noteworthy that stridulatory pegs function not only as a mechanic part of the stridulatory apparatus, but also as the mechanoreceptors (Hustert et al. 1999). It was shown in two distantly related species of Gomphocerinae, C. biguttulus and Syrbula montezuma (Saussure, 1861), that two sensory cells innervate each peg in themale and each tubercle in the female. These mechanoreceptors can deliver specific propriocep tive information about the contact between the stridulatory file and the vein of the fore wing. A subtle sensory control is required for measuring the pressure of the leg against the wing. The current study of the C. miramae songs shows that the loud maritimuslike echeme is apparently produced by legs being more pressed to the wings than during the softer brunneuslike echeme (Fig. 7). The latter echeme may be also produced with different leg pressure depending on the calling or courtship behavior; during courtship, the sound can be even absent despite the appropriate leg movements (Fig. 8).
In species of the C. albomarginatus group, the peg number and density differ only at the proximal parts of the stridulatory files (Vedenina and Helversen 2009). The vari ous species of this group produce different and very conspicuous visual displays in a particular part of the courtship: during the stroke with the hind tibiae, the femora are kept at the extrahigh, almost vertical, position. At this moment, the proximal pegs may participate in producing sound. Therefore, the divergence in visual display and the changes in the peg morphology in the albomarginatus group could strengthen each other. A similar assumption can be made for the evolution in song and the stridulatory file structure in the biguttulus group.

Peculiarities of the Chorthippus miramae song
The calling song of C. miramae is conspicuously different from the songs of C. brunneus and C. maritimus, by the presence of two types of echemes, which were recorded in 82% of males. In the calling songs of 18% of C. miramae males, however, only the maritimuslike echemes were recorded. The latter specimens, however, clearly belong to C. miramae based on morphology and courtship and rivalry songs.
Until now, the calling songs of C. miramae were only presented under the name C. yersini by Bukhvalova (1993) and C. porphyropterus by Benediktov (2005). Both authors claim the presence of the two echeme types. From the oscillograms presented, one could see many similarities with the songs of C. maritimus and C. brunneus. The current song analysis that includes not only the sound but also the legmovement analysis indicates that both maritimuslike and brunneuslike elements have some peculiarities in the C. miramae song. The maritimuslike echemes rarely show the distinct pulses within syllable, whereas such pulses in the calling song of C. maritimus are typically present. This may be determined by the larger shift between the two legs in C. miramae than in C. maritimus. The brunneus like echeme in the C. miramae song is produced by simple up and down legmovements, whereas each leg generates a simple upstroke but a twostep downstroke in the C. brunneus song. The sound pulses, however, are of a similar temporal structure in both species. The similarities between the calling songs could explain why C. miramae is not found together with C. maritimus and C. brunneus in the same biotopes (despite the latter two species often occur syntopically). According to the concept of 'acoustic niches' (Bukhvalova 2006;Tishechkin and Bukhvalova 2009), the combination of the syllable rate and syllable tem poral pattern determines the species 'place' in the acoustic environment of the grasshopper community. Since these song parameters overlap within the species pairs C. miramae / C. maritimus and C. miramae / C. brunneus, the absence of each pair in the same biotope is not surprising.
C. miramae generally demonstrates a richer song repertoire than the other two spe cies. The courtship song of C. miramae is similar to the brunneus-like echeme, but the sound is very soft. In some cases, leg movements of C. miramae do not produce any sound at all, which may be interpreted by a female as a visual display. Notably, there is no specific courtship song in both C. brunneus and C. maritimus. As for a rivalry song, this is present in C. maritimus and C. miramae but not in C. brunneus. The rivalry song of C. miramae is similar to that in C. maritimus. It comprises the first syllable with dis tinct pulses lasting longer than the subsequent syllables with fuzzy pulses. More often, however, the rivalry repertoire in C. miramae includes short syllables similar to the first one in the maritimuslike echeme but repeated at the rate of 2-2.5/s.
In most species of the biguttulus group, the rivalry song is similar to the calling song (Ragge and Reynolds 1998). The rivalry song may be shorter than the calling song, but similar in temporal structure and usually does not contain any new elements. Only C. maroccanus produces a characteristic rivalry song containing two elements, one element similar to the calling song and the second unique element. Thus, C. miramae is another species of this group, in which the rivalry song is principally different from the calling song.

The relationship of Chorthippus miramae with other members of the biguttulus group
It has been suggested that the biguttulus group comprises many young, closely related species, some of which may be of hybrid origin. Some species of this group were found to hybridize in nature (e.g., Ragge 1976;Bridle and Butlin 2002;Kleukers et al. 2004;Nolen et al. 2020), whereas some of them were hybridized in laboratory in nochoice conditions and produced viable and fertile offspring (Helversen and Helversen 1975;Gottsberger and Mayer 2007). The similarity of the C. miramae song with the songs of C. brunneus and C. maritimus might suggest a hybrid origin of C. miramae.
One of the most well studied hybrid zones within the biguttulus group is a hybrid zone between C. jacobsi and C. brunneus in northern Spain (e.g., Bridle and Butlin 2002;Saldamando et al. 2005;Bridle et al. 2006). The calling song of C. jacobsi is similar to the song of C. maritimus, but of a shorter duration. Songs of F1, F2 and backcross hybrids between C. jacobsi and C. brunneus were intermediate between the songs of both parental species in all song parameters (Saldamando et al. 2005). At the same time, no combination of the parental song elements was found in the hybrid songs. Similarly, natural hybrids between C. maritimus (named as C. bornhalmi) and C. brunneus from northeastern Italy were shown to sing intermediate songs (Kleu kers et al. 2004). In European Russia and Ukraine, these two species often occur in the same biotope allowing them to hybridize. We suggest that the C. brunneus song with unusually long echeme duration and low echeme and pulse rate recorded from loc. 40 (Table 2) may be also attributed to the hybrid. F1 hybrids between C. maritimus and C. brunneus bred in our laboratory only revealed intermediate songs (unpublished data). It is therefore unlikely that C. miramae could evolve from the hybrids between C. maritimus and C. brunneus.
The number of stridulatory pegs in hybrids between C. jacobsi and C. brunneus (Sal damando et al. 2005) and C. brunneus and C. bornhalmi (Kleukers et al. 2004) were shown to be also intermediate between those in parental species. In C. miramae, the peg number is similar to that of C. maritimus, but significantly larger than that in C. brunneus.
Other results were obtained for hybrids between C. biguttulus and C. mollis (Helversen and Helversen 1975) and C. brunneus and C. biguttulus (Gottsberger and Mayer 2007). A combination of the parental song elements and even novel song ele ments were found in hybrids between C. biguttulus and C. mollis. Thus, the hybrid song may be considered as more complex in comparison with the parental songs. In hybrids between C. albomarginatus and C. oschei (unrelated species to the biguttulus group), the values of several song parameters were significantly larger or smaller than those in the parental songs (Vedenina et al. 2007). Notably, the legmovement patterns appeared to be simpler in hybrids than these in both parentals. In hybrids between C. brunneus and C. biguttulus, the speciesspecific syllable structure was largely lost, because the legmovement patterns were also simplified in comparison to the parental patterns (Gottsberger and Mayer 2007). These divergences in inheritance of different song parameters are likely the result from incompatibility of neuronal networks that control stridulatory leg movements in hybrids. This hypothesis was offered by Helvers en and Helversen (1975). They suggested the two patterngenerating neuronal net works to be formed in the central nervous system of hybrids because of nonhomology of the parental elements. The outputs of the two networks converge in a common final pathway, probably at the level of the motoneurons, and may lead to the superimposed pattern of the hybrid song. In C. brunneus, C. biguttulus, and C. mollis, the song ele ments in terms syllable structure are suggested to be nonhomological. In C. biguttulus, for example, the first and the loudest pulse in each syllable is generated by an accentu ated downstroke of the legs; each syllable is usually produced by three upanddown leg movements; the two legs moving in slightly different patterns (e.g., Elsner 1974;Helversen 1983, 1994). It is no coincidence that some authors attribute these three species to different subgroups of the biguttulus group (Willemse et al. 2009;Sirin et al. 2010). By contrast, syllable of the calling song in C. brunneus, C. jacobsi and C. maritimus ( = brunneus subgroup) is produced by similar leg movements (simple upstroke and stepwise downstroke) and may be considered as homological element. It is therefore not surprising why hybrids between the species within the brunneus sub group generate purely intermediate songs without novel elements or combination of the parental elements.
Considering all the aforesaid, what can we say about the origin of C. miramae? We hypothesize that this species could have evolved as a result of hybridization between other species of the biguttulus group, for example, between C. biguttulus and C. maritimus. The two species are vicariant: the first one occurs in the north, the second one -in the south. For example, in the Ukraine C. biguttulus is found more in the north, where as C. maritimus more in the south. Eastwards, this border is shifting, the ranges overlap, and the species may occur syntopically. In the latter case, however, C. maritimus can be found in the first half of summer, whereas C. biguttulus -in the second half of summer. This indicates that the species tend not to meet, probably because the syllable rate in calling songs is quite similar; the syllable structure, however, is very different. Mean while, we do not exclude that hybridization may occur between these species when one of them is rare and another is abundant. To date, no laboratory hybrids were bred between them, and nothing is known about biguttulus × maritimus hybrid song. The hybridization experiments between these species could be a subject of future studies.
We also hypothesize that C. miramae could diverge from C. maritimus. The latter species is widespread in Anatolia, where it occurs in highlands, thus forming iso lated populations. In Anatolia, there is also another species of the biguttulus group, C. re licticus, occurring very locally in the Southern Anatolian Taurus (Sirin et al. 2010). Its calling song is similar to the brunneuslike echeme of C. miramae, which is produced by simple up and down strokes of the legs moving in antiphase. Sirin et al. (2010) suggest that this species could have radiated from a C. maritimus (named as C. bornhalmi in the paper) like ancestor in an interglacial refugium. In southern territories, the members of the biguttulus group, being the coldresistant species, are suggested to be isolated during interglacial periods and spread down and expanded their ranges during glacial periods. If we suggest the divergence of C. miramae from C. maritimus, the spreading of the former to the north could occur, on the contrary, during interglacial periods.
To test both hypotheses (hybrid origin of C. miramae or its divergence from a C. maritimuslike ancestor in a glacial refugium), it is necessary to conduct genomic studies. A recent analysis of mitochondrial and nuclear genomes in the biguttulus group in Western Europe (Nolen et al. 2020) shows that four species, C. brunneus, C. biguttulus, C. rubratibialis and C. mollis, experienced a long period of geographic isolation, followed by secondary contact and extensive introgression. According to Nolen et al. (2020), C. mollis was the first species to split, C. biguttulus was the next, followed by C. rubratibialis and C. brunneus. Mitochondrial ge nomes suggest that the radiation is relatively recent, dating to the midPleistocene. Thus, the species of the biguttulus group must have experienced multiple episodes of contraction and expansion during the multiple glacial periods that affected the European continent. Taking this into account, it would be especially interesting to sample other species of the biguttulus group, especially those at or near the de scribed refugia in Eurasia.