Eugaster, Serville, 1838, Serville, 1838

Cassar, Louis-F, Massa, Bruno, Gauci, Adam, Bartolo, Angela G. & Schembri, Sarah, 2024, A review of the genus Eugaster Serville, 1838 (Orthoptera, Tettigoniidae, Hetrodinae): a multifaceted approach, Zootaxa 5506 (4), pp. 501-532 : 509-522

Results

The findings of this multifaceted research are presented in sections and subsequently addressed holistically for integration of results.

Colour forms— The resulting spatial distribution maps of the main colour forms ( Figure 5A and B ) demonstrated three key outcomes. First , a well-defined clustering of colour forms, specific to location, is especially noticeable, with no evidence of any two main colour forms occurring within the same locality. Such finding does not only result from the suite of specimens examined for the present contribution, but also from direct field observations carried out independently by two of the authors [BM and LFC], as also by other colleagues (pers. comm. J. J. Borg / A. Catania / A. Seguna-Cassar, 2023) during their field study visit to Morocco in 2006. Second , the two most pervasive colour forms, namely the overall black form often with some blood red coloration on the pronotum (code: Blk O), and the light-coloured form with chequered markings on the abdomen (code: Br /bm), occurred on either flank of the High Atlas , with the former also occurring on either side of the Middle Atlas , that is, on the Western Meseta (loc. Settat) and the High Plateau (loc. Azrou), and as far east as El Kantara in Algeria. Specimens of the latter (Br /bm), which were examined for the present study, were recorded from the Western Meseta (loc. Sidi Bou-Othmane and north of Marrakech ) and from the Anti-Atlas (loc. north of Tiznit ), as well as from around Jebel Chaambi in Tunisia (4 and 5). On the basis of records pertaining to this study, other main colour forms were noticeably more spatially restricted. For example, the lighter, cream-coloured variant with chequered markings (code: CR/bm) was recorded only on the Western Meseta, the tri-coloured form (paratype of E. fernandezi ) solely at Sidi Ifni on the coastal foothills of the Anti-Atlas, and the uniformly brown form (code: U Brm) from hyper-arid locations of Arbaa ait Ahmed on the Moroccan Anti-Atlas and Tamerza in Tunisia ( Figures 5 A and 5 B ).

Apart from these generally recognised colour forms, additional marked variances in coloration, albeit minor, do occur. For instance, a number of specimens belonging to the overall black form (Blk O) were recorded from Midelt, on the High Plateau, at the end of June 2006. Their coloration was either characteristically black with blood red thoracic spines and posterior pronotal region, and with a vivid ochre background on the entire pronotum, or entirely black. Although both variants of the overall black form seemed abundant at the time of observation, the majority of individuals encountered belonged to the former. Most were observed, in significantly large number, on a stony plain environment or reg-type terrain ( Figure 6 A ), as also along the verge of the thoroughfare between Midelt and Aouli ( Figure 6 B ) and on the road proper, where roadkill incidence was conspicuously high. Several were being consumed by individuals while seeking to cross the asphalted road (pers. comm. J.J. Borg-Cassar, 2023).

The third and perhaps most significant finding, where colour forms are concerned, is the fact that the black colour form ( Table 3 , colour code Blk O) occurred at altitudes of between 700 and 1795 metres above mean sea level, with one exception, that of a single specimen taken at Settat, south of Casablanca, on the Western Meseta [specimen code: EM-Set-9]. If the location data is indeed correct, then this would be the only anomaly for this colour form in terms of altitude. The altitude of other sites within which this colour form was recorded in Morocco always exceeded 1200 m AMSL [Azrou on the Western Meseta—Middle Atlas boundary at> 1200 m; Aoulouz at ≈ 1250 m; two localities in the limits of Taliouine at 1425 m and 1483 m respectively; Ouarzazate at 1450 m; 20 km north of Taznakht at 1510 m; and, Anezal all on the Anti-Atlas at 1550 m; Midelt at 1677 m on the High Plateau; and, Gorges du Todrha on the High Atlas—Anti-Atlas interface at 1795 m]. In Algeria, the Blk O colour form occurred at a marginally lower altitude, namely at ≈ 700 m at Bou Saada and 710 m on Jebel Metlili at El Kantara.

The brown colour form with black markings ( Table 3 , colour code Br/bm; Figure 7 ) was recorded at altitudes that ranged between 210 and 560 metres AMSL [North Tiznit at 210 m on the Anti-Atlas; Sidi Bou-Othmane at 516 m; and, north of Marrakech at between 516 m and 560 m, all three sites on the Western Meseta; and, on Jebel Chaambi, within the Aurès range, in Tunisia at 490 m]. The specimens pertaining to the cream colour form with black markings ( Table 3 , colour code Cr/bm) were recorded near Rabat, on the Western Meseta at an altitude of <95 m AMSL.

The uniformly brown colour form ( Table 3 , colour code U Brn) was recorded at practically the two extreme ends on an East–West axis of the Eugaster distribution range, that is, at Arbaa ait Ahmed on the Anti-Atlas, near the Atlantic coast of Morocco, and Tamerza ( Tunisia), located on the eastern extreme of the Aurès mountain range (which extends from eastern Algeria to western Tunisia), respectively, at altitudes of 360 m and 10 m AMSL.

The trichromatic colour form ( Table 3 , colour code Trkrm), with a relatively unique coloration, was recorded solely from the coastal area at Sidi Ifni on the Anti-Atlas, at an altitude of <15 m AMSL.

Whether or not the various colour forms belong to different species, it is assumed that some parallel exists between colour form variation and surrounding landscape, extant biotope, and/or prevailing environmental conditions. Might the physiological significance of the dominant black coloration of individuals occurring within high altitude regions be a thermoregulatory adaptation, in line with the thermal melanism hypothesis? ( Huey & Kingsolver, 1989; 1993; Clusella-Trullas et al., 2007, 2008). Conversely, might the lighter coloured individuals that occur at lower altitude be adapted to enhanced reflective capability to prevent excessive heating of the body? Such physiological adaptation may be analogous with that of the west African cetoniid, Goliathus goliatus (Drury, 1770) , for which it has been suggested that coloration of elytra (which range from white to dark brown) may be contingent upon the forest type (open or closed canopy) in which the larvae develop, as a result of the amount of light that actually penetrated to the forest floor ( Sibilia et al., 2018; Zverev et al., 2018; Dendi et al., 2021). Black Eugaster forms, as would be the case with Goliathus , would conform with the hypothesis, since these large, dark-coloured exotherms will tend to benefit in cooler climatic conditions of high-altitude locations, since it is envisaged they would absorb more solar radiation than their lighter counterparts. Colour form distribution in relation to altitude is presented in Figure 8 . As indicated above, apart from a single exception of a Blk O individual taken at Settat (altitude: 252 MASL), all other records demonstrate a clear correlation between colour form type and altitude.

Biometric assessment and statistical analysis— Visually, some individual specimens appear to be different from others, not only in terms of coloration, but also in respect of some physical attributes. ANOVA was performed, separately for males and females, to compare the means of all colour form types, as well as to determine if the means of selected biometric parameters [length of pronotum; length from upper rim of pronotum to femoral tip (knee); thorax width with spines; thorax width excluding spines] are significantly different (full data in Table 12 ).

In all cases pertaining to male specimens, the p -value is greater than.05 and there is therefore no statistically significant difference between the means of the parameters considered among the five colour form types ( Table 4).

In the case of the female specimens, the mean of the parameters relating to biometrics of the pronotum [length of pronotum; length from upper rim of pronotum to femoral tip (knee)] were not found to be significantly different between the different colour form types, with p -values again above the.05 threshold, as in the case of the males. However, differences in the means of parameters involving biometrics of the thorax [thorax width with spines; thorax width excluding spines] were significantly different between the colour form types ( Table 5). The spines or thoracic protuberances are further discussed in the next sub-section (Feature extraction and edge detection).

Feature extraction and edge detection— A cursory glance at the thoracic structure of the various colour forms gives the impression of characteristically very diverse outlines; the examples in Figure 9 are a case in point. Yet, when one looks carefully at the number, configuration and sequence of protrusion of each thoracic spine, it soon becomes evident that there is minimal distinction between these, even if some protuberances may be different in terms of shape, with some edges being rounded and often stocky, while others thin and more pointed.

Pearson’s R (Pearson correlation coefficient) tests analysed the relationship between selected variables pertaining to or associated with the thoracic structure (length from upper pronotum tip to the knee of hind femur, pronotum length, thorax width including spines, and thorax width excluding spines) ( Table 6 ). Scatter graphs were generated for each pair of variables. In the case of male specimens, a strong correlation was observed between the length of the pronotum-hind femur tip and thorax width (both including and excluding spines), with significance at the 0.01 level. In the case of female individuals, all parameters exhibited significant correlations at the 0.01 level.

Visualisations indicated that all samples followed a general linear trend, with only one male sample (specimen code: EM-SBO- 6 ♂) being identified as a clear outlier. There was also one outlier among the female cohort, which however was not as pronounced as its male counterpart. The specimen (code: ET-Tmz- 53 ♀), which appears to exhibit some manner of deformity ( Figure 9 J ) in terms of thoracic morphology, can be considered appreciably uncharacteristic. Its inclusion in Figure 9 was purely to illustrate the ‘uniform brown’ (U Brn) colour form. From the demonstrable strong linear relationship (particularly evident for the male specimens) resulting from the Pearson Correlation Coefficient tests, it can therefore be concluded that the size (length and width) of the thoracic structure of each individual insect is generally proportional to its overall size.

Although Eugaster thoracic cuticular armour morphology had often been suggested as a means of characterisation to determine taxa at species or subspecies level by various scientific workers, the statistical analysis carried out, results of which are presented in Tables 6 A and 6 B , and Figures 10 A and 10 B , demonstrate that the entire suite of specimens examined is biometrically analogous. This observation is further validated by results obtained from the edge detection process ( Figures 11 A and 11 B ).

Examination of stridulatory file— Following extraction of the tegmina, high resolution images of the sclerotized forewings and stridulatory files were taken, first in pairs ( Figure 12 top and middle) and then singly ( Figure 12 bottom). The reason for the former was to compare the stridulatory file on the left tegmen with that on the right tegmen within each pair, while the latter allowed a closer and more detailed examination of the pegs (also referred to as serrations or teeth) on individual sound-producing structures. In addition to functioning as stridulatory mechanisms, these also function as proprioceptive organs ( Hustert et al., 1999). Table 7 lists the details, including number of pegs on each stridulatory file and the length of each file (when such structures were wholly visible).

The number of pegs of the stridulatory files of all the specimens examined ( Table 7 ) fell within the respective numerical range reported by Grzeschik (1969) and Heller et al. (2022), that is 20–65 “teeth” on the stridulatory file of the left tegmen. In terms of file length, with the exception of three borderline cases, all specimens examined ( Table 7 ) also fell within the respective file length range reported, that is, of 1.4–4.3 mm (Grzeschik, 1969; Heller et al., 2022). Clearly, such broad numerical ranges may render comparisons somewhat difficult. Notwithstanding, it was deemed pertinent to juxtapose measurements taken for this present study with those established by Grzeschik (1969) and Heller et al. (2022) to ensure a more complete investigative protocol.

A statistical comparison of means (amplitude of variation) was performed, respectively, of the lengths of the left and right stridulatory files, and of the number of pegs that occurred on the files examined, between (i) the two main colour forms (Blk O and Br/bm) and (ii) specimens collected from either side of the High Atlas, in this case, the Western Meseta and the Anti-Atlas.

Length of stridulatory files relative to the two main colour forms: On average, the length of left stridulatory files of Br/bm individuals was found to be longer than those of Blk O individuals, respectively, with values of 2.73 (SD.358) and 2.13 (SD.527). This trend is similarly reflected in the lengths of the right file. Br/bm individuals were found to have an average of 2.03 (SD.234), while Blk O individuals were found to have an average of 1.32 (SD.035).

Levene’s test was used to verify if both samples have equal variances (Levene, 1960). For the left stridulatory file length parameter, the test showed an F value of 1.681 with a significance of.224, while for the right stridulatory file length parameter, the test indicated an F value of 3.272 with a significance of.108. In both cases, the p- value was>.05, indicating no statistically significant differences between the variances of the two samples.

The Independent Samples t -Test was then used to assess if there was statistical significance in the difference between the means of the two main colour forms (Blk O and Br/bm). The t -test for the left stridulatory file length resulted in a value of −2.355, with a 2-tailed p -value of.040 ( Table 8), while for the right file length, the t -test was of −5.077, with a corresponding 2-tailed p -test of less than.001. In both cases, the results indicate a statistically significant difference between stridulatory file lengths of the two main colour forms.

Length of stridulatory files relative to geographical locations: On average, the left stridulatory file lengths of individuals from the Western Meseta were found to be longer than those of individuals from the Anti-Atlas, with respective values of 2.67 (SD.402) and 2.43 (SD.560). This trend is also reflected in the lengths of the right file, where Western Meseta individuals averaged 2.04 (SD.305), while Anti-Atlas individuals were found to have an average of 1.67 (SD.394).

Levene’s Test was also carried out to check if both samples have the same variance. For the left file length parameter, the test shows an F value of.443 and a significance of.521. For the right file length parameter, the test shows an F value of 2.33 with a significance of.165. In both cases, the significance is>.05, indicating no statistically significant difference between the variances of the two samples.

The Independent Samples t -Test was then performed to assess if there was statistical significance in the difference between the means of the Western Meseta and the Anti-Atlas samples. The t -test for the left stridulatory file length produced a result of −.816, with a 2-tailed p -value of.434 ( Table 9). For the right stridulatory file length, the t -test was of −1.595 and the corresponding 2-tailed p -test was.149. In both cases, the results are therefore not significant.

Number of pegs on the stridulatory files relative to the two main colour forms: On average, the pegs on the left file of Br/bm individuals were found to be similar in number to those of Blk O counterparts (≈ 41). The average number of pegs on the right file of Br/bm individuals was 31 while on Blk O specimens 27 (SD, respectively, was 1.574 and 6.506).

As in the case of the file length, Levene’s Test was carried out to investigate whether the two main colour form samples have the same variance. For the number of pegs on the left file, the test showed an F value of.750 with a significance of.407, indicating no statistically significant difference between the variances of the two samples. For the pegs on the right file, the test generated an F value of 5.842 with a significance of.042. In this case, the t -test was not carried out in view of a significance of <.05, indicative of a statistically significant difference between the variances of the two samples.

Subsequently, the Independent Samples t -Test, used to assess if there is any statistical significance in the difference between the means of the number of pegs on the left file in the two main colour forms, resulted in a value of −.394, with a 2-tailed p -value of.702 ( Table 10). There is therefore no significant difference between the two main colour forms, Blk O and Br/bm.

Number of pegs on the stridulatory files relative to geographical locations: Regardless of colour form, the number of pegs on the left file of specimens from both the Western Meseta and Anti-Atlas was, on average, ≈ 41, while the number of pegs on the right stridulatory file was also similar (≈ 30) for both sample groups on either side of the High Atlas.

As in previous cases, Levene’s Test was performed to check whether the two geographical samples have equal variances. For the number of pegs on the left stridulatory file parameter, the test showed an F value of.012 and a significance of.914, while for the number of pegs on the right file parameter, the test generated an F value of 2.44 with a significance of.157. In both cases, the significance is>.05, indicating no significant difference between the variances of the two samples.

The Independent Samples t -Test was again used to assess if there was statistical significance in the difference between the means of the specimens from the Western-Meseta and the Anti-Atlas. The t -test for the number of pegs on the left stridulatory file resulted in −.329, with a 2-tailed p -value of.749 ( Table 11). For the number of pegs on the right file, the t -test produced a result of −.634 with the corresponding 2-tailed p -test of.544. In both these cases, the results are not significant (accepting the null hypothesis), indicating no statistical difference between the two geographical sub-groups in terms of the number of pegs on both stridulatory files.

Genital sclerite structure examination— The two ♂♂ specimens (spec. codes: EM-Alz-35 and EM-Orz-51) that were examined in the conventional manner, using KOH solution to dissolve soft tissue, did not yield any structure that resembled, even remotely, the titillators of a Tettigoniid. The other four ♂♂ specimens, selected from either side of the High Atlas Mountain range (spec. codes: EM-SBO-7, EM-Mrk-10, EM-Anz-12, and EM-Tzt-29), were examined using a commercially available ‘relaxing fluid’, and tepid water in which drops of antiseptic were added to discourage the subsequent formation of mould. As indicated in the Methodology above, abstaining from the use of KOH has the advantage of preserving non-sclerotized tissue; in the event that Eugaster titillators are so constituted, as suggested by Heller et al. (2022), such approach would have assured the preservation of these copulatory appendices. Conversely, however, this also meant that all other material present, including other anatomical structures and undigested organic matter, formed a conglomerated mass, which presented various challenges in this aspect of the investigation, not least in distinguishing the various anatomical characters from one another. Although some structures that bear a strong resemblance to Tettigoniid titillators were observed within the mass of organic material examined belonging to both specimens that were softened with tepid water ( Figure 13 ), it could not be incontrovertibly ascertained that these were indeed the copulatory appendices being searched for. At this juncture, it may be premature to assume the presence of such structures based on the examination of dry specimens, given their condition, and further investigations may be warranted, using freshly collected individuals.

Molecular and Phylogenetic analyses— The quality of dried specimens used varied considerably and, as a result, DNA degradation did not allow the amplification of most sequences for each specimen. Only the mitochondrial cytochrome c oxidase I (COI) marker was amplified successfully for specimens EM-AaA-27 (U Brn colour form) and EM-Anz-22 (Blk O colour form).Although this data is not comparable to the nuclear data presented in Grzywacz et al. (2015), it is however suitable for differentiating species, particularly among insect groups (Zhang and Bu, 2022). Reported herein are the first COI sequences of Eugaster spinulosa . The BLAST searches did not result in any close hits. The consensus tree for the COI data shows that the sequence for the specimens EM-AaA-27 and EMAnz-22 clustered together and was highly supported ( Figure 14 : 1.00 and 100%).

Another noteworthy aspect concerning the methodologies employed for the present DNA analyses is that the specimens used were collected between 1986 and 2022, including two specimens (EM-AaA-27 and EM-Anz-22), collected in 2015 and 2022 respectively. In a recent study, Mullin et al. (2023) noted that most DNA fragments from old museum insect specimens were below 100 bp, even for samples collected as recently as 2004, which shows that DNA fragmentation occurs rapidly after dry preservation of the insect specimens.

In the present study, it was deemed pertinent to include both successful and negative outcomes since the evaluation of results involving genotyping from museum specimens has often been ambiguous, with most studies being limited only to successful results (Lalonde & Marcus, 2020). In order to heighten the chances of success, when employing the methodology outlined herein, it would be advisable to focus on fresh material in future genetic work on this genus.

Limitations in connection with DNA analyses— DNA degrades post-mortem as a function of time and heat (Lindahl, 1991) and, as has been attested, molecular-based studies have mostly been limited to recently collected specimens preserved in a specific manner for such work (Gilbert et al., 2007). Studies that investigated old and degraded DNA have, in general, demonstrated a characteristically high ratio of mitochondrial to nuclear template molecules, with mitochondrial DNA remaining PCR-amplifiable for longer in comparison to nuclear DNA (Lindahl, 1991). In fact, Lalonde and Marcus (2020) reported that successful works on museum specimens have mostly used either nuclear microsatellites or fragments of mitochondrial DNA, with the latter being the more successful approach. However, recent literature has also shown that some efforts were successful in recovering both nuclear and mitochondrial data from historical insect specimens, which suggests that it may be possible to characterise nuclear genetic data from a specimen where mitochondrial methods have been successful (Lalonde and Marcus, 2020). Thus, whilst nuclear DNA generally requires fresh specimen or specimen frozen at -20 °C, mitochondrial DNA is in general more obtainable from older specimen. However, noting that the only previous study with published sequences for the genus Eugaster used nuclear DNA from fresh material, the present authors attempted to amplify both mitochondrial and nuclear DNA.

Figures

FIGURE 5 A [top] and B [bottom]. Spatial distribution of all records pertaining to the present study, for sites in Morocco [A] and sites in Algeria and Tunisia [B], based on specimen data. ●Blk O; ●Br/bm; Cr/bm; ●U Brn; ●Trkrm.

FIGURE 6 A [left] and B [right]. Female specimens in the vicinity of Midelt (High Plateau) in June (Photos: P.M. Sammut).

FIGURE 7 A [left] and B [right]. Female specimens at Sidi Bou-Othmane (Western Meseta) in June (Photos: J.J. Borg).

FIGURE 8. Contour map highlighting altitude—colour form gradient across localities in which specimens were recorded (Base map source: ETOPO).

FIGURE 9. Head and thoracic regions of the various colour forms and variants, which marked the initial phase of the feature extraction and edge detection process. A–D: Blk O; E: Trkrm; F–G: Br/bm; H: Cr/bm; I–J: U Brn.

FIGURE 10 A [top—males (blue data points)] and B [bottom—females (red data points)]. Scatter plots showing the relationship between the length of the pronotum and the overall body length (approach based on firmly sclerotised parts of the body structure, i.e., upper rim of pronotum to the knee of the hind femur—see Methodology for rationale).

FIGURE 11 A and B. Skeletised outlines of selected males [top] and females [bottom] representing the five colour forms from different geographical locations. The general distribution of spines on all three sections of the pronotum (prozona, mesozona, metazona) is characteristically similar across the entire suite of specimens examined. Additional cuticular protuberances occur randomly in many of the representative samples of all five colour forms.

FIGURE 12. Tegmina and stridulatory files of specimens collected from either side of the Moroccan High Atlas. Top and middle: Left and right tegmina of 2♂♂ from 25 km north of Marrakech, Western Meseta, at 560 m (EM-Mrk-10) and North Tiznit, Anti-Atlas, at 210 m (EM-Tzt-32) respectively; note the much longer file on the left tegmen in either case. Bottom: Left tegmen and stridulatory file of ♂ collected from Anezal, Anti-Atlas, at 1550 m (EM-Anz-20).

FIGURE 13. A titillator-like appendage observed during examination of the terminalia of a male from Sidi Bou-Othmane on the Western Meseta (spec. code: EM-SBO-7).

FIGURE 14. Consensus phylogenetic tree of species inferred from COI sequences. Bayesian Inference (BI) and Maximum likelihood (ML) analysis involved 9 specimens (of six species) and one outgroup taxon. The numbers on branches are Bayesian posterior probabilities (BPP) and bootstrap (BS) values (> 0.75 and 75%, respectively). The scale bar represents the number of substitutions per site.

Tables

TABLE 6 A [top—males] and B [bottom—females]. Correlation matrix for variables associated with the thoracic structure of male and female specimens, respectively, summarising the bivariate Pearson correlation values for all the variables evaluated. Correlation is significant at the 0.01 level (2-tailed). Values in bold within shaded cells show statistically significant correlations.

A. Males	Length pronotum— Hind femur tip	Pronotum length	Thorax width + spines	Thorax width excluding spines
Length pronotum—Hind femur tip	1	0.248	0.674	0.673
Pronotum length	0.248	1	0.45	0.354
Thorax width + spines	0.674	0.45	1	0.322
Thorax width (excluding spines)	0.673	0.354	0.322	1
B. Females	Length pronotum— Hind femur tip	Pronotum length	Thorax width + spines	Thorax width excluding spines
Length pronotum—Hind femur tip	1	0.716	0.852	0.607
Pronotum length	0.716	1	0.814	0.746
Thorax width + spines	0.852	0.814	1	0.821
Thorax width (excluding spines)	0.607	0.746	0.821	1

TABLE 7. Details and measurements pertaining to the 12♂♂ specimens whose stridulatory files were examined.

Specimen code	Colour form code	Number of pegs on Left file	length of Left File	Number of pegs on Right file	length of Right File
		n	mm	n	mm
EM-SBO-1	Br/bm	47	3.25	33	2.30
EM-SBO-6	Br/bm	40	2.80	30	2.12
EM-SBO-7	Br/bm	41	2.72	29	1.60
EM-Mrk-10	Br/bm	39	2.30	31	2.15
EM-SBO-11	Br/bm	41	2.28	-	-
EM-Anz-12	Blk O	43	2.48	-	-
EM-Anz-20	Blk O	41	2.64	21	1.36
EM-Tzt-29	Br/bm	39	2.55	32	2.05
EM-Tzt-31	Br/bm	42	3.18	29	1.85
EM-Tzt-32	Br/bm	43	2.76	32	2.15
EM-Alz-35	Blk O	44	1.90	34	1.30

......continued on the next page