The study of morphological variability of invertebrate populations allows to evaluate the resistance of a population system, its capacity to retain constancy in dynamic environmental conditions, and the possibility of change within a species, deviation of individuals from average sizes (Sota et al., 2000; Brygadyrenko & Reshetniak, 2014). Those species of invertebrate animals which have a wide range in different continental climatic zones face a great variety of moisture and temperature conditions, which indirectly affects their trophic resources, causative agents of parasitic and contagious diseases. This finally leads to certain morphological types obtaining advantages within a population. Pathological morphological changes of insect imagines with complete metamorphosis occur under the influence of larval parasites, bacterial, viral and fungal pathogens. They can be evaluated using an index of asymmetry, the values of which reflects the intensity of somatic mutations when a whole complex of environmental factors influence an invertebrate's organism (Brygadyrenko & Slynko, 2015). The absence of disfiguring mutations and of sharp deviations in the body structure of certain individuals in a population need not be an unambiguous indication that such a population is in a normal condition (Daly et al., 1998).

Morphometric study of variability of populations of ground beetles involves the expenditure of a significant amount of time in collecting the beetles, measuring them and analyzing the results, which may explain why very little work has been done in this sphere. While it would be of interest to study the variability of populations of rare species of ground beetle and those species that require protection, it would be equally rewarding to study the characteristics of common species of this family, which have a wide range and high degree of ecological flexibility (Brygadyrenko & Fedorchenko, 2008; Brygadyrenko & Slynko, 2015). One of the most abundant species in the ecosystems of almost the entire territory of Europe is Pterostichus melanarius (Illiger, 1798) (Thiele, 1977; Sharova, 1981; Den Boer, 1985; Kryzhanovskij et al., 1995; Hurka, 1996; Irmler, 2003).

The development cycle of P. melanarius can vary in different types of ecosystems: within populations the duration of certain stages of ontogenesis can change, as can the seasonal linkages of these stages (Hurka, 1975; Matalin, 2007). Under certain conditions this species undergoes two-year summer-autumn recycles, in other microclimatic conditions a one year variant of the cycle with overwintering of older larvae, and in others a two year cycle variant with overwintering imagines (Jørum, 1980; Basedov, 1994; Matalin, 2006, 2007). The high abundance of this species and its significance in destroying agricultural pests is acknowledged in many studies in field and laboratory conditions (Desender et al., 1985; Wallin, 1988; Wallin & Ekbom, 1994; Frampton et al., 1995; Fadl et al., 1996; Symondson, 1997; Thomas et al., 1998; Chapman et al., 1999; Raworth & Choi, 2001; Collins et al., 2002; Fournier & Loreau, 2002; Raworth et al., 2004; Kitaev et al., 2011). The morphological variability of populations of this ground beetle species has until now remained poorly studied. The major focus of research attention has been the wing polymorphism of this species (Lovei, 1996; Matalin, 2003). Lovei (1996) points out that the percentage of P. melanarius individuals with fully functional wings is 2% lower in stable ecosystems (that are not affected by periodic flooding with water, for example in forests), and increases in conditions of unstable hydrologic regimes (for example in the polder ecosystems of the Netherlands, small river and lake islands of Poland).

The feeding of this species in laboratory conditions is decidedly nonselective. Until now it has been considered that P. melanarius consumes practically all species of invertebrate animals, even those over twice their weight (Thiele, 1977; Pollet & Desender, 1985; Currie & Digweed, 1996; Thomas et al., 2008). Invertebrates that contain in their body poorly studied poisonous compounds form an exception to this dietary pattern (Symondson, 1997). A detailed study of the diet of P. melanarius diet in laboratory conditions (Brygadyrenko & Korolev, 2006; Korolev & Brygadyrenko, 2012a, 2012b) allowed us to show that this species has an extremely wide range of diet compared to other species of the Pterostichus genus. Searching for an explanation for such a wide trophic range, we attempted to define the individual preferences of certain individuals (Korolev & Brygadyrenko, 2014). Thus individual trophic preferences were identified for the first time for the ground beetle family. It was shown that trophic preferences did not correlate with any morphologic indicator (Korolev & Brygadyrenko, 2014). For further analysis of the reasons for individual variability of trophic preferences, evaluation of the morphological variability of P. melanarius populations will be necessary. If a clearly defined morphotype in P. melanarius populations is found, further research into differences in their trophic preferences can be conducted.

Therefore we have set the following objectives for this research: (1) to define the general variability of morphometric characteristics in a P. melanarius population, (2) to find characteristics according to which sexual dimorphism will be most acutely manifested and (3) to define factors that contribute the most to morphometric variability of the studied population of P. melanarius.

The specimens were collected in June and July 2013 from a P. melanarius population in an artificial forest plantation in the area surrounding Zapadniy, Dnipropetrovsk, Ukraine (48°27′N, 34°56′E). The beetles were collected using pit fall traps and manually from under logs, stones, plant remains. The studied forest plantation (Robinia pseudoacacia L., Quercus robur L., Ulmus laevis Pall., Acer negundo L.) is subjected to intensive anthropogenic impact: recreational pressure (accumulation of garbage, bonfires, damage from trampling), aerogenic pollution with industrial pollutants (heavy metals, formaline, phenol, sulfur and nitrogen oxides) as a result of the operation of the biggest industrial plants in the region (Pakhomov & Brygadyrenko, 2005).

The collected ground beetles (130 males and 95 females) were put for a day in a freezing chamber and then dried out on paper-cotton wool sheets. The collected specimens were assessed according to 14 metric (Fig. 1) and 10 non-metric characteristics (Table 1). These characteristics are labeled as followed and this forms the key to Table 1 (all body measurements are in millimeters): Lb – length of body, from front edge of clypeus to elytra tips, Ll – length of labrum, Sl – width of labrum, Lcl – length of clypeus, Lc – length of head from front edge of clypeus to cervical suture, Sc – width of head before eyes, Lp – length of prothorax along the central line, Spl – width of prothorax front edge, Sp2 – maximum width of prothorax, Sp3 – width of prothorax back edge, Hm – maximum height (thickness) of body at the level of central part of metathorax, Le – length of elytra, Se – maximum width of elytra, Lpen – length of flight wings.

The following indices were calculated: Lb/Hb – ratio of body length to body height, (Sc+Sp+Se)/3Lb – ratio of average width of head, prothorax and elytra to body length, Lp/Sp2 – ratio of prothorax length to its maximum width, Le/Lp – ratio of elytra length to prothorax length, Se/Sp2 – ratio of elytra width to maximum prothorax width, Sp2/Sp3 – ratio of maximum prothorax width to width of its back edge, Le/Se – ratio of elytra length to their width, Sl/Sc – ratio of labrum width to head width, Spl/Sp3 – ratio of prothorax width on front and back edges, Sc/Sp – ratio head width to prothorax width.

The following non-metrical characteristics were analysed: F1 – shape of labrum front edge (1: straight, 2: slightly concave, 3: strongly concave), Chl1 – number of setae-bearing pores on labrum, Chl2 – extent of indumentation of labrum front angles (1: weak indumentation, 2: strong indumentation), Ch111 – depth of setae-bearing pores on labrum (1: the surface is flat, 2: pores of average depth, 3: setae-bearing pores are very deepened), Fcl – shape of clypeus front edge (1; concave, 2: straight, 3: convex), M – arrangement of mandibles (1: left mandible is on top of the right, 2: right mandible is on top of the left), Ml and Mr – depth of furrow on left and right mandibles respectively (1: slightly manifested furrows, 2: furrows of average depth, 3: deepened furrows), Psl – presence of transverse grooves on prothorax front edge (0: transverse grooves are absent, 1: one and 2: two curved transverse grooves), Ps2 – presence of indentations on front edge of prothorax (0: absent, 1; indentations present). For the assessment of the most important body proportions 10 morphometric indices suggested by Sharova (1981) were used, and in addition ratios suggested by ourselves (Table 1).

The measurements of the beetles were taken using an eyepiece micrometer MBC-10 with accuracy to 0.05 mm. When analyzing the characteristics average and standard deviation, minimal and maximal meanings, asymmetry and excess were determined. For determining the statistical significance of differences in averages single factor dispersion analysis was used. For finding factors that identify the variability of the morphological characteristics of P. melanarius we used PCA analysis (Statistica 8.0 program).

For 13 out of 14 metric characteristics (except the length of flight wings) of P. melanarius there is a strongly significant (P=1·10⁻⁹ and lower) sexual dimorphism (Table 1). Females are larger than males, the greatest difference (Figs. 2-4) is for the labrum (the value for females is on average 8.8% higher than for males), maximum width of elytra (7.4%), width of labrum (7.3%), length of body (6.8%), maximum body height (6.8), width of front edge of prothorax (6.6%), length of clypeus (6.5%), width of head (6.5%), length of elytra (6.1%), length of head (5.9%), maximum width of prothorax (5.8%), width of back edge of prothorax (5.8%), length of prothorax (4.9%), length of flight wings (3.7%). If we take out of consideration the length of the flight wings, then on average for 13 metric characteristics of P. melanarius the size of females is 6.56±0.96% greater than those of males.

Apart from average values the range of variability of the studied metric characteristics also varies (Figs. 2-4): females have a 88.9% greater length of clypeus, 18.8% greater length of prothorax, 11.1% length of labrum and 8.3% greater length of head. At the same time, males have a 33.3% greater variability of length of elytra, 200% width of elytra, 14,3% width of labrum, 2.6% length of body, 5.0% width of front edge of prothorax, 4.8% body height.

For body length a normal distribution of individuals in the population was observed for females (excess and asymmetry are statistically insignificant, Table 1) Males, on the contrary, showed a statistically significant (P<0.001) positive excess. According to Geodakyan (1983), the smaller size of males compared to females (in this case by 6.8%) indicates a gradual decrease in body sizes of the species in the process of evolution. However, in the case of P. melanarius this decrease has stabilised, which indicates an absence of asymmetry in distribution of the body length of the males collected for study. A similar distribution is seen in the head width of P. melanarius: females have a normal distribution, males have a statistically significant (P<0.001) positive excess (a sharply peaked distribution is observed).

The length of prothorax of males and females shows a normal distribution (Table 1, Fig. 3), and within this distribution were observed the least significant sexual differences out of the 13 metric characteristics (only 4.9%), which indicate the microevolutionary stability of this characteristic.

For width of front and back edges, and also maximum width, of the females’ prothorax (Table 1, Fig. 3) a statistically significant positive excess was registered (P<0.001, i.e. distribution of the characteristic “peaked”) and a negative asymmetry (P<0.01, i.e. individuals dominate with greater values of measurements, compared to average). Males have no statistically significant asymmetry according to the three measurements of prothorax width.

Body height is, on one hand, connected with the ability to shelter in narrow crevices under stones, in cracks in the soil etc., and on the other hand with the ability of females to increase reproductive capacity (increase in belly volume not only due to length and width of elytra, but also due to body height). The distribution of P. melanarius males according to body height is normal (i.e. body height is optimal for living in the shelter of litter and soil horizons), among females there was observed a negative asymmetry (P<0.01), connected with presence (among the females collected) of a small amount of individuals that oviposited at the time of collecting beetles (June, July).

It is interesting to note that a normal distribution of individuals was registered in the studied P. melanarius population according to length and width of elytra of females and a sharply peaked distribution (positive excess, P<0.001) for these characteristics among males (Table 1, Fig. 4). The absence of statistically significant asymmetry for length and width of elytra among individuals of both sexes either sex indicates the stability of these parameters.

Length of the flight wings for P. melanarius is characterized by acute polymorphism (Table 1), which reaches 14.3-15.5 mm in long-winged individuals in the population we studied (1 of 95 females and 2 of 130 males). Among short-winged males (3.75±0.37 mm) and females (3.94±0.33 mm) the length of wings differed significantly between the sexes (F=15.17, F_0.05=3.88, P=1.3·10⁻⁴). On average, according to the number of long-winged individuals (1.3%), the studied population can be characterized as a stable lowland P. melanarius population, in which long-winged individuals do not obtain advantage (Lovei, 1996; Matalin, 2003).

For determining the most significant body proportions of the ground beetles we examined 10 indices, which in general showed a much lower sexual dimorphism (0.26±0.86%) compared to linear measurements (6.56±0.96%). In other words, females and males are according to 10 the examined indices, practically isomorphic (with decrease in one of the metric characteristics, another one also decreases, and on the contrary with an increase in one, there is an increase in another). Statistically significant differences (Table 1, Figs. 5-6) between males and females were registered only for ratio of length of elytra to prothorax length (Le/Lp, P=0.04, among females this on average is 1.1% higher), width of elytra to maximum width of prothorax (Se/Sp2, P=0.0013, among females is 1.5% higher compared to males) and length of elytra to their width (Le/Se, P=0.01, among males this is 1.2% higher compared to females). Among females a highly significant (P<0.001) excess was registered in 6, and among males in 5 out of the 10 examined body proportions (Table 1).

The joint variability of the examined metric characteristics of P. melanarius (Figs. 5-6) indicates that there are more complex regularities in the distribution of body proportions then was earlier described by Sharova (1981). Regression analysis showed significant differences for the slope of the curve (coefficient a in the equation y=ax+b) among males and females. From this it can be seen that certain individuals can significantly diverge from the cluster of the main selection, which is more evident among females.

The coefficient of variation (ratio of average quadratic deviation to average quantity, expressed as a percentage) is minimal for proportions and is maximal for metric characteristics (Fig. 7), which indicates isometric differences between P. melanarius males and females. The most constant index in the studied selection was the ratio of average arithmetical mean of head width, prothorax and elytra to body length (males: 2.1%, females: 2.3%).

Out of the 10 analysed non-metric characteristics (shape of front edge of the labrum, number of setae-bearing pores on the labrum, extent of indumentums on corners of the labrum, depth of setae-bearing pores on the labrum, shape of front edge of clypeus, disposition of mandible, depth of furrow on left and right mandibles, transverse groove on front edge of prothorax, and impression near front edge of prothorax) a statistically significant sexual dimorphism (P=0.01) was registered only for shape of front edge of the labrum. Among males front edge of the labrum is mostly slightly concave, among females highly concave (Table 1, Fig. 8).

Among most specimens the left mandible is disposed on top of the right (79% of males and 75% females), though about a quarter of individuals in the studied P. melanarius population have another disposition of the maxillae. The biological meaning of differences in mandible disposition remains uncertain, though we can presume that their proportions are significant for eating gastropods, the cuttlebones of which also have different variants of spatial orientation. This question needs further research.

A combined analysis of metric and non-metric characteristics enables the general factors to be identified that define the morphological peculiarities of P. melanarius individuals. Using PCA analysis (Figs. 9-10, Table 2) we found that 45.2% of the morphologic variability of the studied population is determined by factor 1, which we identified as the sex of the P. melanarius specimens (in Fig. 10 in the positive zone of the factor are mostly males with numbers from 1 to 130, in the negative mostly females with numbers from 131 to 225). The whole complex of metric characteristics (length and width of head, prothorax, elytra and parts of the head, body height) together is determined by factor 1, the individual's sex (Fig. 9). The consistency of variation of these characteristics is upheld also by low coefficients of variations in body proportions (indices) compared to linear measurements (Fig. 7).

The remaining 20 factors contributing to the morphologic variability of P. melanarius individuals (Table 2) ranged from 0.3% to 6.3% of the total dispersion (total: 54.8%). The point of directional turning, where the derivative changes its sign from positive to negative, is situated between factors 10 and 11 (Table 2), thus only factors 1-10 have a practical effect upon the morphology of P. melanarius individuals.

Factor 2 (6.3% of dispersion) differentiates non-metric characteristics, not affecting the metric characteristics. It affects the shape of the clypeus and formation of impressions on the front edge of the prothorax. Factor 3 (6.0% of dispersion) manifests itself in the extent of indumentums at the corners of the labrum and depth of furrows in the mandibles. Factor 4 (5.4% of dispersion) is identified by the formation of transverse grooves on the prothorax. Factor 5 (5.1% of dispersion) influences the disposition of mandibles (right on top of left or left on top of the right), and is also connected with depth of furrows in the mandibles. Factor 6 (4.7% of dispersion) influences the quantity of setae-bearing pores on the labrum and formation of impressions near the edge of the prothorax.

In this way the results of PCA analysis upheld the high consistency in variations in linear body sizes of P. melanarius individuals in the studied population, and also allowed us to numerically evaluate the significance of such non-metric characteristics as shape of clypeus, impressions on front edge of the prothorax, indumentation of front corners of the labrum, depth of furrows on mandibles, disposition of mandibles, number of setae-bearing pores on the labrum and impressions near the front edge of the prothorax.

In the course of our previous research we registered the individual differences in the food range of certain P. melanarius individuals (Korolev & Brygadyrenko, 2014). In our current research we attempted to distinguish morphologically distinct groups of individuals according to a single (Figs. 2-4, 8), several (Figs. 5-6) or all examined characteristics (Figs. 9-10). However, no more or less distinguished morphological groups (morphotypes) were found in the population. Morphological homogeneity (normal distribution of individuals) of population was violated for a number of characteristics by the appearance of asymmetry or excess for certain features/parameters or ratios of metric functions. The use of certain invertebrate species (for example ground beetles with no large-scale migration capacity and which accumulate large amounts of toxins in the conditions of anthropogenic activity) is suitable for incorporation into regional programs of environmental monitoring (Lindqvist & Block, 2001; Pakhomov & Brygadyrenko, 2005).

For most metric characteristics (except the length of flight wings) P. melanarius shows a clear sexual dimorphism (females are larger than males by 6.56±0.96%). The studied population, judging by the proportion of long-winged individuals (1.3%), can be classified as a stable lowland population of P. melanarius. For evaluation of the most significant body proportions 10 indices were analysed that manifested much lower sexual differences (0.26±0.86%) compared to linear measurements (6.56±0.96%). Females and males are practically isomorphic: with decrease of one metric characteristic, another decrease proportionately. A statistically significant excess was registered for 6 body proportions for females; and 5 out of 10 evaluated proportions for males. Regression analysis of metric characteristics shows statistically significant differences of angle of slope of the curve for most metric characteristics of males and females. The most constant in the studied selection was index of ratio of arithmetical mean for width of head, prothorax and elytra to body length (males’ CV: 2.1%, females’ CV: 2.3%). Out of 10 non-metric characteristics statistically significant sexual dimorphism is registered only for shape of the front edge of the labrum. Males mostly have a slightly concave and females have a strongly concave upper edge of the labrum.

With the use of regression analysis it was found that 45.2% of morphological variability is conditioned by sex. The whole complex of metric characteristics (length and width of head, prothorax, elytra and parts of head, height of body) varies simultaneously among individuals of different sex. We did not succeed in finding more or less distinct morphological groups (morphotypes) in the studied P. melanarius population. A normal distribution of individuals in the population was violated for a number of characteristics by the appearance of a statistically significant asymmetry or excess of certain features or ratios of metric characteristics.