Biotope Impact on Fluctuating Asymmetry Manifestation in Ground beetles (Coleoptera, Carabidae)

In our study we used the data set on morphimetric traits in beetles species. It has been constantly replenishing for 20 years by the samples, received from different regions of Russia and abroad. In this case we have selected data on nine species for which the left and right sides had been measured and fluctuating asymmetry (FA) could be estimated. The samples were from 6 provinces of Russia and Belarus, which ranged in 3 degrees in latitude and 57 degrees in longitude and included more than 150 plots in different types of biotopes. FA was assessed according to the standard method in 5265 specimen in one dimensional trait and one - meristic. ANOVA showed that biotope, species and their interaction affected FA in both traits, that is different species reacted differently to biotope type. In uncommon biotopes (according to accepted in carabidology classification) FA was increased. In forest species the negative relationship between FA in dimensional and meristic traits in the range in biotopes was revealed. In those species only FA values were higher in males than in females. In generalist species FA varieв similarly in both sexes and in both traits being the highest in open biotopes. In eudomonant of arable lands biotopes – Poecilus cupreus – the highest values of FA were recorded in the meadows, being about equal in all types of crops.


Introduction
The body size of an animal is one of the most widely studied traits in biology because it has a potent impact on various aspects of the life-history of a given organism, from physiology to ecology and evolution 1. In a long-term macroevolutionary context, body size may have serious effects on speciation and extinction processes mediated by its effects on population density, resource exploitation, generation time, etc. 2-3. On the short-term ecological and microevolutionary timescale, body size is a crucial feature affecting individual fitness.
That is why body size variation is frequently used in the studies concerning the level of adaptation of certain species. Apparently it is seen in the researches when fluctuating asymmetry (FA) is taken into account. The latter is the admitted indicator of development stress in the certain environmental conditions. In this aspect the most valuable studies were done in insects with holometabolism, because their imagoes do not change body size and its proportions during the remaining ontogenesis 4-5. Ground beetles are no exception. Being the recognized bioindicators 6-9, they are widely used in bioindicative practice. Effects of urbanization and sampling region on FA manifestation were revealed, besides FA in carabids turned to be species and sex biased 10.
The aim of this study was to research biotope peculiarities impact on FA in ground beetles, emphasizing biotope vegetation. Ground beetles are characterized by biotope confinement and in unaccustomed ones they decrease in size 11-12, change their population characteristics 13-25.
We hypothesized that: (i) factors "biotope and species" interact when affecting FA; (ii) FA level is biotope biased; (iii) in uncommon biotopes FA level is increased; (iv) FA in males is higher than in females.

Materials and Methods
Study organisms. We analyzed FA in nine carabid species: Carabus aeruginosus (Fischer von Collection sites and sampling methods. In our investigations we used our data set of morphometric measurements in ground beetles. It was funded in 1996 and has been complementing to date by new information: in the scientific contracts frames our colleagues -carabidologists have been sending the beetles to our laboratory, where we have been measuring them for several morphological traits. At present this data set includes more than 60000 rows and this data is used for different purposes. For the purposes of this study we have extracted the data, including the left and right sides measurements of beetles. So we formed the data set of our further analysis including data in seven species of ground beetles (Table 1). They were sampled in different years and in different provinces of Russia and Belarus (Fig. 1). But we compiled all data taking into account only two variablesspecies and biotope. The others were ignored, because factor "Province" impact had been studied earlier 10. Sampling year impact can be ignored also because sampling period in each province is no less than three years and the possible fluctuations in FA year by year neutralize each other.  So studied beetles were sampled by the unified method (pitfall traps), all photos were taken by one person using the same method. Morphometric data was collected from images taken by Nikon D5100 camera with custom opaque light disperser and a box with opaque reflective surface.

Results
FA in dimensional trait appeared to be affected by species*biotope, but not biotope*sex interactions ( Table 2, S1). In other words, FA level in the beetles from the same type of biotope was depended from species, but it is not dependent by those beetles sex. This is especially clear forC.
granulatus, C. cancellatus and P. melanarius (Table S1). This conclusion became more pronounced when modeling different variables impact against the background of others (Table 3): biotope type and species affected significantly FA in studied samples. This was again especially clear for C.granulatus, C. cancellatus and all types biotopes except shrubs and swamp (Table S2). . Though all species included to biotope*species interaction and Intercept was significant, the only one case of such interaction was significant (Table S3) and there were no interactions of biotope and sex. So FA in meristic trait was practically independent of biotope type where the carabid species dwelled. That conclusion was confirmed by the next analysis (Table 5, S4): FA in meristic trait was species-specific in all studied species. Similarly we analyzed data set in Poec. cupreus. FA in dimensional trait was significantly affected by biotope, but not biotope*sex interactions, sex by itself did not affect FA also (Tables 6,7, S5,6). In meristic trait, on the contrary, FA was significantly affected by biotope, biotope*sex interaction, but not by the sex by itself (Tables 8,9    The actual values of FA in different species and in different biotopes are presented at the following series of figures. In C. granulatus FA in dimensional trait in females fitted in the row: linden=shrubs=birch<meadow<lawn, that is the highest FA level was in the lawn. In males already FA was roughly speaking similar in all biotopes (Fig. 3). If we exclude some biotopes from the figure due small samples in them (i.e. pine, birch, shrubs, swamp), then we can obtain the following picture of dimensional trait FA variation in different biotopes: linden<meadow<lawn. Noteworthy that for meristic trait in females relation was the opposite: lawn=meadow<linden. So for C. granulatus the opposite trend occurred: FA in dimensional trait was the highest in the forest biotope, and in meristic trait -in the open one.
In C. cancellatus FA in dimensional trait in females fitted in the row: spring wheat<linden<birch<lawn<raspberry (biotope "pine" we counted out because of the small sample). In males the tendencies of FA variation were similar (Fig. 4).
In meristic trait (similar to the previous species-C. granulatus) the opposite trend of FA in dimensional and meristic traits: the highest FA in C. cancellatus dimensional trait was observed in open habitats (lawn), and in meristic trait -in forest one (linden).  In C. arvensis in a single type of biotope FA was significantly higher in males than in females in studied traits (Fig. 6), but in P. niger in the same biotope FA was similar in males and females (Fig. 7). In P. melanarius FA in both traits and in females and males as well fitted in the row: birch<pine<linden<meadow, thus, FA being the highest in open biotope (Fig. 8). The opposite trends in dimensional trait were observed in P. oblongopunctatus, the highest value of FA being in open biotope (lawn) and the lowest -in the forest (pine) (Fig. 9). But in meristic trait FA values were the highest exactly in forest biotopes (pine, linden), that is the opposite trends in relation to two studied traits were revealed. Those trends were similar to observed in C. granulatus and C. cancellatus.
In Ps. rufipes FA in both traits and in females and males as well fitted in the row: raspberry<linden<pine (Fig. 10). By the similarity of the variability of asymmetry for both traits it is similar with P. melanarius.
As for the last species studied -Poec. cupreus -the highest value of FA in it was in meadow. It was true both for females and males by dimensional trait and meristic trait (Fig. 11 -14).

Discussion
In previous paper 10 significant species, anthropogene, sex and locality effects on FA manifestation were shown. In this article similar biotope effect was revealed. Especially important is that that effect interacted with species level. In other words, FA realized differently in different species in the same biotope. This conclusion has the direct relation to bioindication where FA is used. That is to say, for instance, if we estimate FA in the C. cancellatus in the city biotopes (these are lawns and meadows as a rule) and in natural ones (linden, for instance), the high level of FA in the city populations will be determined exactly by biotope peculiarities, but not by factor "the city". And the conclusions about negative environment in the city will be wrong.   We confirmed hypothesized biotope*species interaction when affecting FA. Every species has its own preferences for humidity, lightness, canopy cover etc.
The fact remains unexplained in negative relation on FA variation in dimensional and meristic traits in C. granulatus and C. cancellatus again: in both species FA levels in dimensional and meristic trait changed in opposite in relation to biotope type directions. And exactly in those species FA in males was higher than in females. In other studied species males responded to biotope characters about the same as females. So unlike the previous studies 10 we did not find sex*biotope interaction in effect to FA in most cases. Perhaps it can be explained by species-level peculiarities of C. granulatus and C. cancellatus. Both are large forest zoophagous and very rarely are met in open habitats and arable landscapes. These uncommon habitats affect negatively those species including developmental stability and lead to high FA values in lawns and raspberry (Fig. 3,4). P.
oblongopunctatus is the forest species too. And the discussed negative relation on FA variation in dimensional and meristic traits was revealed in it too, though it is relatively small species.
Unlike discussed above forest species, P. melanarius and Ps. rufipes are ubiquists. They feel equally good both in forest and open biotopes. Perhaps due to their "generalists" characters they did not show differences in male/female response to habitats and multidirectional variation of FA in dimensional and meristic traits (Fig. 8,10). So we confirmed three out of four hypothesized suggestions: factor "biotope" and its interaction with factor "species" affected FA in ground beetles the latter being higher in uncommon biotopes. FA were higher in males than in females in a small number of cases.

Conclusions
The use of FA as the bioindicator of environmental stress is complicated by the fact that Funding: This research received no external funding.