Aphid-ant Relationships: The Role of Cuticular Hydrocarbons and Different Chemical Stimuli

In ant-aphid interactions, various aphid species offer honeydew to the ant partner and increase their density by ant attendance, whilst others are never attended by ant species, in which case ants tend to treat them as prey. In this regard, ants should have the capability to distinguish myrmecophile aphid species from not mutualistic species, and consequently mutualistic aphids have to be accepted as partners rather than prey. Based on previous studies it is known that ants use the cuticular hydrocarbons (CHCs) pattern of aphids to distinguish mutualistic and non-mutualistic species, but it is unclear, whether the capability to recognize their partners from mutualistic aphid species is innate or depends on learning. Therefore We investigated the cuticular hydrocarbon pro�le of four myrmecophile aphid species ( Aphis pomi , Aphis spiraecola , Dysaphis devecta , Dysaphis plantaginea ) using GC-MS. Then we investigated the behaviour of the ants workers ( Lasius niger L., Hymenoptera: Formicidae) towards the different chemical stimuli deriving from the aphids. We applied four treatments: real Aphis pomi individuals; total extract of the CHC pro�le of Aphis pomi; sucrose solution, and untreated control. The behaviour of the ant workers was different towards the source between the treatments: the real aphid and the sugar solution proved to be the most attractive to the ants, and the presence of pure extract of the individuals is enough to disturb the behaviour of the ants. We provide evidence that the key stimuli of the tending behaviour could be the CHC patterns of the aphids and the CHC pro�le of aphids tend to be genus-specic. This research will promote further investigations to test the behavior of ant workers towards other species of aphids and treatment combinations.


Introduction
The behaviour of ants towards aphids can be of two types; attending behaviour (touching behaviour only) and predatory behaviour(H.Sakata 1994) Ant-aphid interaction is a well-known model of mutualism: ants protect the aphids against potential natural enemies such as; predators and parasitoids while in return aphids provide honeydew to ants as a nutrient.(Addicott 1978;Banks 1962;Breton and Addicott 1992;Bronstein 1998;BRONSTEIN and BARBOSA 2002;Buckley 1987;Carver et al. 2003 Youngsteadt and Devries 2005), prevent potential competitors (Ando and Ohgushi 2008); An aphid's longevity was signi cantly correlated with the daily mean number of workers tending it and increase the tness of the attended aphids through matured earlier, lived longer, a higher expected number of offspring than aphids not tended by ants (Ando and Ohgushi 2008), and had a higher rate of reproduction (Flatt and Weisser 2000).In contrast, (Billick et al. 2014) observed that aphid management by ants is considered to be a less speci c relationship than fungus-farming by ants and relationship among ants and aphids extend from mutualistic to antagonistic, depending on aphid species and environmental conditions.Ant attendance can also has a negative effect on aphid colonies (J.Offenberg, Nielsen, and Damgaard 2019; B Stadler and Dixon 1999;Bernhard Stadler and Dixon 2005).Depending on the ants' available diet options (Joachim Offenberg 2001), mutualism relationship might also turn over into predation.In general, ants favour to species that give large amounts of honeydew, and which produce high quantities of amino acids and/or di-and tri-saccharides in the composition of honeydew (Lang and Menzel 2011;Wolfgang Völkl et al. 1999).Attending and guarding large number of aphids is costly for individual ants (Endo and Itino 2012;Bernhard Stadler and Dixon 2005), therefore it is adaptable for them to prey on the extra aphids as a protein source instead of attending them to assemble honeydew.
Differences in ant attendance can be a consequence of either the different amino acid composition of the honeydew (Lanza and Krauss 1984), or of the differences in frequency and volume of honeydew production(Fischer, M. K., & Shingleton 2001; Wolfgang Völkl et al. 1999) provided by the aphid partner.
Seasonal changes can affected on the concentration of sugars and amino acids of honeydew of aphids, consequently can effected on the behavior of red wood ants (Formica rufa) (N.E.L. Madsen and Offenberg 2020).(Fischer et al. 2002) found that amino acids (especially asparagine and glutamine) are less important for ant resolutions than the honeydew quantity.(Glinwood, Willekens, and Pettersson 2003; H. Sakata 1994) noted that the behaviour of Lasius niger L. (Hymenoptera: Formicidae) might be changed from attending to predation when ants meet a rising proportion of aphids which have not been attended by the ants' colony before.There is a hypothesis that aphids perhaps favour the rapid growing plants, as their sap carries higher quantities of nitrogen and sugars.This can affect the quality and quantity of the honeydew-rewards excreted by the aphids to the ants, and therefore it can increase the degree of protection (number of ants) ( The major sugar components existing in the honeydew of Aphis fabae are sucrose and fructose, showing the highest concentrations (14.3 and 8.1 g L−1) as well as the highest molarities (approximately 0.04 m) among all the speci ed sugars (Detrain et al. 2010).In a maze test, time spent by ants in the foraging area after drinking the sugar solution (glucose, melezitose, ra nose and sucrose) was signi cantly lower, than after drinking pure water (Detrain et al. 2010).Outcomes of (Natalia E L Madsen, Sørensen, and Offenberg 2017) revealed that ants favoured disaccharides over monosaccharides, however the presence of trisaccharides also increased the attractiveness of the sucrose solutions.(Endo and Itino 2012; H. Sakata 1994) revealed that ants are able to mark the aphids they attend somehow to facilitate the decision to prey on or not.(H.Sakata 1994) observed that the ants were less aggressive towards previously marked aphids which provided honeydew earlier, compared to the aphids which have not attended.
Ants stopped attacking when the aphids provide honeydew.Recognition of potential partners mostly depends on chemical cues between connected living organisms, particularly insects that have developed extremely advanced ways to produce and reveal semiochemicals for interspeci c and intraspeci c interactions (Brezolin et al. 2018).(Lang and Menzel 2011) suggested that n -alkanes are the rst nominees as recognition signals, because they are the main ingredient in aphid CHCs, and their comparative quantity varies between mutualistic and non-mutualistic aphid species.In contrast, hydrocarbons with more complex constructions are more simply detectable for the ants when they are connected with sugar solution (van Wilgenburg et al. 2011).However, methylalkanes might also be the key stimuli.Nevertheless, there is no de nite proof, as yet, to demonstrate which structural category of aphid CHCs might operate as partner identi cation signals for the ant partners(I.Sakata, Hayashi, and Nakamuta 2017).In addition, aphid species are classi ed into two groups based on their relationship with ants: myrmecophilous and nonmyrmecophilous species.Myrmecophilous aphid species might use a shared cuticular hydrocarbon signal with the partner ants, which can be used by L. niger to classify aphid individuals into probable trophobiont or probable prey groups.(vanWilgenburg et al. 2011) suggested also, that L. niger uses the cuticular hydrocarbon patterns of aphids to distinguish between different species.Cuticular hydrocarbons of aphids in general are formed mainly of n-alkanes, with fewer amounts of branched alkanes and n-alkenes.Therefore, it is suggested that ants might detect aphid colonies based on (E)-b-farnesene (aphid alarm pheromone), but thereafter they use their CHC pattern to recognise the trophobiotic species (van Wilgenburg et al. 2011).
The aim of this work was to identify the cuticular hydrocarbon patterns of the four aphid species which are the most famous species in Central Europe that causes damages to apple orchards (Aphis pomi, Aphis spiraecola, Dysaphis plantaginea, Dysaphis devecta) feeding on apple; and to examine the behaviour of L. niger against different chemical stimuli, such as; aphid, cuticular hydrocarbon extracted from aphids, sugar solution, and untreated paper control) to reveal the factors playing role in the development and maintenance of the tending behaviour of the ants.There are three circumstances in which ants choose to attend aphids or to assault them: (1) ants prey less on aphid that they attended previously, even though they are myrmecophilous aphids; (2) ants are less probable to assault individual aphids which their nestmates have attended, even within the same aphid species; and nally (3) ants are less likely to assault the aphids that provide honeydew to their nestmates.

Plant Material
Apple trees (Malus domestica 'Idared') were grown from two years old seedlings in plastic pots under a climate-controlled room (24 ± 2°C, 16 L: 8D) and were irrigated regularly, three times a week (Fig. 1 in SI)

Study Organisms
Ant Lasius niger queens were collected from the Botanical Garden of Hungarian University of Agriculture and Life Sciences (volt SZIE) in Budapest (longitude N 47.480867 and latitude E 19.038809).The queen ant was reared under a climate-controlled room (24 ± 2°C, 16 L: 8D).The same ant colony was used for all experiment: it had approximately 50 workers and the same number of offspring.The colonies were reared on 30 m/V% sucrose solution and larvae of Bombyx mori (as a protein source).Ant colonies were kept in glass tubes (16 mm diameter, 150 mm long) in a dark carton case (240 × 175 mm, 100 mm high).
Aphid individuals for bringing up clean, monoclonal strains from the four species (A.pomi, A. spiraecola, D. devecta and D. plantaginea) were collected from the Research Farm of Szent István University in Soroksár (longitude N 47.398575 and latitude E 19.149264).Aphids were reared on apple trees Malus domestica 'Idared' (pot dimensions; 25 mm diam., 20 mm deep).Aphids were kept under a climatecontrolled room (24 ± 2°C, 16 L: 8D).Predators and parasitiods might be present in the laboratory, so apple trees were enclosed by sticky barriers and insect exclusion nets.
Detection of the chemical composition of the cuticular extract of the aphid species by GC-MS 25 aphid individuals (different species and developmental stages, see Table 1) were collected in clean GC-MS vials with ne and clean painting brush, and killed keeping them in a freezer on -20°C for ve minutes.We added markings to the vials to identify the samples easily after the extraction.For the total CHC extraction we added 200 µL n-hexane (analytically pure) to the samples by micropipette.The micropipette tips were changed after every sample to prevent the contamination from each other.As next step, we were shaking the vials for 1 minute with a high-frequency mini shaker (IKA vortex).The extract (200 µL) was transferred by a micropipette with clean micropipette tips separately to a GC vials equipped with micro vial insert.The chemical analysis of the samples was conducted by GC-MS using an Agilent 6890N GC with 5973i MSD device.The GC-MS process had the following parameters: Oven temperature: 190°C for 0.75 minutes (start temperature) then heating up with 8°C/minutes to 320°C, and kept on the nal temperature for 8 minutes.The GC device was equipped with an HP-1MS capillary column (DB-5ms 30 m x 250 µm x 0.25 µm; Cat.No.: J&W 122-5532).The volume of injection was 1 µL.The carrier gas was H2 at a constant ow of 2 mL/min.A split/splitless injector was set to splitless mode on 300°C, with 10:1 split ratio.The solvent delay was three minutes.In the SIM acquisition mode the selected ions were 55; 57; 71; 85.For the data evaluation we used Agilent MSD ChemStation E.02 software.The samples in the order of injection shown in ESI Table .1.

Behavioural tests
Ant responses were recorded to different chemical stimuli in maze tests.A goal area of the tube was connected to the ant colony with a plastic tube (6 A goal area of the tube was connected to the ant colony with a plastic tube (6 cm long, 5 mm in dia.).Cotton bud was used to avoid the escaping of the ant from the tube.The goal arena was a one-centimeter-long, 6 mm in diameter plastic tube closed with a small cotton wool plug.We carried out the treatments (chemical stimuli) on a 3 mm x 3 mm cleaned lter paper pieces on the base of the goal area cm long, 5 mm in dia.), considering the base of the goal arena was clean after every observation to avoid contamination.The duration of the observation was 120 minutes.We placed the maze to a closed room, and xed a high speed camera over it.The sequence of the treatments was randomised.We started the different treatments at the same time on following days, consequently the ant colony has 22 hours rest period between the observations.Totally we performed 3 replicas from all treatments.The evaluation of the videotapes was made manually.We distinguished three different types of the behaviour of the ants (Fig. 2, 3 and 4 in SI).
The treatments in the maze were the following (1) Alive A. pomi aphids xed to the arena (3 individuals); (2) Extract of the full CHC pattern of 10 A. pomi individuals (the solvent was n-hexane) adsorbed on lter paper; (3) 30m/V% sugar solution (imitating the aphid-extracted honeydew) adsorbed on lter paper; (4) Untreated control (clean lter paper).
The Following protocol to get the full cuticular hydrocarbons (CHCs) of Aphids First we washed the painting brush with clean n-hexane then put 20 adults of A. pomi into a 1.5 ml clean Eppendorf tube then were killed through storing them in a freezer (-20°C, for ve minutes).We added 800 µl water to the tube, and shook it for one minute (to remove honeydew remains) then we replaced the aphids to a new clean Eppendorf tube and threw away the previous one, and then added 100 µl n-hexane to the tube which contained the cleared aphids, and shook the tube for one minute then removed the aphids (and threw away).Finally, we adsorbed the extracted CHC components to the clean lter paper quadrants.

Statistical analysis
We xed our observations to an excel database second by second.We distinguished the beginning and the ending of every kind of behavioural state as well as the duration.For deeper analysis, we divided the total duration (120 minutes) into four 30-minute-long quarters to detect the potential behavioural differences within the longer (120 minutes) period.For the statistical analysis of the data, we used robust mixed two-way ANOVA: Welch's test with Games-Howell pairwise comparison to compare the effect of the different treatments (touching, around, away), and Geisser-Greenhouse test with Games-Howell pairwise comparison to compare the effect of the different quadrants within the observation period and to analyse the interaction of the two explanatory variables.

Chemical Analysis
The GC-MS was calibrated based on the different concentration standard mixture solutions, and the retention times of the four standards (nC25; nC27; nC29; nC31) (Fig. 1.).The GC-MS detectable hydrocarbon components (comprising 16 peaks) of the cuticule extract of the four aphid species, were initially distinguished based on their mass spectra and retention index values.(Fig. 2, 3, 4, 5; Table 1, 2).
peak with close retention time to the standard, which was chemically not identical with the standard (Table 2.).In the case of A. pomi we measured three replicates from the adult stage to estimate the standard division of the GC-MS device (Table1, 2).
Three of the four standards (pentacosane (n-C25), heptacosane (n-C27), and nonacosane (n-C29), hentriacontane (n-C31)) were detectable from the cuticle of the four aphid species (A.pomi, A. spiraceloa, D. devacta, D. plantingeinea), and from the different developmental stages of A.pomi including 1-2 nymphs, 3-4.nymphs, winged nymphs, non-winged adults, and the skin), whilst in the case of n-C31 we found a detectable Above the four hydrocarbon standards, we could detect 12 more hydrocarbon peaks from the cuticule extract of the four aphid species (Fig. 2, 2, 4, 5; Table 2).Some of the peaks appeared by all of the aphid species (Table 2; A,C,D,E), while some of them seemed to be genera-speci c (Table 3; speci c for Aphis spp.: B,F; speci c for Dysaphis spp.: H,I,J,K).Within the different developmental stages of A. pomi we could detect no qualitative, but quantitative differences (Table 2, 3.).On the other hand, qualitative and quantitative differences appeared between the different genera (Fig. 2, 3, 4, 5; Table 1, 2): the cuticular hydrocarbon pro le of the two species from the genus Aphis (A.pomi and A. spiraecola) was very similar to each other, but signi cantly differed from the pro le of the twoDysaphis species (D. plantaginea and D. devecta), suggesting that the CHC patterns of the aphids are genera-speci c.

Behavioural Assays
Behavioural state 'Touching' The different chemical stimuli treatments (real, alive A. pomi individuals xed in the goal arena, extract of the cuticular hydrocarbon pro le of A. pomi, 30m/V% sucrose solution and untreated control) affected the activity of the Lasius niger ant workers differently.Signi cantly different activity occurred against different chemical stimuli as indicated by the touching period of the ant workers in the different treatments [Table .3,blue; Fig. 6, Welch-corrected df= (3; 4); f= 7854; p = 0.0383*].The activity of ants had the highest level towards dummies treated with sucrose solution and real aphids, nevertheless they did not differ signi cantly from the extract and untreated control.The only signi cant difference appeared between the untreated control and the extract, thus the total CHC extract of A. pomi can in uence the ant behaviour signi cantly.There was no signi cant differences in the ant activity between the different quadrants within the observation period (two hours) [Table .3,green; Fig. 7, Geisser-Greenhouse test df Behavioural state 'Around' In the case of the behavioural state 'Around' (when the ant worker is in the arena, but it is not touching the stimuli) there was no signi cant difference in the ant activity against different chemical stimuli [Table .Behaviour state 'Away' For the behavioural sate "Away" (the ant in not in the arena) no signi cant difference was found in the ant activity against the different chemical stimuli [Table 5, blue;

Discussion
Aphids and ants are of one of the most studied models of mutualistic associations in the animal kingdom.Aphids produce honeydew, the waste product of their sugar rich but amino acid poor diet, plant sap.After probing by ant antennae, aphids excrete honeydew droplets without dumping them to simplify their collection and assimilation by ants.In exchange for this sugar supply, ants not only protect aphids against natural enemies, but also supply them with hygiene by cleaning the colony from aphid exuviae and honeydew, and decrease the hazards of the fungal contagion (Buckley 1987 ).In the nestmate recognition in ants for instance, some studies have revealed that methyl-alkanes and -alkenes are the main signal components, whereas other studies have showed that n-alkanes play also an important role (Sturgis and Gordon 2012).In such situations where a complex cuticular hydrocarbon pattern is involved, it is not easy to determine which compounds and/or in what quantity are adequate to inform the signal receiver (Liepert and Dettner 1996).In our research, we focused on the n-alkane components of the n-hexane extract of the CHC pro le of the aphids.On the other hand, methyl alkanes-which also appeared in our samples as detectable, but not identi ed components-have been shown to be physiologically effective in several contexts of insect synergy (Howard 1993).We examined the cuticular hydrocarbon pro le of four aphid species feeding on apple (A.pomi, A. spiraecola, D. devecta; D. planatginea) using GC-MS.For the further studies, we attempted to optimize the extraction methods of the CHC-s from the cuticule of the aphids using different extraction periods to gain the optimal extraction period to get the possible highest concentration of the CHC components from the aphids.The optimal extraction period proved to be 30 minutes, and the optimal extraction temperature is 70°C (under raised pressure).The splitless injection process of GC show remarkably greater peak areas and less background noise level, that the split method.Accordingly, the optimal solvent/ aphid individual ratio is 250 µl hexane to 25 aphid individuals.As for detecting method, FID (Flame Ionization Detector) gave more separated peaks and lower background noise level compared to the MS (Mass Spectroscopy).The cuticular hydrocarbon pro le of the two species from the genus Aphis was very similar but differed signi cantly from the pro le of the two Dysaphis species, which also matched to each other.In the CHC pro le of the four species, we found 16 detectable peaks, where two coincided to the normal alkane standards (nC25 and nC27).Our results approved that the CHC pro le of aphids tend to be genus-speci c.On the other hand, simple exposing resemblances among species in their cuticular hydrocarbons are not su cient to prove, that the cuticular hydrocarbons operate as a disguise or they describe species or colony-speci c recognition signals (Allan et al. 2002).It is very probable that Me-11 heptacosene and Me-11 nonacosene compounds promote the ants to prey on Trioxys angeticae.Preliminary bioassays, in which the T. angelicaespeci c amounts of these compounds were applied on dead, hexane-washed parasitoids, and then this dummies were offered in ant-attended A. fabae colonies, found that L. niger responded very aggressively towards such test objects (Liepert and Dettner 1996).Due to the importance of the different compounds, although we were able to detect three (pentacosane nC25, heptacosane nC27, nonacosane nC29) of the examined four standards in the CHC pro le of the four examined aphid species, more investigations are necessary to identify the other 12 detectable chemical components and their biological roles.For hentriacontane (nC31), we could detect a peak with close retention time to the standard, but it has a chemically different structure.Cuticular hydrocarbons might be gained by a myrmecophile aphid easily as a by-product of living in an ant colony and might not transmit any signal to the ant therefore, identi cation bioassays are necessary to prove that chemical mimicry or camou age happens when the cuticular hydrocarbons of a myrmecophile aphid are similar to its host's CHCs (Allan et al. 2002) partner: the honeydew under ant attendance composed of a noticeable lower ratio of glucose and noticeable higher ratios of trehalose and sucrose in a comparison with sugar composition under the absence of ants (Yao and Akimoto 2001).Sucrose in honeydew is considered as an undigested part of the over owing sugar sucked from the host plant (Srivastava 1987).In this case a raised ratio of sucrose in honeydew means that the effectiveness of its hydrolysis in the alimentary canal is reduced under ant attendance.In this regard, we examined the behaviour of workers of L. niger ants towards different chemicals connected to aphids: real A. pomi individuals; total extract of the CHC pro le of the aphids; 30 m/V% sucrose solution and untreated control.Our results revealed that L. niger ants show the highest preference towards sucrose solution.The reason of the high ant activity can be based on the biological role of the sugars: ants use sugar solutions as a primary food source (Beattie 1985;Joachim Offenberg 2001).The second highest ant activity was detectable on real aphids, then on the total extract of the CHC pattern of A. pomi, and just a very weak action was detectable toward the untreated control.Additionally, the total duration of observation by video camera was two hours, and we found that the behaviour of the ants did not differ signi cantly between the quarters (30 minutes long periods).

Conclusions
we can state that the sucrose solution and real aphid had remarkably higher effect on the ant behaviour than the other two treatments.Further investigations are needed to clarify if high ant activities in these two different treatments (sugar solution and real aphid) are similar or different behaviours (just feeding on the sugar solution, and tending/protecting behaviour on real aphids).The CHC extract of the A. pomi had signi cant effect to the behaviour of the ants (compared to the untreated control), but it did not reach the level of the real aphids and the sugar solution.Our results suggest that the ant-aphid mutualisms are not based exclusively on cuticular hydrocarbons, we need further investigations to clarify the role and importance of the different factors (CHC pattern and its concentration, honeydew, shape of the aphid, behaviour of the aphid) as determinants of ant attendance of aphids.
Tables Table 1: The amount (ng/individual) of the cuticular hydrocarbon standards (nC25, nC27, nC29, nC31) detected from the cuticular extracts of the four aphid species and the different developmental stages of A. pomi (nd: not detectable, red colour: a peak with close retention time). The The different components of the cuticle hydrocarbon pro le of the adult stage of A. pomi The different components of the cuticle hydrocarbon pro le of the adult stage of A. spiraecola.
The different components of the cuticle hydrocarbon pro le of the adult stage of D. devecta.The period of behavioural state 'Touching' in the different quarters.
The period of behavioural state 'Around' toward the different chemical stimuli.
The period of behavioural state 'Around' in the different quarters.
The period of behavioural state 'Away' toward the different chemical stimuli.
The period of behavioural state 'Away' in the different quarters.

Fig. 10 ,
Welch-corrected df = (3; 4.4); F=0.681; p = 0.6050].However, the ant activity was signi cantly different during the observation period (two hours): in the quarters three and four ants spent signi cantly longer time away from the goal arena compared to the rst quarter with the second quarter between the rst and third periods [Table.5, green; Fig. 11, Geisser-Greenhouse test df (3.0; 24.0) = 4682 p = 0.0103].
GC-MS chromatograms (total ion counts) of the four standards (nC25, nC27, nC29, nC31).The arrows below the X axes indicate the retention times of odd-numbered n-alkanes for comparison.

Figure 5 The
Figure 5

Figure 7
Figure 7 ; Detrain et al. 2010; El-Ziady 1960; EL-ZIADY and Kennedy 1956; Van Emden 2007; Hölldobler and Wilson 1990; Pontin 1959; Samways 1983; Way 1963; Yao, Shibao, and Akimoto 2000).Cuticular hydrocarbons have a substantial function in the chemical communication of social insects such as ants (Blomquist and Bagnères 2010; I. Sakata, Hayashi, and Nakamuta 2017 (Ness and Bronstein 2004)et al. 1992;Feener 1995;Garnett, Akre, and Sehlke 1985;W Völkl and Mackauer 1993;Wolfgang Völkl 1995;Weiss og et al. 1995)ted the morphological or behavioural characterisation of the ants(Allan et al. 2002;Breed et al. 1992;Feener 1995;Garnett, Akre, and Sehlke 1985;W Völkl and Mackauer 1993;Wolfgang Völkl 1995;Weiss og et al. 1995).Therefore, we investigated the behavioral chracterisation of the ants towards different chemical stimuli.Quantitative differences in cuticular hydrocarbons are supposed to have a signi cant role in nestmate recognition processes in ants.Examinations of the functional subcaste distinction in the ant species Camponotus vagus, assumed that the ants are able to differentiate foragers and brood-tenders by means of quantitative differences in their hydrocarbon pro les (Bonavita-Cougourdan, Clement, and Lange 1993).Nevertheless, it is still not clear whether such quantitative variations between models and mimics are important in chemical mimicry systems.For instance, (Howard, Stanley-Samuelson, and Akre 1990) stated variations in relative abundances of cuticular hydrocarbons in the mimicry system of the syrphid Microdon albicomatus and its host ant Myrmica incomplete.Therefore, we began an optimisation process of the CHC extraction and detection methods to gain more peaks on GC-MS which in turn can lead to determination of further components that might be still undetected.As a result of this study, we detected 16 CHC components in different amounts from the four aphid species with different retention time.As we showed that the CHC pattern of the four species is genera-speci c, and the tested two aphis genera (Aphis and Dysaphis) have a different level of ant attendance (Csaba Nagy, personal observation), we concluded that ant attendance might be partly driven by the quality and quantity of the CHC components of the different aphid species or genera.(ParisandEspadaler2009)assumedthat aphid abundance and honeydew production would rise when the aphids are tended by L. neglectus ants, as this effect appeared by other invasive ant species as well(Ness and Bronstein 2004).In this investigation the density of ant-attendance to the honeydew by the invasive ant L. neglectus was remarkably higher than that of the native ant L. grandis.Aphid honeydew carries a mixture of saccharides: monosaccharides (glucose, fructose), disaccharides (sucrose, trehalose, maltose), and (Auclair 1963s (melezi-tose, ra nose, fructomaltose)(Auclair 1963; Hendrix, Wei, and Leggett 1992; Mittler 1958; Němec and Starý 1990; Wolfgang Völkl et al. 1999; Walters and Mullin 1988; Yao and Akimoto 2001), as well as several kinds of amino acids (Barlow and Randolph 1978; Maltais and Auclair 1962; Mittler 1953, 1958; Sasaki et al. 1990; Yao and Akimoto 2001).attending ants (Formica yessensis) completely changed the sugar composition of the honeydew of the aphid (Tuberculatus querciuola)

Table 2 :
The amount (ng/individual) of the unidenti ed cuticular hydrocarbon peaks detected from the cuticular extract of the four aphid species and the different developmental stages of A. pomi (nd: not detectable, red colour: the retention time of the component).

Table 5 :
Mean time period (standard deviation) of the behavioural state 'Away' toward different chemical stimuli in the four time quarters.