Performance of pea aphid (Acyrthosiphon pisum) clones on host plants and synthetic diets mimicking the same plants phloem amino acid composition

The performances of three clones of pea aphids, with different host affiliations, were evaluated on four host plants species and on four artificial diets. The amino acid compositions of the diets mimicked those of the phloem sap of the respective host plants. The total concentration of amino acids was the same in all the diets. The pea aphid clones performance were significantly affected by amino acid composition of the diets in different ways, implying physiological and/or behavioural differences among coexisting pea aphid clones in response to amino acids in artificial diets. The observed differences in performance on diets between clones were not related to host plant affiliations. Thus, even if the variation in amino acid composition in phloem sap among the host plants affects the pea aphid clones when tested on artificial diets, this variation has no observable effect on pea aphid performance on natural host plants.     

Resistance of different lucerne cultivars to the pea aphid Acyrthosiphon pisum: influence of phloem composition on aphid fecundity

Sugars and amino acids were analysed in the phloem sap (sampling by the exudation method) of four clones of lucerne (Medicago sativa L.) characterized by their resistance to the pea aphid, Acyrthosiphon pisum (Harris) (Homoptera: Aphididae). Differences in the qualitative amino acid balance are shown and a negative correlation between the reproductive rate of the aphid and the sugars/amino acids ratio is underlined. To demonstrate any causal relationship between these two variables, the reproduction of A. pisum on artificial diets copying the sap of the two extreme clones (resistant and susceptible) was measured. The results show that the resistance is not a simple nutritional effect; however, the amino acid balance contributes to the resistance exhibited by some cultivars.

---

Résumé L'analyse (sucres et acides aminés) de la sève de 4 cultivars de luzerne (Medicago sativa) caractérisés pour leur résistance au puceron du pois (Acyrthosiphon pisum (Harris)) (Homoptera: Aphididae), est réalisée sur des exsudats de phloème. La teneur en sucres (exprimée en équivalent saccharose) montre d'importantes différences: rapport de 1 à 3 respectivement entre les clones sensible et résistant. La teneur globale en acides aminés exsudés varie en sens inverse, entraînant une corrélation négative pour les 4 clones étudiés entre le rapport sucre/acides aminés (0,6 à 3,6) et la résistance mesurée par le taux net de reproduction à 14 jours (78,5 à 11,2 larves/femelle). L'analyse des acidses aminés révèle de plus des différences qualitatives entre cultivars, notamment pour certains acides aminés essentiels méthionine, histidine, lysine, aromatiques.Le rôle de l'équilibre en acides aminés et en sacharose dans la résistance de certains cultivars au puceron du pois est apprécié par des mesures de la reproduction de A. pisum sur des milieux artificiels reproduisant le rapport sucre/acides aminés et/ou le spectre des acides aminés des 2 clones extrêmes (sensible et résistant). La résistance n'apparaît pas être le résultat d'une simple différence de la qualité nutritionnelle de la sève; la balance en acides aminés doit cependant participer pour une part à la résistance au puceron.

     

Melon resistance to the aphid Aphis gossypii: behavioural analysis and chemical correlations with nitrogenous compounds

In the melon, the Vat (monogenic, dominant) resistance gene governs both an antixenotic reaction to the melon aphid Aphis gossypii Glover (Homoptera, Aphididae) and a resistance to non-persistent virus transmission, restricted to this vector species. We investigated the behavioural features and tissue localisation of the antixenosis resistance by the electrical penetration graph technique (EPG, DC system). We also compared the chemical composition in amino compounds and proteins of the phloem sap collected from two isogenic lines of melon (Cucumis melo L.), carrying the Vat gene or not. All behavioural and chemical data indicated that this resistance is constitutive. EPG analysis clearly showed that access to phloem, although delayed by alterations in pathway activities, was not impaired in terms of frequency of access or initiation of feeding. The most striking feature was, however, a very reduced duration of ingestion from phloem of resistant plants, making this compartment one of the tissues where the effects of the Vat gene are unambiguously expressed. This was confirmed by clear differential activity of phloem extracts in artificial no-choice bioassays. Chemical analyses have shown that phloem saps from the two isogenic lines were extremely similar in profiles of ninhydrin positive compounds, and contained a low total amount of free amino acids (less than 10 mM). Out of more than 40 distinguishable peaks in the chromatograms (protein and non-protein amino acids, as well as small peptides), only five differentiated the two genotypes. Two of them were increased in the resistant genotype: glutamic acid and a major unknown peak, probably a non-protein amino acid (different from pyrazolyl-alanine, a Cucumis-specific amino acid). The three others were depressed in resistant plants, and included the sulphur amino acid cystine and a peptide peak partly composed of the cysteine-containing peptide glutathione (reduced form). Sap collection also showed that phloem exudation rates, as well as total protein and glutathione levels, were depressed in phloem sap from resistant plants. Such data are all indicative of a modified phloem-sealing physiology, linked to sulfhydryl oxidation processes, in plants carrying the Vat gene. The originality of the mechanism of Vat resistance to aphids is discussed.     

Patterns in aphid honeydew production parallel diurnal shifts in phloem sap composition

Variation in phloem sap composition is important in determining aphid performance and is known to occur at both diurnal timescales and in response to plant age. For field grown potato plants, Solanum tuberosum L. (Solanaceae), we determined diurnal variation in components of phloem sap, measured by ethylene diamine tetra‐acetate exudation, and tested for impacts of plant age. The effects of plant age and diurnal cycles on honeydew production by Macrosiphum euphorbiae (Thomas) and Myzus persicae (Sulzer) (both Hemiptera: Aphididae) were also quantified. Both the ratio of sucrose to amino acids and the composition of amino acids in phloem sap varied significantly with time of day. Dietary essential amino acids contributed a smaller proportion of amino acids in the phloem sap of older plants and during early phases of the diurnal cycle. The only significant effect on aphid honeydew production was of the diurnal cycle for Ma. euphorbiae, although increased honeydew production during the day when compared with the production at night, was consistent across the two species. In contrast with studies carried out at seasonal scales, we found limited evidence for variation in phloem sap composition in response to plant age, consistent with our results for honeydew production. These data highlight the need for improved understanding of how seasonal and diurnal physiology of plants influence performance in phloem sap feeding insects.     

Difficulties in location and acceptance of phloem sap combined with reduced concentration of phloem amino acids explain lowered performance of the aphid Rhopalosiphum padi on nitrogen deficient barley (Hordeum vulgare) seedlings

Effects of nitrogen deficiency in hydroponically grown barley seedlings (Hordeum vulgare L.) on the development and reproduction of the aphid Rhopalosiphum padi (L.) (Hemiptera: Aphididae) were investigated.Plant growth was significantly reduced in seedlings grown without nitrogen. Aphid intrinsic rate of increase (r _m) was also significantly lower on these plants compared with that on plants grown with 8 mol m^–3 nitrogen. Phloem sap was collected from seedling stems by aphid stylectomy and amino acids quantified by HPLC. There was a significant reduction in the concentration of non-essential amino acids as a group, but not of essential amino acids. Electrical penetration graphs (EPG) indicated that aphids reached the phloem more quickly and fed for longer on plants grown with nitrogen. This is the first reported study in which this combination of techniques has been used to understand the interactions of an aphid and plant under different environmental conditions.     

Genotypic variation in induced resistance and induced susceptibility in the peach-Myzus persicae aphid system

In response to herbivore damage or stress, plants may express physiological or morphological changes known as induced responses. We tested whether previous herbivory by the aphid Myzus persicae differentially altered the expression of resistance and susceptibility among five genotypes of peach that differ in their resistance phenotype (avoidance resistance, antibiosis resistance or susceptibility). We measured behavioural and performance parameters of aphid success on plants previously infested by conspecifics as compared to uninfested controls. Significant variation was found both among genotypes and among resistance phenotype, including between genotypes showing a same resistance phenotype. Genotypes with avoidance resistance showed either induced resistance to aphid settling or no response. Genotypes with antibiosis resistance showed induced susceptibility to aphid settling, but the effects of previous herbivory on aphid development were either positive or negative depending on the genotype. In the susceptible genotype, most parameters of aphid settlement and performance, including reproduction, were positively influenced by previous herbivory. Using electronic recording, the aphid probing behaviour was examined to tentatively identify host plant tissues most likely to play a role in induced defenses. Probing behaviour was significantly affected by plant genotype, previous herbivory, and their interaction, indicating complex relations between the two factors. In the genotypes with avoidance resistance, aphids were deterred before they reach the phloem. In the genotypes expressing susceptibility or antibiosis resistance, previous herbivory triggered instead the induction of a phloem‐mediated response, that however diverged depending on the resistance status (facilitation or reduction of phloem sap uptake respectively). Genotypic variation in induction found in the peach‐M. persicae system establishes a useful framework to improve our knowledge of the ecological role of induced plant responses to aphids.     

The role of sinigrin in host plant recognition by aphids during initial plant penetration

Plant penetration behaviour (probing) of the cabbage aphid, Brevicoryne brassicae, and the pea aphid, Acyrthosiphon pisum, was studied on excised leaves of broad beans, Vicia faba, kept in water or in a 1% aqueous solution of sinigrin. Using the DC EPG (Electrical Penetration Graph) technique it was shown that the cabbage aphid on sinigrin-untreated bean leaves showed numerous short probes into epidermis and mesophyll. None of these aphids showed either phloem salivation or ingestion waveforms on untreated leaves. In contrast, on sinigrin-treated bean leaves, 35% of the probing time was spent on phloem sap ingestion (E2) and almost all aphids reached phloem vessels and started feeding. The duration of phloem salivation before phloem ingestion and the mean duration of phloem ingestion periods were similar on a host and a sinigrin-treated non-host plant. However, the total probing time by B. brassicae was 10% longer, the total phloem sap ingestion time was twice as long, and the time to the first phloem phase within a probe was three times shorter on the host plant compared to sinigrin-treated broad beans. Acyrthosiphon pisum also responded to the addition of sinigrin to broad beans, but in this case sinigrin acted as a deterrent. On sinigrin-treated leaves, A. pisum terminated probes before ingestion from phloem vessels, and none of these aphids showed phloem salivation and ingestion on treated leaves. Glucosinolates were detected in the mesophyll cells of the brassicaceous plant, Sinapis alba. Based on this finding and in addition to the foregoing EPG analysis of aphid probing on these plants and broad beans, our hypothesis is that aphids may recognise their host plants as soon as they probe the mesophyll tissue and before they start ingestion from phloem vessels.     

Changes in the amino acid composition and protein modification in phloem sap of rice

Pure phloem sap was collected from insects feeding on rice (Oryza sativa L.) leaves by a laser technique similar to the aphid stylet technique. Rapid circulation of nitrogen in the sieve tubes was demonstrated directly using ¹⁵N as a tracer. Application to the roots of the metabolic inhibitors of amino acids, aminooxyacetate and methioninesulfoximine, changed the amino acid composition in the sieve tubes. Feeding methionine to leaf tips resulted in its bulk transfer into the sieve tubes. In vitro experiments confirmed the existence of protein kinases in the pure rice phloem sap. The phosphorylation status of the sieve tube sap proteins was affected by the light regime. The possibility that changes in chemical composition or protein modification such as phosphorylation in the sieve tubes might affect plant growth are discussed.Analysis of pure phloem sap collected from rice plants by insect laser technique has shown dynamic changes in the chemical composition and the quality of proteins in the sap.     

Analysis of wheat resistance to the cereal aphidSitobion avenae using electrical penetration graphs and flow charts combined with correspondence analysis

The behaviour ofSitobion avenae (F.), was compared on resistant wheat lines ofTriticum monococcum (L.) and a susceptible variety ofTriticum aestivum (L.). Firstly, stylet penetration activities were monitored with the Electrical Penetration Graph (EPG) technique and subsequently analysed using flow charts combined with correspondence analysis. Plant resistance was shown to be associated with repeated penetrations without access to either the xylem or the phloem, and with numerous failures in starting a sustained sap ingestion (as represented by pattern E2). Access to sieve elements of the phloem did not seem to be much affected on resistant plants but it took the aphid three times as long to produce a sap ingestion pattern when maintained on the resistant lineT. monococcum n^o 44 (Tm44) as compared with aphids maintained on susceptible plants. As a result the total time spent in ingesting from sieve elements was reduced by 72% on Tm44. Secondly, direct observations of freely-moving apterous adults were performed. Aphids did not discriminate between resistant and susceptible wheat during the first 30 min of access to test leaves, but only 4 out of 25 aphids were still probing after eight hours on resistant Tm44. The relevance of these results to possible location of the resistance factor(s) are discussed. Although detection of plant resistance before sieve elements are reached can not be rigorously excluded, the factors involved inT. monococcum resistance toS. avenae undoubtedly occur within the phloem vessels.     

Phloem specific aphid resistance in Cucumis melo line AR 5: effects on feeding behaviour and performance of Aphis gossypii

The feeding behaviour, excretion rate, and life history traits of the cotton-melon aphid, Aphis gossypii (Glover) (Homoptera, Aphididae), were measured on a resistant melon, Cucumis melo L., breeding line, AR 5. The site of resistance detection by the aphids was determined using the electrical penetration graph (EPG) technique. EPG recordings showed that resistance is expressed within the host plant, rather than on its surface, because the time to first stylet penetration was not significantly different between AR 5 and the closely related susceptible breeding line, PMR 5. EPG patterns associated with stylet pathway activities of the aphids were not significantly different between the resistant and susceptible lines. Significant behavioural differences were observed only after stylets contacted phloem sieve elements. On AR 5, the duration of salivation after sieve element puncture (waveform E1) was significantly longer, and the number of aphids showing phloem sap ingestion (waveform E2) was significantly reduced. We conclude that the resistance mechanism producing the effects seen in this study acts within the phloem sieve elements. Monitoring of excretion rates on the two genotypes showed that aphid feeding was delayed and greatly reduced on the resistant genotype. Comparisons of aphid life history traits and population development between host plant genotypes showed that the effects of resistance act throughout aphid development and are highly effective at slowing down population increase.     

High variation in host adaptation among clones of the pea aphid,Acyrthosiphon pisum on peas,Pisum sativum

The performance of one clone of the pea aphid,Acyrthosiphon pisum (Harris), was assessed on 37 different cultivars and species ofPisum L. In addition, random samples of 36 pea aphid clones collected on alfalfa and clover were tested on a selection of fivePisum sativum L. cultivars. Aphid performance was evaluated in terms of the mean relative growth rate (MRGR) during the first five days of life or other life history variables. The MRGR of the first-mentioned pea aphid clone differed little between cultivars. No significant differences in MRGR were found between wild and cultivatedPisum species or between modern and oldP. sativum cultivars. There was considerable variation in host adaptation among the 36 pea aphid clones within each sampled field. The pea aphid clones showed no consistent pattern in performance on four of the five pea cultivars i.e. there was a significant pea aphid genotype —pea genotype interaction. On one of the cultivars all clones performed well. Pea aphid clones collected from red clover generally performed relatively poorly on pea cultivars, in contrast to the pea aphid clones collected on alfalfa. There was no difference in performance between the two pea aphid colour forms tested. Possible reasons for the high variation and the observed adaptation patterns are discussed. The fact that all clones were collected in two adjacent fields indicates thatA. pisum shows high local intraspecific variability in terms of host adaptation.     

Biological and chemical characteristics of a genetic resistance of melon to the melon aphid

Conclusion Our behavioural data show that the vat resistance in melon is detected early by the aphid during stylet penetration, and is strongly reinforced during phloem feeding. The repulsive effect of phloem sap has been confirmed in vitro, thus demonstrating the presence of extractable chemical factors discriminating the genotypes. The near isogenicity of the two lines is biochemically confirmed by the high degree of homology of their tested composition in many nitrogenous metabolites. Two small peptides are significantly modified when the vat gene is present and they are presently under investigation. Such extractable phloem factors have already been demonstrated, although not yet identified, in another aphid resistance system (van Helden et al., 1995).Author for correspondence     

How nutritionally imbalanced is phloem sap for aphids?

Aphids harbour intracellular symbionts (Buchnera) that provide their host with amino acids present in low amounts in their diet, phloem sap. To find out the extent to which aphids depend on their symbionts for synthesis of individual essential amino acids, we have evaluated how well phloem sap amino acid composition matches the aphids' needs. Amino acid compositions of the ingested phloem sap were compared to amino acids in aphid body proteins and also to available information about optimal diet composition for other plant feeding insects. Phloem sap data from severed stylets of two aphid species, Rhopalosiphum padi (L.) (Homoptera: Aphididae) feeding on wheat, and Uroleucon sonchi (L.) (Homoptera: Aphididae) feeding on Sonchus oleraceus (L.), together with published information on phloem sap compositions from other plant species were used.Phloem sap has in general only around 20% essential amino acids, with a range from 15–48%. Aphid body proteins and optimal diets for two other plant feeding insects have around 50%. The phloem sap of early flowering S. oleraceus ingested by U. sonchi contained 48%, which seems to be exceptional. Aphids feeding on different plants appear to be very differently dependent on their symbionts for their overall essential amino acid synthesis, due to the large variation in proportion of essential amino acids in phloem sap from different plants.The profile of the essential amino acids in phloem sap from different plant species corresponds rather well to profiles of both aphid body proteins and optimal diets determined for plant feeding insects. However, methionine and leucine in phloem sap are in general low in these comparisons, suggesting a higher dependence on the symbiont for synthesis of these amino acids. Concentrations of several essential amino acids in phloem from different plant species seem to vary together, suggesting that levels of symbiont provisioning of different amino acids are adjusted in parallel.     

Tissue localisation of lettuce resistance to the aphidNasonovia ribisnigri using electrical penetration graphs

Electrical penetration graphs (EPG's) ofNasonovia ribisnigri (Mosley) (Homoptera, Aphididae) on resistant and susceptible lettuce (Lactuca sativa, Compositae) showed a large reduction in the duration of the food uptake pattern (E2) on the resistant line. No differences in EPG's were observed before the phloem was reached. Therefore, resistance is believed to be located in the phloem vessel. Both mechanical blocking of the sieve element after puncturing and a difference in composition of the phloem sap are possible resistance factors. However, a chemical factor seems more likely because of the specificity of the resistance againstN. ribisnigri.     

Quantifying the symbiont contribution to essential amino acids in aphids: the importance of tryptophan for Uroleucon ambrosiae

A complete amino acid budget was constructed using the aphid Uroleucon ambrosiae (Strecker), feeding on a suboptimal host, Tithonia fruticosa. The availability of amino acids was estimated from phloem analyses and phloem intake rates at each stage of development. Requirements for amino acids were estimated from gravimetric studies and from analyses of body amino acids. Because the budget was found to be well balanced, estimates of specific needs and shortfalls of essential amino acids were calculated, thus quantifying the role of symbiotic bacteria in fulfilling needs for these amino acids. The most dramatic shortfall was for tryptophan, consistent with the amplification of relevant genes in the symbiont.     

Electrical penetration graphs of the damson-hop aphid, Phorodon humuli on resistant and susceptible hops (Humulus lupulus)

Electrical penetration graphs (EPG's) were used to locate resistance to Phorodon humuli (Schrank) (Homoptera, Aphididae) in hops (Humulus lupulus, Cannabinaceae). Aphids on those hops showing resistance had a much reduced E2 pattern (uptake of phloem). In addition, many aphids on the resistant plants spent time non-probing within two minutes of withdrawing from the phloem. This was not observed with aphids on susceptible hops. The results suggest that resistance is located in the phloem. The involvement of a mechanical factor such as the blocking of aphid stylets, the presence of antinutritional factors, or simply an inadequate supply of nutrients, are discussed as possible resistance mechanisms.     

Probing behaviour of the green peach aphid Myzus persicae on resistant Prunus genotypes

Electrical penetration graphs of Myzus persicae (Sulzer) (Homoptera: Aphididae) feeding behaviour on four resistant and two susceptible genotypes of peach (Prunus persica L. Batsch) and related species showed that resistance was mainly linked to (i) reduced duration of phloem sap uptake, (ii) reduced percentage of pattern E1 (salivary secretion into sieve elements) followed by pattern E2 (sap ingestion) and (iii) increased number of shifts from E1 to E2 and back. These results suggest the unsuitability of phloem sap, and thus repetitive failures to initiate sustained ingestion. Extensive comparisons of the EPGs also revealed more specific trends. Aphids on the most susceptible cultivar GF305 produced significantly longer potential drops than on other peach genotypes. On the resistant Rubira, aphids generated more penetrations before the first E occurred, indicating the possible presence of a resistance factor before the phloem was reached. The clone P1908 of the wild species Prunus davidiana displayed traits of both susceptibility (less but longer probes) and resistance. In particular, aphids produced more E1, suggesting difficulties in preparing sieve elements before feeding. The aphid probing process could be correlated with aphid settling behaviour and bionomics, as previously reported, and gave evidence for the existence of different mechanisms underlying resistance in the tested genotypes against M. persicae.     

Accumulation,selection and covariation of amino acids in sieve tube sap of tansy (Tanacetum vulgare) and castor bean (Ricinus communis): evidence for the function of a basic amino acid transporter and the absence of a γ‐amino butyric acid transporter

Sieve tube sap was obtained from Tanacetum by aphid stylectomy and from Ricinus after apical bud decapitation. The amino acids in sieve tube sap were analyzed and compared with those from leaves. Arginine and lysine accumulated in the sieve tube sap of Tanacetum more than 10‐fold compared to the leaf extracts and they were, together with asparagine and serine, preferably selected into the sieve tube sap, whereas glycine, methionine/tryptophan and γ‐amino butyric acid were partially or completely excluded. The two basic amino acids also showed a close covariation in sieve tube sap. The acidic amino acids also grouped together, but antagonistic to the other amino acids. The accumulation ratios between sieve tube sap and leaf extracts were smaller in Ricinus than in Tanacetum. Arginine, histidine, lysine and glutamine were enriched and preferentially loaded into the phloem, together with isoleucine and valine. In contrast, glycine and methionine/tryptophan were partially and γ‐amino butyric acid almost completely excluded from sieve tube sap. The covariation analysis grouped arginine together with several neutral amino acids. The acidic amino acids were loaded under competition with neutral amino acids. It is concluded from comparison with the substrate specificities of already characterized plant amino acid transporters, that an AtCAT1‐like transporter functions in phloem loading of basic amino acids, whereas a transporter like AtGAT1 is absent in phloem. Although Tanacetum and Ricinus have different minor vein architecture, their phloem loading specificities for amino acids are relatively similar.     

Seasonal changes in the free amino acids of oat and barley phloem sap in relation to plant growth stage and growth of Rhopalosiphum padi

The concentrations and composition of free amino acids in phloem sap from two cultivars of oats and barley, both susceptible to the aphid Rhopalosiphum padi, were determined by means of high performance liquid chromatography. Sap was collected from excised aphid stylets at three developmental stages (seedlings, tillering plants and plants undergoing stem elongation) from plants given or not given fertiliser and grown outdoors. In connection, the growth of individual R. padi nymphs was estimated at the same phenological stages on plants grown in the greenhouse. The content of free amino acids was consistently higher in seedlings than in plants at the early tillering stage. Only in seedlings did the addition of fertiliser increase amino acid levels. Barley phloem sap contained more free amino acids than that of oats when fertiliser was added and at later developmental stages. Phloem sap of oats and barley showed similar patterns in their composition of free amino acids at the seedling stage, but as the plants grew older the patterns became increasingly different. Plants given fertiliser had higher amounts of dicarboxylic amino acids (glutamic and aspartic acid) than unfertilised plants. The concentrations of γ-amino butyric acid, glycine, histidine, and methionine were very low in all treatments. The relative growth rates of R. padi nymphs were low when amino acid content was low and vice versa. The results are discussed in relation to host plant suitability and plant resistance mechanisms.     

Phloem-sap feeding by animals: problems and solutions

The incidence of phloem sap feeding by animals appears paradoxical. Although phloem sap is nutrient-rich compared with many other plant products and generally lacking in toxins and feeding deterrents, it is consumed as the dominant or sole diet by a very restricted range of animals, exclusively insects of the order Hemiptera. These insects display two sets of adaptations. First, linked to the high ratio of non-essential:essential amino acids in phloem sap, these insects contain symbiotic micro-organisms which provide them with essential amino acids. For example, bacteria of the genus Buchnera contribute up to 90% of the essential amino acids required by the pea aphid Acyrthosiphon pisum feeding on Vicia faba. Second, the insect tolerance of the very high sugar content and osmotic pressure of phloem sap is promoted by their possession in the gut of sucrase-transglucosidase activity, which transforms excess ingested sugar into long-chain oligosaccharides voided via honeydew. Various other animals consume phloem sap by proxy, through feeding on the honeydew of phloem-feeding hemipterans. Honeydew is physiologically less extreme than phloem sap, with a higher essential:non-essential amino acid ratio and lower osmotic pressure. Even so, ant species strongly dependent on honeydew as food may benefit from nutrients derived from their symbiotic bacteria Blochmannia.