Version classiqueVersion mobile

Cassava-Mealybug interactions

 | 
Paul-André Calatayud
, 
Bruno Le Rü

2. Host-Plant selection

Texte intégral

1Phenacoccus manihoti and P. herreni are oligophagous insects which mainly colonize Manihot species. In Africa, limited infestation of Talinum triangularae Jack. (Portulacaceae) by P. manihoti also has been reported (Neuenschwander et al., 1986). In South America, however, P. manihoti has been found on Citrus spp. (Rutaceae) and soybean, Glycine max (L.) Merr. (Fabaceae), while P. herreni has been reported only on M. esculenta (Williams & Granara de Willink, 1992). Under laboratory conditions, P. manihoti can be reared on poinsettia, Euphorbia pulcherrina Wild. (Euphorbiaceae)(Boussienguet, 1984).

2The different resistance components of cassava to P. manihoti have been studied by Tertuliano and colleagues (1993). These authors identified varieties of cassava with different degrees of antixenosis or non-preference to the pest, i.e. cassava varieties more or less preferred by the insect (see also Chapter 5). This type of resistance generally occurs during initial plant selection and involves the physical and chemical characteristics of the plant. To further explain how host plants are identified by the insect, the types of sensory information mealybugs can detect using either the antennae or the labium were first described by Le Rü and colleagues (1995a and b). To determine the role of plant surface characteristics in plant selection, the probing behaviour on the surface and on the outer tissues of the plant by the mealybug was analysed (Renard et al., 1998; Renard, 1999). Studies on the feeding behaviour have helped to determine the final decision made by the insect in plant selection (Calatayud et al., 1994a; 2001a).

SENSORY SYSTEM OF MEALYBUGS

3During plant surface exploration, the major sensory organs used by Sternorrhyncha are the antennal and labial sensilla. While the antennal sensilla of aphids and psyllids have been well studied, the labial sensilla have been poorly investigated, with only three studies on aphids and whiteflies that show a high degree of anatomical variability among species (Backus, 1988; Walker & Gordh, 1989). In the Pseudococcidae, the external morphology of the antennal receptors has been examined by light microscopy only. A study on the citrus mealybug, Planococcus citri Risso (Pseudococcidae) has shown the presence of one and two basiconic sensilla with olfactory function on the subapical and apical segments of the antenna, respectively (Salama, 1971). Koteja (1980) studied the external morphology and the distribution of sensory receptors on the antennae of a few species of Pseudococcidae by light microscopy. In Phenacoccus aceris Geoffroy, he observed the presence of basiconic sensilla and sensilla chaetica (thin-walled pegs) with putative olfactory function on the last antennal segment. More recently, Le Rü and co-workers (1995a, b) investigated the external morphology of both antennal and labial sensilla of P. manihoti by scanning electron microscopy and their internal ultrastructure by transmission electron microscopy.

Antennae of P. manihoti

4As in most of the Sternorrhyncha studied to date (Backus, 1988), the antennae of P. manihoti are equipped with a small number of sensilla (Figure 2.1). Its sensory system appears to be well diversified, however. Trichoid sensilla, distributed on all segments of the antenna and innervated by a single mechanoreceptive dendrite, have the characteristics of exteroceptors. They are located mainly on the distal part of antennal segments and it is thought that they may be involved in the mealybug’s sense of touch. Their numbers are similar to those previously reported in other Sternorrhyncha, such as the Aphididae and Psyllidae (Backus, 1988).

5A campaniform sensillum located on the pedicel and one basiconic sensillum on the flagellum were also detected on the antenna of P. manihoti (Le Rü et al., 1995a). These sensilla, which have the characteristics of proprioceptors (Koteja, 1980), have been found in other mealybug species and also in Aleyrodidae, Aphididae, Psyllidae and Thysanoptera, among other orders (Koteja, 1980). Such sensilla are likely to help the insect in guiding the movement of its antennae (Bromley et al., 1980). As previously described by Koteja (1980) for the antennae of many Coccinae species, P. manihoti antennae also possess coeloconic sensilla located ventrally on the pedicel and flagellum. The author did not attribute any function to those receptors, however, as his study was based on only external cuticular structures. Le Rü and colleagues (1995a) have shown that these sensilla are related to poreless sensilla with inflexible sockets, probably having thermo/hygroreceptor functions.

Figure 2.1
Antenna of 1st instar of Phenacoccus manihoti. Ventral view of the 9th antennal segment (a) showing the position of the different sensilla; the longest sensilla, which have a chemoreceptive function, are indicated by the arrows. Detail of one of these sensilla (b) showing a perforated cuticle with numerous pores (Le Rü et al., 1995a)

6Koteja (1980) reported the presence of several types of sensilla on the three distal segments of the flagellum in Pseudococcidae. This was confirmed for P. manihoti by Le Rü and co-workers (1995a), who reported five different types of sensilla. It was shown that these sensilla are related to uniporous chemosensilla and multiporous chemosensilla with probable gustatory and olfactory functions, respectively. Unlike the other families of Sternorrhyncha (Aphididae, Aleyrodidae and Psyllidae), the antennae of Pseudococcidae are not equipped with plate organs, whose olfactory function has been demonstrated in aphids by Bromley & Anderson (1982). In Pseudococcidae, the olfactory function appears to be mediated by peg sensilla localized on the apex of the antenna. Moreover, similar to Aphididae and Psyllidae, the gustatory function is also localized on the apex of the antenna in Pseudococcidae.

Labium of P. manihoti

7In Sternorrhyncha, the sensory receptors on the labium have been poorly described compared to those on the antennae. However, it has been observed that the tip of the labium of the aphid Brevicoryne brassicae L. possesses only mechanoreceptors (Wensler, 1977; Tjallingii, 1978a), while in the whitefly Parabemisia myricae Kuwana, the labium is equipped with chemoreceptors (Walker & Gordh, 1989). The morphology and ultrastructure of sensilla present on the tip of the labium of Pseudococcidae were described for the first time on P. manihoti by Le Rü et al. (1995b), who reported 10 pair of trichoid hairs with a probable mechanoreceptive function distributed over the labium (Figure 2.2). It has been suggested by Backus (1988) that such hairs may be involved in the identification of the physical characteristics of the plant surface while the insect is selecting the site of the probe, and may likely detect the depth and angle of the labium when the mouthparts or the stylets are inserted during feeding. On the labium of P. manihoti, six uniporous chemosensilla and two multiporous chemosensilla with likely gustatory and olfactory functions, respectively, have been recorded. Le Rü and colleagues (1995a) reported the presence of gustatory sensilla on the labium of Pseudococcidae, while Walker & Gordh (1989) described such sensilla on the labium of Aleyrodidae. The former authors reported for the first time the presence of olfactory sensilla in Sternorrhyncha; the presence of such receptors had been described previously only in Auchenorrhyncha and Nilaparvata lugens Stal (Delphacidae) by Foster et al. (1983).

Figure 2.2
Labium of 4th instar larvae of Phenacoccus manihoti, showing six sensilla having contact chemoreceptor as well as mechanoreceptor functions (stars) and two sensilla having olfactory functions (arrows), according to Le Rü et al. (1995b)

8In conclusion, as in most Sternorrhyncha studied to date, the apex of the antenna and the tip of the labium of P. manihoti are equipped with a small number of sensilla, generally characteristic of a high specificity in feeding (Chapman, 1982). The sensory system of P. manihoti, however, appears well diversified, with 58 sensilla of nine different types on the antenna and 30 sensilla of four different types on the labium. Together, the antennal and labial sensilla most probably mediate orientation to the cassava by crawlers (larvae) of P. manihoti and the initial stages of plant acceptance, while dabbing the plant surface with the ventral side of the last antennal segment and the tip of the labium.

HOST-PLANT RECOGNITION BY MEALYBUGS

Behaviour on plant surface

9Video observations done on P. manihoti (Renard et al., 1998; Renard, 1999) have shown that when selecting the fixation site, the mealybug alternates between walking and stopping (called the ‘probing sensus’ by Klingauf, 1970). When the mealybug walks, the antennae are pointed forward and the labium quickly taps the leaf surface (brief contact). When the insect stops, it rubs the surface with the tip of its forelegs and slowly taps it with the lower side of the last antennal segment and with the tip of the labium (more prolonged contact). The first stop may take place on the upper or lower face of the leaf (Figure 2.3). Prior to probing, simply walking on the leaf surface and tapping the leaf with the antennae and labium allows the insect to differentiate between the more- or less-preferred plants. On a preferred cassava variety such as ‘Incoza’, the passage on the lower leaf face is quickly followed by a nearly immediate fixation. On a less preferred cassava variety such as ‘M’pembe’, an increase in the duration of walks is observed. On a non-host plant such as Talinum triangularae (Portulacaceae), the walks are long (especially on the upper leaf face) and accompanied by numerous changes of side as the insect attempts to escape. This pattern of behaviour occurring on a less preferred plant and on a non-host plant is observed in many Sternorrhyncha (Backus, 1988), and especially in aphids (Ibbotson & Kennedy, 1959; Klinghauf, 1970) and whiteflies (Noldus et al., 1986; Walker, 1987).

Figure 2.3
Characteristic ethograms (a succession of behavioural items observed over a time period) of the Phenacoccus manihoti testing behaviour observed over 90 min on leaves of two host plants (two cassava varieties) and one non-host plant (Talinum triangularae) (after Renard, 1999)

10The test probing behaviour by the insect before feeding, constituting the first step in host-plant acceptance, has been studied in detail for P. manihoti (Renard et al., 1998). To observe the relationship between the movement of mealybug organs (body, legs, antennae and labium) and the stylets’ pathways inside the plant tissues, video observations of the mealybugs combined with electropenetrography (EPG), a technique to electrically follow the stylets’ pathway, were used. Phenacoccus manihoti exhibits three phases of test probing (Renard et al., 1998). During the first phase, the mealybug repeatedly drums and rubs the plant surface with the ventral part of the antennae, labium and foreleg tips. Thereafter, the antennae vibrate. At the end of this phase, the labium remains in contact with the plant surface without any stylet penetration into the plant tissues. According to Städler (1986), the insect probably evaluates the nutritional quality of the plant during this phase, the recognition having already taken place during this behavioural event. The repeated contacts of the olfactory and gustatory sensory organs of the mealybug, located principally on the antennae and labial tip, suggest that, as is hypothesized for aphids (Klingauf, 1971; Greenway et al., 1978), the mealybug is able to perceive the odours present in the thin air layer above the leaf surface (the boundary layer).

11In the second phase, the insect quickly moves its head up and down, moves the antennae forward and backward, and rubs the plant surface with its forelegs. During this phase, the stylets pass through the epidermal and inner tissues of the plant. Finally during the third phase, the mealybug becomes more agitated. It can be observed to stand up using its rear legs and push the upper part of its body against the plant. The stylets continue their progression into the tissues, which is principally extra- or intercellular (see also section on feeding behaviour, below), until they reach the phloem. Renard and colleagues (1998) observed that the three phases described above and the progression of the stylets in the leaf tissues are rendered difficult in less preferred cassava varieties. This last aspect has been confirmed in EPG studies by Calatayud and co-workers (1994a).

Influence of host-plant surface characteristics

12Upon contact with the plant surface, the insect is able to obtain both chemical and physical information on plant quality by tactile (mechanosensory) and contact chemosensory (taste or gustation) stimuli (Schoonhoven et al., 1998).

13Physical characteristics:

14The physical characteristics of the leaf surface may not play an important role in host-plant recognition. Renard (1999) has shown that for P. manihoti, for example, neither the trichome density nor the waxy thin layer of the lower leaf epidermis (the leaf surface colonized by mealybugs) have a relationship to preference for a cassava variety or the status of a plant as a non-host. Host-plant acceptance by P. manihoti therefore appears to be mainly determined by the chemical characteristics of the plant surface.

15Chemical characteristics:

16Mealybugs prefere to colonize Manihot species. Cassava has a high content of cyanide compounds in the leaves, stems and roots (Arihantana & Buckle, 1986; Ezeala & Okoro, 1986; Pancoro & Hughes, 1992). The predominant cyanogen is linamarin (Butler et al., 1965) (Figure 2.4), a cyanogenic glucoside which is hydrolyzed into glucose and free hydrogen cyanide (HCN) after tissue damage (Conn, 1980). This hydrolysis is called ‘ecyanogenesis’.

Figure 2.4
Linamarin structure according to Conn (1980)

17Further experiments done on P. manihoti by Renard (1999) demonstrated that if cassava leaves are previously soaked in methanol, thus preferentially removing the polar compounds (including linamarin) from the leaf surface, the insects do not recognize the host leaves. Without solvent treatment, 80% of the insects were fixed on leaves after infestation versus only 30% after methanol treatment. When these methanolic extracts were applied to the Parafilm enclosing an artificial diet (see Chapter 3 for details of the set-up), an increase in the proportion of insects fixed on the diet rose from 10% in the control to 43% after treatment by the extracts. Analysis of the methanolic extracts by gas chromatography coupled with mass spectrometry (GC-MS) revealed the presence of linamarin at a concentration of 1–4 ng/cm2 (Renard, 1999). To confirm the role of linamarin in host-plant recognition by the insect, commercial linamarin (620 µg/L) was deposited on Parafilm enclosing the artificial diet. The proportion of insects fixed on the diet increased from 33% in the control to 93% after treatment (Renard, 1999).

18These results all indicate that linamarin, present on the phylloplane of cassava, induces the host-plant recognition behaviour of P. manihoti. Even at the very low linamarin concentration of only a few ng/cm2, this compound could be detected by P. manihoti larvae because of the high sensitivity of the sensory organs found generally in insects. Moreover, the repeated contacts of the olfactory and gustatory sensory organs, located principally on the antennae and labial tip of P. manihoti (Le Rü et al., 1995 a, b), enable the insect to perceive compounds such as the linamarin present in the thin air layer of the phylloplane. These repeated contacts of the sensory organs with the plant serve likely to increase the amount of sensory input to the central nervous system.

19Primary compounds such as sucrose and free amino acids also are found on the cassava leaf surface. Nevertheless, there is no evidence that these compounds play a crucial role in host-plant recognition by P. manihoti before probing (Renard, 1999). Additional analyses of cassava leaf surface compounds reveal also the presence of triterpenoids, likely including ß-amyrin (Renard, 1999). These compounds, frequently present on the plant phylloplane (Wollenweber et al., 1999), are known to be repulsive to many insect species (see review by Eigenbrode & Espelie, 1995), however their effect on the cassava mealybugs has not been studied.

20In conclusion, linamarin, a cyanogenic compound present on cassava phylloplane and characteristic of M. esculenta, the host-plant of P. manihoti and P. herreni, appears the most plausible chemical involved in host plant recognition and acceptance by the insects. This has been clearly demonstrated by Renard (1999) in the mobile stages (crawlers) of P. manihoti, the stage most involved in plant colonization.

Feeding behaviour

21By virtue of possessing stylets, the mealybugs have a sap-sucking mode of feeding. The mouthparts of the mealybugs, and in Sternorhyncha in general, are composed of four stylets (two mandibularies and two maxillaries) (Grassé, 1951) (Figure 2.5). In the Coccidae and Pseudococcidae, the stylets are folded into a loop inside the crumena in the labium. The total length of the stylets of P. manihoti is estimated to be about 870-900 µm.

Figure 2.5
Anatomical relationships of the labium (lb) and the stylets (st) in Phenacoccus manihoti (cl, clypeus; cr, crumena) observed in a longitudinal section of the body (a). Transverse section of the stylets folded up into a loop inside the crumena (b). The diameter of the alimentary canal is estimated at about 2 µm. Schematic view of the stylets in Sternorhyncha according to Grassé (1951) (c)

Figure 2.6
The DC-EPG (electropenetrography) system (Tjallingii, 1978b). A gold wire is fixed on the dorsum of the insect with a water-based silver paint (a). The insect is then connected to a probe (b). The probe is connected to an amplifier linked to a computer for electrical signals storing (c). The experiment is carried out in a Faraday cage (d)

22The feeding behaviour of the Coccidae and Pseudococcidae has been described by several authors (Pesson, 1944; Albrigo & Brooks, 1977; Campbell, 1990; Molyneux et al., 1990), all of whom report a phloemophagous (phloem-sap feeding) behaviour. An assessment of the pathways followed by the stylets through the leaf tissues is now possible using electropenetrography (EPG) (Tjallingii, 1978b) (Figure 2.6). This technique, developed on aphids, allows direct electrical monitoring of the stylets’ pathway inside the plant tissues and was used for the first time on mealybug species (particularly P. manihoti and P. herreni) by Calatayud and co-workers (1994a; 2001a). Similarities in the EPGs from mealybugs and those of aphids and whiteflies (Tjallingii, 1978b; Janssen et al., 1989) allowed adoption of a standard labelling pattern defined by Tjallingii (1988). The EPG data from aphids has been used as a reference for correlating the EPG patterns with various components of the stylet penetration process (see Figure 2.7 for interpretation of EPG patterns).

23The EPG recordings show clearly the phloemophagous feeding characteristic of mealybugs. How these insects locate the sieve tubes is still not clearly understood, however. The guiding factors are probably a combination of physical and chemical stimuli (Rahbé et al., 2000). Mealybugs seem to display an exclusive extracellular route to the phloem, with periodic intracellular punctures identified by drops in the electrical potential (Figure 2.7). Phases resembling xylem ingestion are occasionally observed.

Figure 2.7
Typical EPG (electropenetrograph) patterns registered with Phenacoccus manihoti and P. herreni using the DC-EPG system of Tjallingii (1988). Model of 4 hours’ recording, showing among the paths sequences of contact of mouthparts with leaf tissues (A and B, beginning of penetration showing big waves of salivation); cell-wall activities (C, showing small waves of salivation); cell punctures (pd, potential drop due to crossing of the membrane cell and its transmembrane potential); xylem sap ingestion (Xyl—no potential drop is registered because the xylem cells do not possess a cell membrane); and phloem cell ingestion, beginning with a potential drop due to the puncture of the phloem cell membrane and showing two sub-patterns (E1 and E2) of peaks of sap ingestion (Calatayud et al., 1994a and 2001a). np: non-penetration, i.e. stylets outside the leaf tissues

24A comparison of EPGs from mealybugs (P. manihoti and P. herreni) with those from aphids shows some differences that could be significant in distinguishing their interactions with the host plant. As observed by Calatayud and colleagues (1994a), the main differences include the following: first, the cell puncture (pd pattern in Figure 2.7) duration is higher (about 10.20 s) in mealybugs, including two other Pseudoccoccidea species (Ferrisia virgata Cock. and Rastrococcus invadens Williams), while it is only about 5.7 s in aphids. Secondly, the number of cell punctures before reaching the phloem sap is much lower in mealybugs (10.20/h) on all plants than in aphids (50.60/h). Lastly, the minimal time to reach the phloem is much higher in mealybugs, at 1.5 h on a favourable host and 2.8 h on an unfavourable host for P. manihoti, while it is often less than 15 min for aphids. These differences emphasize the lower mobility of mealybugs within their feeding sequence as compared to aphids.

25Phenacoccus herreni males feed only on the phloem sap of the host plant until the second instar; while in the cocoon, they do not feed until they emerge as winged adults. The females, on the other hand, feed on phloem sap throughout their life cycle, indicating that they cause more damage to the cassava plant than males (Polania et al., 1999).

26Influence of plant species: A comparison of the behaviour of P. manihoti on different plant species of varying preference reveals significant correlations between some EPG parameters and the host status or the plant’s susceptibility index (Calatayud et al., 1994a). Host status appears to be linked to phloem accessibility, suggesting that early plant rejection due to delays in phloem-finding may result in antixenosis.

27Influence of location on leaf: It has been demonstrated that in the case of P. herreni, the insect’s location on a leaf strongly influences the EPG parameters, and thus its feeding behaviour (Calatayud et al., 2001a). Studies have shown that feeding near a major leaf vein—the general location of cassava mealybugs on leaves—facilitates phloem-finding behaviour. In conclusion, it is clear that pre-phloem interactions (mainly the intercellular pathways of the stylets) are the most important for host-plant acceptance or feeding site location.

28Influence of plant chemistry: To investigate the links between the feeding behaviour of P. manihoti and the leaf chemistry of the host plant, chemical analysis of cassava leaves was performed by Calatayud and co-workers (1994a). Some secondary compounds were found in leaf apoplastic (intercellular) fluids, but none of the cassava varieties tested, including wild cassava species such as M. cecropiafolia, M. violaceae and the hybrid M. esculenta x M. glaziovii, showed detectable amounts of alkaloids (P.-A. Calatayud, unpublished data).

29Cyanogenic compounds have been found to be restricted to true hosts (cassava). Cyanogenic compounds in the bound form of glucosides, mainly linamarin, are generally present in the vacuolar compartment of cassava meristematic tissues (Conn, 1980). When correlating cyanide content with EPG parameters, the best correlation has been observed with the duration of intracellular punctures (Calatayud et al., 1994a), indicating a possible use of cyanogenic compounds as allelochemicals for host recognition during the cell penetration process by P. manihoti.

30No correlation has been shown between flavonoid levels and any of the individual probing parameters, nor with the susceptibility index (cf antixenosis), indicating that the flavonoids are not involved in the initial interaction between the mealybug and its host. On the other hand, the level of phenolic acids in the leaves has been shown to be strongly correlated to the EPG parameters, and is associated with a delay in reaching the phloem vessels (Calatayud et al., 1994a). On the less preferred cassava variety with a high phenolic acid level in its extracellular fluids, the mealybugs spent the longest time in searching for the phloem. This is interesting in view of the role of phenolic acids in cell wall structures and as precursors of lignins, cutins/suberins and phenolic-coupled pectins, all of which could interact with the oxidizing enzymes found in the saliva of such phloemophagous insects (Fry, 1983; Goodman, 1986).

Table des illustrations

Légende Figure 2.1Antenna of 1st instar of Phenacoccus manihoti. Ventral view of the 9th antennal segment (a) showing the position of the different sensilla; the longest sensilla, which have a chemoreceptive function, are indicated by the arrows. Detail of one of these sensilla (b) showing a perforated cuticle with numerous pores (Le Rü et al., 1995a)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-1.jpg
Fichier image/jpeg, 105k
Légende Figure 2.2Labium of 4th instar larvae of Phenacoccus manihoti, showing six sensilla having contact chemoreceptor as well as mechanoreceptor functions (stars) and two sensilla having olfactory functions (arrows), according to Le Rü et al. (1995b)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-2.jpg
Fichier image/jpeg, 146k
Légende Figure 2.3Characteristic ethograms (a succession of behavioural items observed over a time period) of the Phenacoccus manihoti testing behaviour observed over 90 min on leaves of two host plants (two cassava varieties) and one non-host plant (Talinum triangularae) (after Renard, 1999)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-3.jpg
Fichier image/jpeg, 98k
Légende Figure 2.4Linamarin structure according to Conn (1980)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-4.jpg
Fichier image/jpeg, 27k
Légende Figure 2.5Anatomical relationships of the labium (lb) and the stylets (st) in Phenacoccus manihoti (cl, clypeus; cr, crumena) observed in a longitudinal section of the body (a). Transverse section of the stylets folded up into a loop inside the crumena (b). The diameter of the alimentary canal is estimated at about 2 µm. Schematic view of the stylets in Sternorhyncha according to Grassé (1951) (c)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-5.jpg
Fichier image/jpeg, 180k
Légende Figure 2.6The DC-EPG (electropenetrography) system (Tjallingii, 1978b). A gold wire is fixed on the dorsum of the insect with a water-based silver paint (a). The insect is then connected to a probe (b). The probe is connected to an amplifier linked to a computer for electrical signals storing (c). The experiment is carried out in a Faraday cage (d)
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-6.jpg
Fichier image/jpeg, 216k
Légende Figure 2.7Typical EPG (electropenetrograph) patterns registered with Phenacoccus manihoti and P. herreni using the DC-EPG system of Tjallingii (1988). Model of 4 hours’ recording, showing among the paths sequences of contact of mouthparts with leaf tissues (A and B, beginning of penetration showing big waves of salivation); cell-wall activities (C, showing small waves of salivation); cell punctures (pd, potential drop due to crossing of the membrane cell and its transmembrane potential); xylem sap ingestion (Xyl—no potential drop is registered because the xylem cells do not possess a cell membrane); and phloem cell ingestion, beginning with a potential drop due to the puncture of the phloem cell membrane and showing two sub-patterns (E1 and E2) of peaks of sap ingestion (Calatayud et al., 1994a and 2001a). np: non-penetration, i.e. stylets outside the leaf tissues
URL http://books.openedition.org/irdeditions/docannexe/image/9876/img-7.jpg
Fichier image/jpeg, 135k

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search