Aphis glycines Matsumura (Hemiptera: Aphididae), a major pest of soybean, Glycine max (L.) Merril (Fabaceae), is an invasive species in the United States. It causes direct and indirect injury by removing photo assimilates from plant vascular tissues, stunting plant growth, reducing photosynthetic rates (Beckendorf et al. 2008), and transmitting plant viruses (Clark & Perry 2002).

Plant resistance has been proposed as a valuable tool for managing A. glycines (Hill et al. 2012). Resistance mechanisms are classified into 3 categories: antixenosis, antibiosis, and tolerance. Antixenosis is defined as a plant defense strategy in which the presence of morphological, physical, or chemical plant traits affect the behavior of arthropods. Antibiosis occurs when plant resistance mechanisms deleteriously affect the biology (physiology) of the arthropod without affecting its behavior. In tolerance, the plant does not cause any biological or behavioral change to the arthropod, but it has the ability to withstand or recover from the attack of the arthropod while maintaining its productive capacity (Smith 2005).

Soybean accessions from the USDA soybean collection have been screened to identify genes that confer resistance to A. glycines (Hill et al. 2006). At least 8 resistant soybean lines involving Rag (Resistance to Aphis glycines) genes linked to 4 chromosomes, numbered from Rag1 to Rag6, have been documented (Hill et al. 2010; Zhang et al. 2017). In ‘Dowling,’ antibiosis is expressed and is controlled by the dominant gene Rag1 (Hill et al. 2006). The gene Rag2 was identified as responsible for antibiosis resistance to A. glycines in PI 243540 (Kang et al. 2008), and in PI 200538 (Hill et al. 2009; Kim et al. 2010). Studies evaluating genotypes with both Rag1 and Rag2 genes have demonstrated that A. glycines development was reduced relative to lines with either Rag1 or Rag2 alone (Hill et al. 2006, 2009). From the known A. glycines biotypes (defined by the capacity to develop on aphid-resistant soybeans genotypes), biotype 1 is unable to colonize soybean plants containing Rag1 or Rag2 (Hill et al. 2010; Chandran et al. 2013), biotype 2 can colonize Rag1, but does not colonize Rag2 soybeans (Kim et al. 2008; Chandran et al. 2013), and biotype 3 is able to colonize Rag2 plants as well as some others expressing different Rag genes (Hill et al. 2010, 2012). In other studies, virulent populations of A. glycines (biotype 4) were capable of overcoming either the Rag1, Rag2, or both genes (Kim et al. 2008; Hill et al. 2010; Alt & Ryan-Mahmutagic 2013).

The electrical penetration graph (EPG), using alternate current (AC), has been used to describe the feeding behavior of insects, and to analyze resistance expression in plant genotypes (Reese et al. 2000; Diaz-Montano et al. 2007; Crompton & Ode 2010; Zhu et al. 2011; Todd et al. 2016). The electrical penetration graph technique is based on an electrical circuit in which an insect and plant are integrated. Once the insect inserts its mouthparts into the plant, the circuit is complete and changes in electrical resistance, and bio-voltages (waveforms) are recorded over time, providing continuous information on feeding and probing phases that can be correlated to insect stylet penetration behavior (Sauvion & Rahbe 1999; Backus & Bennett 2009). The resulting waveforms can provide important information about physical location of plant resistance factors, and the time course of insect responses to these factors (Tjallingii 1985; Jiang et al. 2001).

Table 1.

Mean number and duration of electrical penetration graph feeding parameters (± SE) for A. glycines feeding on resistant and susceptible soybean entries for 15 h.

During probing, the location of A. glycines stylets can be established, providing the basis for correlating feeding behavior with interactions within plant tissues, which is useful to determine the location of resistance factors in plant tissues (Jiang et al. 2001). The feeding process of the insects present specific waveforms (Reese et al. 2000). The waveforms described are the sieve element phase (represented by waveform E, and divided in the process of active salivation (E1) and phloem ingestion (E2), xylem feeding (G), pathway phase (P), potential drop (pd), and non-probing (Z) (Todd et al. 2016). The duration of the sieve element phase is longer in susceptible genotypes and shorter in resistant genotypes (Diaz-Montano et al. 2007). Researchers concluded that antibiotic metabolites reside in the phloem tissue (Crompton & Ode 2010). In PI 243540 (Rag2) and the susceptible genotype ‘Wyandot,’ it was observed that aphids had fewer prolonged phases of active salivation (E1), pathway events, and non-probing intervals, and that fewer insects reached the phloem on PI 243540 (Todd et al. 2016).

Although there are relevant studies using the electrical penetration graph technique to evaluate the feeding behavior of A. glycines, there is no study simultaneously evaluating soybean genotypes expressing Rag1, Rag2, and the tolerant KS4202. The contribution of a simultaneous study is the opportunity in evaluating the resistant expression of the soybean genotypes in the feeding behavior of A. glycines under a standardized condition (insect colony and generation; plant condition; and electrical penetration graph system). In addition, for the first time, KS4202 was used and compared to other genotypes as a tolerant, not a susceptible genotype, to A. glycines. Therefore, there is a need for electrical penetration graph studies in order to understand more thoroughly the feeding behavior of the aphids on resistant soybean lines. The objective of this work was to use the electrical penetration graph technique to compare the feeding behavior of A. glycines in soybean genotypes expressing antibiosis, antixenosis, and tolerance resistance.

Materials and Methods

Results

The electrical penetration graph feeding parameters of the 4 soybean genotypes are reported in Table 1. The time that A. glycines spent in the first sieve element phase, the number of xylem phases and sieve element phases, and duration of xylem phases did not differ among the genotypes. Aphids feeding on ‘SD76RR’ differed from the other genotypes by having a prolonged period for time to the first pathway (first probe).

The number of potential drops was higher in ‘Dowling’ and PI 200538, differing from KS4202 (Fig. 1). Genotypes ‘SD76RR’ and KS4202 had significantly fewer number of pathway phases than the other genotypes (Table 1). Additionally, aphids had a significantly greater number of non-probing events on ‘Dowling’ and PI 200538, relative to the other genotypes (Table 1).

Fig. 1.

Mean number of potential drops by Aphis glycines on soybean genotypes for a 15 h (900 min) period.

In relation to duration of the evaluated parameters (Table 1), the pathway phases were shorter for aphids feeding on KS4202 compared to the other genotypes. For duration of non-probing events, aphids feeding on ‘Dowling’ had the longest value while those on ‘SD76RR’ had the shortest.

Aphids fed on ‘Dowling’ had the shortest duration of sieve element phases (Fig. 2). The number of aphids that reached the sieve element phase were similar among the genotypes, although the number was lowest for ‘Dowling. ’

Discussion

The results of this study demonstrated that the different resistance categories present in the evaluated soybean genotypes significantly affect the feeding behavior of the aphid. There were no differences in the time that aphids took to reach the sieve element phase among the soybean genotypes. Similarly, ‘Dowling,’ PI 200538, and KS4202 did not differ from each other for the time to first probe. Because the genotypes had a short or similar period of time for these parameters, resistance factors do not appear to be present in the epidermis or mesophyll tissue layers (Crompton & Ode 2010).

Aphids spent similar periods in the xylem phase on all genotypes, as has been reported in another study involving other genotypes, as well as KS4202 and ‘Dowling’ (Diaz-Montano et al. 2007). Xylem ingestion is important for the insects to preserve a water balance (Spiller et al. 1990); however, xylem sap may not provide aphids with an adequate balance of nutrients as phloem sap (Powell & Hardie 2002).

The shortest duration of sieve element phase and the longest duration of non-probing occurred for aphids feeding on ‘Dowling,’ demonstrating the expression of resistance of ‘Dowling’ to A. glycines. This feeding behavior indicates the expression of antixenosis because the phloem provokes a feeding deterrence to the aphids (Roux et al. 2008).

In another study involving the A. glycines resistance genotypes ‘Jackson,’ ‘Dowling,’ Pioneer 95B97, and K1639, the aphids also spent less time in sieve element phases relative to KS4202 (Diaz-Montano et al. 2007). In a study examining A. glycines feeding behavior on plants treated with thiamethoxam, the insect spent less time feeding in the sieve element phase and had a long duration of non-probing events, which also demonstrates aphid feeding behavior under negative conditions (Stamm et al. 2013).

Although there were no significant differences, out of 14 replications for each genotype, fewer aphids reached the sieve element phase in ‘Dowling’ compared to the other genotypes. In another study, ‘Dowling’ also had the lowest number of aphids that reached the sieve element phase in 8 hr (Zhu et al. 2011), so we expected that fewer aphids would reach the sieve element phase on ‘Dowling’ in our study, and consequently, a higher difference in this parameter among the genotypes would be observed. However, it is important to point out that aphid feeding behavior in our study was evaluated for a period of 15 hr (Todd et al. 2016), while many other studies evaluated the aphids for approximately 8 or 9 hr (Diaz-Montano et al. 2007; Crompton & Ode 2010). This longer period of study would inform possible variation in feeding behavior and the insects would have more time to reach the phloem phase of the host plants.

Fig. 2.

Mean duration of sieve element phase by Aphis glycines on soybean genotypes for a 15 h (900 min) period.

The results of this study suggest that ‘Dowling’ expresses antixenosis or antibiosis to A. glycines through phloem resistance factors, the phloem characteristics being the principal component of induced resistance (Diaz-Montano et al. 2007; Crompton & Ode 2010). Several phloem characteristics have been correlated with antibiosis to A. glycines, including phloem protein coagulation and callose accumulation (Tjallingii 2006), deposition of lectins (Bostwick et al. 1992; Down et al. 1996), protease inhibitors (Kehr 2006), secondary metabolites (Dinant et al. 2010), or lack of sufficient nutrients in phloem sap (Smith 2005). These compounds or resistance factors may affect the aphids by affecting prolonged feeding, which would suggest antixenosis (Will et al. 2007), or by negatively affecting their development (e.g., weight and fecundity), suggesting antibiosis (Halkier & Gershenzon 2006; Kim et al. 2008).

In addition, ‘Dowling’ also showed long pathway phases, which are often described to reflect antixenosis (Diaz-Montano et al. 2007; Roux et al. 2008). Waves of pathway phases represent non-feeding behavior in the phloem or xylem. They comprise a sequence of feeding events where aphids puncture individual cells with their stylet tips to sample cell sap and sieve elements. This feeding behavior is fundamental in the process of accepting or rejecting a host plant (Jiang & Walker 2001).

Aphids feeding on PI 200538 had a greater number and duration of pathway phases, similar to ‘Dowling’; however, those aphids had more sustained sieve element phases and spent less time in non-probing behavior. Although aphids fed on PI 200538 demonstrated inconsistent values under the conditions of this study, previous studies and the presence of the Rag2 gene verify the resistance to A. glycines (including biotype 2). The PI 200538 is used as the male parent in crosses with ‘Dowling’ to obtain resistant genotypes in soybean breeding programs (Hill et al. 2009).

Due to the previous studies that verify tolerance of KS4202 to A. glycines, it was expected that this genotype would express different electrical penetration graph values compared to the resistant ‘Dowling’ and PI 200538. In this electrical penetration graph study, aphids feeding on KS4202 had long durations of sieve element phases and short durations of non-probing and pathway phase, supporting the expression of tolerance and not antibiosis or antixenosis in KS4202. In another electrical penetration graph study comparing KS4202 with other genotypes, including ‘Dowling,’ the results indicated that KS4202 had the longest duration of sieve element phase (Diaz-Montano et al. 2007). Similar results were documented when KS4202 was compared to KS16121, KS1613, and KS1642 (Zhu et al. 2011).

Genotype KS4202 had the lowest potential drop value compared to the other genotypes. The potential drops are considered brief intracellular punctures by the stylets along their pathway behavior (Tjallingii & Esch 1993). Aphids inject watery saliva into the punctured cell, using their stylets to collect sap samples (Martin et al. 1997). Different chemical cues can be present in these samples and thus be correlated to host plant acceptance by the insects (Tjallingii 2006). Because of this characteristic, a high number of potential drops is associated with greater difficulty of reaching phloem phases, and consequently, resistance of soybean genotypes (Chandran et al. 2013). On other hand, some research hypothesizes that a greater number of potential drops is an indication of host plant acceptance by the insect, enabling it to reach the sieve element phase quickly (Diaz-Montano et al. 2007).

The feeding behavior of soybeans aphids in KS4202 suggests that the insects do not suffer negative effects when feeding on this host plant. In addition, these results support that this genotype expresses tolerance to A. glycines (Prochaska et al. 2015). The expression of tolerance in this genotype may be associated with the upregulation of detoxification mechanisms, such as peroxidases and faster regulation of RuBP (ribulose-1,5-biphosphate) (Pierson et al. 2011). Recently, several genes were identified with roles that are connected to plant defenses and the scavenging of reactive oxygen species, which supports the hypothesis that these genes may be correlated to tolerance to A. glycines (Prochaska et al. 2015). The present study provides an important dataset describing A. glycines feeding behavior on A. glycines resistant soybean genotypes, and contributes to a better understanding of resistance mechanisms. New research should be conducted to investigate the tolerance response of KS4202 to A. glycines, which would contribute to soybean breeding programs aimed at developing A. glycines resistant soybean.