Introduction

Bacterial symbionts are widespread among invertebrates and are quite common in insects. Their associations with host insects can lead to interactions ranging from mutualism to parasitism (Baumann 2005, Duron & Hurst 2013, Klepzig et al 2009). In insects, endosymbiotic bacteria are categorized into two groups: obligate/primary and facultative/secondary symbionts (Buchner 1965, Baumann 2005). Obligate symbionts establish mutualistic interactions with their hosts, often contributing to insect nutrition by providing essential nutrients that are limited or completely absent from the diet of the host (Moran et al 2005, Douglas 2009). Facultative symbionts are not vital for survival and/or reproduction of their hosts, but may influence their fitness traits by altering host response to a number of abiotic and biotic factors (Oliver et al 2003, Oliver et al 2010, Su et al 2013, Asplen et al 2014, Gao et al 2014, Wagner et al 2015).

Several aspects of aphid-symbiont interactions have been studied. The genomic composition and physiological contributions of the aphid obligate symbiont Buchnera aphidicola have been determined for a number of aphid species, and the effects of several commonly reported secondary symbionts of aphids (Hamiltonella defensa, Serratia symbiotica, Regiella insecticola, Spiroplasma sp., Arsenophonus sp., and Rickettsia sp.) have been widely investigated (Fukatsu 1994, Chen & Purcell 1997, Fukatsu et al 2001, Leonardo & Muiru 2003, Oliver et al 2003, Douglas et al 2006, Simon et al 2011, Lukasik et al 2013b, Tsuchida et al 2004, Wagner et al 2015).

Aphid secondary symbionts influence the fitness traits of their hosts in different ways. Arsenophonus, H. defensa, S. symbiotica, Spiroplasma, and R. insecticola have been reported to protect their host aphids from natural enemies. Hamiltonella defensa and S. symbiotica provide protection against macrobials (parasitoids) (Oliver et al 2003), Spiroplasma protect against microbials and R. insecticola against both biological agents (Ferrari et al 2001, Ferrari et al 2004, Vorburger et al 2010, Lukasik et al 2013a), and Arsenophonus have no protective role (Wulff et al 2013). Pea aphid resistance to parasitoids was much higher when associated with H. defensa, and resistance was dependent on symbiont strain and presence of the infective bacteriophage APSE (Oliver et al 2003, 2005, 2009, 2014).

Aphid secondary symbionts have also been shown to affect other host traits. Besides influencing host resistance to natural enemies, S. symbiotica and R. insecticola have also been linked to aphid heat tolerance (Chen et al 2000, Montllor et al 2002, Russell & Moran 2006). Regiella insecticola and Arsenophonus may also interfere in host plant use, as aphid fitness on host plants is different in symbiont-free and symbiont-associated aphids (Leonardo & Muiru 2003, Ferrari et al 2006). Spiroplasma has been shown to affect sex determination in sexual generations of aphids by eliminating male progeny (Simon et al 2011) and to induce fitness costs by reducing aphid fecundity and longevity (Fukatsu et al 2001, Simon et al 2011).

Secondary symbiont diversity can vary greatly within and among aphid species (Tsuchida et al 2002, Haynes et al 2003, Brady et al 2014). Within-species variation may occur due to ecological factors related to aphid geographical distribution and/or host plant use (Ferrari et al 2004, McLean et al 2011), but little information is available on among-species variation (Brady et al 2014). Although a fair amount of information has been published on the diversity and interactions of aphid symbionts, most of the available data focuses on polyphagous aphids, primarily the pea aphid Acyrthosiphon pisum (Harris) (Tsuchida et al 2004, Ferrari et al 2012, Russell et al 2013). Very few studies have focused on aphid species that have a narrow host range or are found in tropical areas (Henry et al 2015, Zytynska & Weisser 2016).

Aphis (Toxoptera) citricidus (Kirkaldy) is an oligophagous aphid commonly found in tropical areas feeding on Citrus plants and a few relative species. It is also an efficient vector of citrus tristeza virus (CTV) to citrus plants. Another aphid species, Aphis (Toxoptera) aurantii (Boyer de Fonscolombe), usually co-occurs with A. citricidus in Citrus groves. However, A. aurantii is polyphagous and a poor vector of CTV (Halbert & Brown 1996, Halbert et al 2004, Lagos & Voeftlin 2014).

As aphid-symbiont associations in aphids with restricted diet breadth still remain to be investigated, we screened for the most commonly reported aphid-associated facultative symbionts in an oligophagous (A. citricidus) as compared to a polyphagous (A. aurantii) aphid species within the same genus and in the same ecological niche.

Material and Methods

Aphid isolines

Adult aphids were collected from an ecological niche on unmanaged Citrus sp. and Murraya paniculata trees in the municipality of Piracicaba, SP, Brazil. Samples were collected randomly from 20 trees throughout a year. Aphids were brought to the laboratory and classified following the key to wingless adults provided by Halbert & Brown (1996).

After species identification, specimens of A. citricidus and A. aurantii were individually placed on new shoots of Citrus sinensis seedlings in a rearing cage (50 cm high × 15 cm diameter, with two lateral openings covered with cloth for ventilation). Cages were kept under controlled conditions (25 ± 2°C, 60 ± 10% RH, 14 h photophase) for a week to observe aphid development and eliminate parasitized and/or unhealthy aphids. Symbiont assessments were carried out by diagnostic PCR to detect the six most common secondary symbionts associated with aphids, as well as the primary symbiont B. aphidicola and the APSE-1 bacteriophage (Sandstrom et al 2001, Tsuchida et al 2002, Oliver et al 2005, Oliver et al 2006).

DNA extraction and symbiont screening

Three adult aphids from each isoline were randomly selected, placed into 1.5-mL plastic tubes, and subjected to genomic DNA (gDNA) extraction following Sunnucks et al (1996). DNA quality was verified by spectrophotometer measurements of the A260/280 ratio and by agarose gel electrophoresis on a 1% gel slab containing 0.5 μg/mL ethidium bromide in Tris-acetate-EDTA buffer (40 mM Tris-acetate, 1 mM EDTA, pH 7.2) at 5 V/cm, followed by visualization on a UV transilluminator (DNR Bio-imaging system).

Diagnostic PCR reactions were set to a final volume of 25 μL and contained 20 ng/μL of gDNA, 1× PCR buffer, 1.5 mM MgCl2, 10 μM dNTP, 0.32 μM of each primer, and 0.5 U of Taq polymerase. Cycling conditions and primer sets depended on the symbiont screened (Table 1). Amplicons were resolved by agarose gel electrophoresis on a 1.5% gel slab and visualized under the same conditions described above.

Table 1 Primers and cycling conditions used for assessing the association of the most common aphid symbionts infecting Aphis (Toxoptera) citricidus and Aphis (Toxoptera) aurantii by using diagnostic PCR.
Full size table

Statistical analysis

Differences in frequency of association of secondary symbionts with the oligophagous A. citricidus and the polyphagous A. aurantii were tested using a contingency table with a chi-square test (p < 0.05).

Results

Establishing isolines from field-collected adults was quite difficult, ultimately limiting the number of isolines screened for symbionts. Even so, 47 isolines of A. citricidus and 37 of A. aurantii originating from different field collections were subjected to diagnostic PCRs to determine their association with the most commonly reported aphid symbionts.

Aphis citricidus and A. aurantii showed high rates of infection by secondary symbionts, as exclusive infections by the primary symbiont B. aphidicola comprised only 18% of all tested isolines of A. citricidus and 7% of A. aurantii isolines (Fig 1). Overall, A. aurantii was more frequently associated with secondary symbionts than A. citricidus (x 2 = 27.0, df = 1, p < 0.001), and the frequency of occurrence of each facultative symbiont was different between aphid species. Spiroplasma was the most frequent secondary symbiont in both A. citricidus (25/47) and A. aurantii (31/37), but occurred at a higher frequency in A. aurantii (x 2 = 7.5, df = 1, p < 0.001). In addition, Spiroplasma was the only secondary symbiont to be persistently associated with A. aurantii in single and multiple infections with other secondary symbionts (Fig 1 and Table 2). Many isolines of A. citricidus harbored H. defensa (14/47) or Arsenophonus (11/47), whereas few isolines carried Rickettsia (3/47), R. insecticola (2/47), or S. symbiotica (2/47). On the other hand, several isolines of A. aurantii were infected with Rickettsia (9/37) or H. defensa (7/37), a few with S. symbiotica (2/37) or R. insecticola (1/37), and none with Arsenophonus. In addition, even though many isolines of A. citricidus and A. aurantii were infected with H. defensa, we were unable to detect infections with the bacteriophage APSE-1 (Fig 1). While Arsenophonus association with A. citricidus was much more common than with A. aurantii (x 2 = 10.2, df = 1, p < 0.001), Rickettsia showed the opposite trend (x 2 = 5.2, df = 1, p < 0.001). No difference in association with H. defensa (x 2 = 1.5, df = 1, p = 0.22), S. symbiotica (x 2 = 0.05, df = 1, p = 0.83), or R. insecticola (x 2 = 0.16, df = 1, p = 0.69) was detected between the two aphids.

Fig 1
figure 1

Occurrence and frequency (%) of facultative symbionts in isolines of Aphis citricidus (A) and Aphis aurantii (B) collected from the same ecological niche.

Full size image
Table 2 Occurrence of facultative symbionts in singly and multiply infected isolines of Aphis (Toxoptera) citricidus and Aphis (Toxoptera) aurantii inhabiting the same ecological niche.
Full size table

In both species, several isolines were infected with multiple facultative symbionts. Nearly two-thirds of A. citricidus isolines were infected with one (34%—16/47) or two (34%—16/47) secondary symbionts, 2% carried three, and 2% carried four facultative symbionts. Almost half of the isolines of A. aurantii (47%) harbored only one secondary symbiont, 26% carried two, 11% carried three facultative symbionts, and none were infected with four facultative symbionts (Fig 2).

Fig 2
figure 2

Isolines (%) of Aphis citricidus and Aphis aurantii from the same ecological niche carrying none, single, or multiple infections of facultative symbionts.

Full size image

An analysis of associations among facultative symbionts in multiple-infected aphids indicated that some multiple infections were more common in A. citricidus than in A. aurantii. In A. citricidus, Spiroplasma and H. defensa were observed in single infection events, followed by several combinations of double, triple, and quadruple secondary symbiont infections, with Spiroplasma as the most common symbiont in aphids with multiple infections. Spiroplasma was present in all multiple infections in A. aurantii and in five out of eight combinations of multiple infections in A. citricidus (Table 2).

Discussion

Richness and abundance of facultative symbionts commonly associated with aphids differed between the closely related aphid species A. citricidus and A. aurantii exploiting the same ecological niche, with the polyphagous species A. aurantii showing higher frequencies of infection by secondary symbionts. Spiroplasma was the most abundant facultative symbiont associated with these aphid species in single or multiple infections. Rickettsia was the second most abundant facultative symbiont associated with A. aurantii, but was seldom associated with A. citricidus; on the other hand, Arsenophonus was common in A. citricidus, but was not associated with A. aurantii. Multiple infections were common. Among these, double infections prevailed, although we observed infections by up to four facultative symbionts.

Hamiltonella defensa, R. insecticola, S. symbiotica, Rickettsia, Spiroplasma, X-type, Arsenophonus, and Wolbachia are regularly reported as facultative symbionts of aphids, although their richness and abundance can differ from one host aphid to another (Sandstrom et al 2001, Tsuchida et al 2002, Augustinos et al 2011, Zytynska & Weisser 2016). The diversity of aphid-associated facultative symbionts has been intensively investigated in polyphagous aphids, and the factors influencing these associations are mostly known for Aphis craccivora, Aphis fabae, A. pisum, and Sitobion avenae (Ferrari et al 2012, Lukasik et al 2013a, Brady et al 2014, Cayetano et al 2015). Aphid genotype, symbiont strain, host plant, and biogeography are among the factors affecting the occurrence of facultative symbionts in aphids (McLean et al 2011, Ferrari et al 2012). Extensive data on the influence of biogeography is still lacking, as most of the available information on facultative symbiont richness and abundance focuses on natural populations of aphids in the Northern Hemisphere (Zytynska & Weisser 2016) and very little is known about aphid species in the Southern Hemisphere, particularly in the neotropics (Telesnicki et al 2012, Arneodo & Ortego 2014). Therefore, our data on A. citricidus and A. aurantii add to current knowledge by providing additional information about tropical areas in the southern part of the Neotropical region, also contributing towards an elucidation of how polyphagy/oligophagy may interfere with aphid-facultative symbiont associations.

Recent research on aphid-symbiont associations has focused on the importance of ecological niche on symbiont distribution, as aphids with similar feeding habits are more likely to share similar symbiont communities (Henry et al 2015). These authors also showed that the mean proportion of aphids carrying S. symbiotica, but not H. defensa or R. insecticola, increased with aphid specialization in host plant use (Henry et al 2015). These data show that certain associations are affected by the life history traits of the host aphid. We did not find such a correlation with S. symbiotica in the polyphagous/oligophagous aphids we tested, as the abundance of infection by S. symbiotica ranged from 4.2% (2/47) in A. citricidus (oligophagous) to 5.4% (2/37) in A. aurantii (polyphagous).

But we did find differences in the frequency at which certain facultative symbionts were associated with A. citricidus and A. aurantii aphids inhabiting the same ecological niche. Arsenophonus was exclusively associated with A. citricidus; frequency of infection by Rickettsia differed between the two aphids, occurring in nearly 24.3% (9/37) of A. aurantii isolines, but only 6.4% (3/47) of A. citricidus isolines tested. Arsenophonus has been associated with host plant specialization in A. craccivora (Wagner et al 2015), and its high abundance in A. citricidus could be associated with the narrow host range of this aphid. Differences in Rickettsia abundance between A. citricidus and A. aurantii could be related to the role this facultative symbiont may play on host defense against microbial infections (Lukasik et al 2013b). This would indicate that A. aurantii may be under higher selection pressure due to pathogen infection than A. citricidus, as maintaining such associations involves fitness costs to the host (Oliver et al 2014). Associations with defensive facultative symbionts are more prevalent under parasitoid selection pressure (Oliver et al 2009, 2014) and are more often observed in aphids that are not tended by ants that provide the same defensive services (Henry et al 2015).

Two other facultative symbionts that reportedly enhance host defense against natural enemies, namely Spiroplasma and H. defensa (Lukasik et al 2013b, Oliver et al 2003), were very frequent in single- or multiple-infected A. citricidus and A. aurantii. This also supports the hypothesis that these aphid species are kept under strong selective pressure by natural enemies in field conditions. Although we found no difference in frequency of infection by H. defensa between the two aphids, which indicates that they may be under the same selective pressure by parasitoids, our data clearly suggest that A. aurantii may be more subject to infection by entomopathogenic fungi than A. citricidus, as Spiroplasma and Rickettsia were detected more frequently in A. aurantii. Spiroplasma is by far the most frequently associated facultative symbiont of A. citricidus and A. aurantii, and although a fair amount of information on the role of this symbiont in fruit flies is known (Cockburn et al 2013, Xie et al 2014, Haselkorn & Jaenike 2015), investigations into its role in aphid associations began only recently (Lukasik et al 2013a, b, c).

Multiple infections by more than two facultative symbionts were also detected in A. citricidus and A. aurantii, but at a lower frequency than single or double infections, as expected. Multiple-infected hosts are subject to a bottleneck as a result of the associated fitness costs (Oliver et al 2006). Most field trials conducted thus far have been carried out with A. pisum, which usually harbors one or two facultative symbionts (Ferrari et al 2012, Russell et al 2013). Data on the dynamics of symbionts associated with A. pisum have suggested that multiple infections may not be efficiently maintained in the population, due to inefficient transmission. In addition, temporal changes in frequency have been reported for all symbionts associated with A. pisum, and increases in the frequency of certain symbionts in response to bioecological stressors have indicated that symbionts participate in seasonal adaptation of aphids (Smith et al 2015). However, initial studies on other species of aphids did not follow this pattern (Chandler et al 2008, Henry et al 2015). Laboratory surveys identified up to four facultative symbionts in a single clone, providing enhanced resistance to parasitoids at the expense of decreased fecundity (Oliver et al 2006). Differences between field and laboratory data can be attributed to the transmission of facultative symbionts, as multiple infections are more prone to transmission failures, which can be more severe under variable environmental field conditions (Oliver et al 2014).

Secondary symbionts occurred in different combinations in multiple-infected A. citricidus and A. aurantii. Surveys of natural populations have rarely reported the abundance at which facultative symbionts co-occur. Co-infections by H. defensa and S. symbiotica have been associated with increased defense against parasitoids, but the cost of this association would decrease its abundance in the field, particularly under low selection pressure (Oliver et al 2006). Hamiltonella and Serratia have also been shown to co-infect A. pisum with Rickettsiella, a facultative symbiont that alters the body color of the aphid due to accumulation of blue-green polycyclic quinones (Tsuchida et al 2010). The accumulation of these compounds and co-occurrence of Rickettsiella with Hamiltonella and Serratia led to the hypothesis that Rickettsiella may also benefit the aphid by improving its immune response against natural enemies (Tsuchida et al 2010). Co-infections involving R. insecticola and H. defensa or the X-type, or Rickettsiella and Spiroplasma, are less frequent in field conditions, probably due to a fitness cost component (Ferrari et al 2012) or the incompatible nature of R. insecticola (Smith et al 2015). Regiella insecticola was very rare, infecting a single line of each species of Aphis studied. In both cases, R. insecticola was detected in triple infections, co-occurring with H. defensa and Arsenophonus in A. citricidus, but with Spiroplasma and Rickettsia in A. aurantii.

We showed that facultative symbiont richness and abundance were different between two aphids with different host ranges in the same ecological niche, with the polyphagous species being more frequently infected with secondary symbionts than the oligophagous one. We also reported Spiroplasma as the most common facultative symbiont of A. citricidus and A. aurantii, but the diversity of associations with other facultative symbionts was quite high between the two aphids, with A. citricidus carrying a considerably higher number of facultative symbiont combinations. However, the relevance of such multiple infections for the host aphid should not be exclusively viewed from the perspective of richness/abundance of associated microbiota, but rather their role and functional complementarity (Zytynska & Weisser 2016). Therefore, the nature of the facultative symbiont associations we detected in natural populations of A. citricidus and A. aurantii, as relates to host range use and role in ecological and fitness traits of host aphids, remains to be investigated.