Introduction

In Mexico, the main pest of soybean (Glycine max L.) is the Mexican soybean weevil Rhyssomatus nigerrimus Fahraeus (Coleoptera: Curculionidae) (López-Guillén et al. 2012b), which causes major damage during the vegetative and reproductive stages of soybeans. The female weevil deposits her eggs inside the pods and the larvae feed on the beans (Terán-Vargas and López Guillén 2014). It has been observed that R. nigerrimus also feeds on the flowers (López-Guillén et al. 2012a). The weevils locate the soybean plant possibly due to the volatiles that the plant produces. Generally, plants emit volatile compounds into the environment (Bautista-Lozada et al. 2012). These volatile compounds play an important role in the plant–insect interaction (Cantúa-Ayala et al. 2019; Hu et al. 2021; Karmakar et al. 2020; Zhang et al. 2016). Among these are the flowers volatiles. Floral volatiles have different functions in attraction, reproduction, dissuasion, and antagonism of herbivorous insects and pollinators (Hetherington-Rauth and Ramírez 2016; Schiestl 2015). Floral volatiles act as attractants, not only of pollinators, but also of herbivorous pests (Wang et al. 2018). In the case of curculionids, several species have been reported to interact with the flowers; for example, the apple blossom beetle Anthonomus pomorum (Linnaeus) (Coleoptera: Curculionidae) is attracted by apple flower buds (Collatz and Dorn 2013), whereas its allied A. rubi (Herbst) prefers strawberry plants with flowers over those without flowers (Mozūraitis et al. 2020) and females of A. musculus (Say) prefer open flowers of cranberry (Vaccinium macrocarpon L.) (Szendrei et al. 2009). Soybean plants are known to emit volatiles that attract herbivores; for example, Riptortus pedestris (Fabricius) (Hemiptera: Alydidae) is attracted by soybean volatiles (Song et al. 2022). It is also known that, when R. nigerrimus damages a soybean plant, induced volatiles are produced (Espadas-Pinacho et al. 2021). However, the composition of the volatiles from soybean flowers, which could attract R. nigerrimus, has not been reported, so it would be helpful to know the behavior of R. nigerrimus and develop a robust base of attractants for the control of this pest. For this reason, in this study we evaluate the response of R. nigerrimus to the volatile flower extracts of both varieties FT-Cristalina-RCH and Flores most frequently cultivated and the compounds that constitute the aroma of the soybean flowers were tentatively identified. Moreover, we tested behavioral and electrophysiological activities of individual compounds as well as blends of the compounds from flowers of both varieties to adults of R. nigerrimus.

Materials and methods

Biological materials

Adults of R. nigerrimus in active state and soybean flowers (varieties FT-Cristalina-RCH and Flores) in stage R2 (complete flowering) were collected in a soybean field located in the first section of the ejido Tinajas (N 14° 22′, W 092° 20′), Tapachula, Chiapas, Mexico. The flowers were placed in 250 mL Erlenmeyer flasks and kept at 5 °C until use. The weevils were sexed following the technique described by López-Guillén et al. (2016). Males and females were placed in separate 1 L plastic recipients. The insects were fed with pieces (0.5 mm × 0.5 mm) of sweet potato (Ipomea batatas L.) daily and kept in the insectarium of ECOSUR (El Colegio de la Frontera Sur), Mexico at a temperature of 25 ± 2 °C, 75 ± 5% relative humidity, and 12:12 h light:dark photoperiod.

Volatile collection

We collected the volatiles emitted by the flowers using the dynamic aeration technique. We placed 5 g in fresh weight of flowers in a 250 mL glass recipient. A current of air (previously purified through a filter with Tenax®) was made to pass over the flowers at a flow of 0.8 L/min. The volatiles were captured in a small glass column with Super Q (50–80 mesh; Water Associates, Milford, MA, USA). The collection process lasted 24 h. The volatiles were then extracted with 400 μL of dichloromethane, placed in 2 mL vials and stored at − 20 °C until analysis. The volatiles collection was performed at a temperature of 27 ± 2 °C, 75 ± 5% relative humidity, and photoperiod of 12:12 h light:dark. Ten replications were done for each variety of soybean.

Behavioral bioassays

The insects used in the bioassays were fasted for 24 h. We assessed the attraction response of R. nigerrimus to volatiles emitted by soybean flowers (var. FT-Cristalina-RCH and Flores) using a Y-shaped olfactometer placed horizontally on the table (main arm 15 cm long, lateral arms at 45° with 10 cm long and 2.3 cm internal diameter). Air, previously humidified and purified with an activated carbon filter, passed at a flow rate of 0.5 L/min through each arm of the olfactometer. In one of the olfactometer arms was placed 1 g of flowers, and the other remained empty as a control. We placed a group of 5 insects at the entrance of the main arm of the olfactometer and allowed them a maximum time of 5 min to select between the treatment and the control. We considered that the insect had made an election when it had entered 2 cm into one of the lateral arms. We changed the position of the Y-tube after five replications, interchanging the place of the glass adaptor (8 cm long, 1.3 cm internal diameter) every five replications to avoid response bias. For both males and females, 60 replications were performed for each soybean variety. The bioassays were carried out between 8:00 and 19:00 h. Before beginning and after each evaluation, the olfactometer and the adaptors were washed with water and neutral soap and placed in an oven at 120 °C for 60 min. The bioassays were conducted at a temperature 35 ± 2 °C, 75 ± 5% relative humidity and used artificial white light (1676 lx). For the bioassays with extracts, 20 µL of the extract or solvent was placed on a 5 cm × 5 cm piece of filter paper (Whatman No. 2, Maidstone, England) and left for 20 s to evaporate the solvent. We then placed a piece of filter paper with the extract in one of the arms of the olfactometer and a piece of filter paper with 20 µL of dichloromethane (control) in the opposite arm. We compared the response of the two sexes of weevils to the volatile extracts of the two soybean varieties simultaneously in the olfactometer. This was replicated 30 times for each volatile extract and each sex.

Chemical analysis

The volatile extracts were analyzed in a gas chromatograph coupled with a mass spectrometer (Shimadzu GC-2010 Plus, Tokyo, Japan) equipped with a 5% phenyl-methyl-silicone column (DB5-MS) 30 m × 0.25 mm internal diameter and 0.25 µm film thickness. The oven temperature program consisted of an initial temperature 50 °C (for 2 min) with increments of 15 °C/min up to a final temperature of 280 °C (for 10 min). The analyzed aliquot was 1 µL per extract of soybean extract obtained by dynamic aeration. Volatile compounds were tentatively identified retention indexes, comparing the spectral data of each compound with the NIST database and then confirmed with the use of reference synthetics when possible. We used the percentage of the peak areas to know their relative proportions and the quantification of selected compounds was carried out using the external standard method, which consisted of preparing a set of solutions of each compound at different concentrations (100 ng/µL, 75 ng/µL, 50 ng/µL, 25 ng/µL and 1 ng/µL), then a calibration curve was constructed with the areas obtained for each compound and related to the concentrations.

Electroantennography (EAG)

We evaluated the antennal response of R. nigerrimus males and females to the volatile extracts from both soybean flower varieties, synthetic blends and individual synthetic compounds through EAG tests. Each weevil was exposed to cold until anesthetized for 1 min to facilitate manipulation. We cut off the insect’s head and inserted the reference electrode at the base, using a glass capillary filled with Ringer solution (NaCl 0.35 g, CaCl2 0.21 g, KCl 0.35 g, and NaHCO3 0.2 g dissolved in 1 L water). The distal tip of the antenna was inserted into the tip of the glass capillary placed in the recording electrode. We used a standard aliquot (1 μL) of the extract, blend, or compound (solution prepared with dichloromethane HPLC grade at a concentration of 1 µg/µL) on a piece of filter paper (0.5 × 1.0 cm, Whatman No.1, Whatman International, Maidstone, United Kingdom) and exposed it to air for 20 s on a glass Petri dish to allow the solvent to evaporate. A glass Pasteur pipette or sample cartridge was then introduced. A new cartridge was prepared for each antenna replicate. As a control, we used a piece of filter paper loaded with 1 µL dichloromethane at the beginning, middle and end of each assay. To present the stimulus, we inserted the tip of the pipette into an orifice located in the upper middle part of a glass tube (10 mm diameter) and passed a flow of air (0.5 L/min) through by activating a pedal connected to a stimulus controller (Syntech CS-05, Hilversum, The Netherlands). The duration of the stimulus was 1 s. A current of purified humid air (0.7 L/min) was constantly directed toward the antenna through a glass tube to eliminate the odors in the system. The signals generated by the antenna were amplified using a controller of intelligent data acquisition (Syntech IDAC-02, Hilversum, The Netherlands) connected to a computer and visualized in a monitor using the software Syntech EAG v.2.7. The response to the different compounds was measured consecutively in the same antenna in a random manner. We evaluated 30 males R. nigerrimus antennae and 30 females R. nigerrimus antennae.

Bioassays with synthetic compounds

Response of adult R. nigerrimus to synthetic compounds

The compounds α-pinene, 1-octen-3-one, 1-octen-3-ol, 2-ethyl-1-hexanol, limonene, undecane, nonanal, and α-copaene were assessed individually at a concentration of 1 µg/µL and blends of the eight compounds from both varieties of flowers were prepared in accord with the concentrations of the aeration extracts (Fig. 4): FT-Cristalina-RCH (6.5, 8.5, 23.0, 49.0, 7.0, 7.0, 14.0, 82.0 ng, respectively) and Flores (4.5, 22.5, 25.0, 25.0, 4.5, 8.0, 11.0, 80.0 ng, respectively). The bioassays were conducted following the methodology described previously in a Y-shaped olfactometer using 1 µL of the synthetic blend to be evaluated and placing as a control a piece of filter paper with 1 µL of dichloromethane, leaving it to evaporate for 20 s before commencing each replication. The two blends were assessed separately with each sex, and also the two synthetic blends were also assessed simultaneously in the Y-shaped olfactometer. We performed 30 replications with R. nigerrimus males and females.

Statistical analysis

We analyzed the percentage of the peak areas of the compounds from flowers of each soybean variety and the quantities of each compound with the Kruskal–Wallis test. We analyzed all the data of the behavioral bioassays using the G test, previously transformed to square root to achieve normality of the data. The insects that did not respond in the bioassays were not included in the analysis. We conducted an analysis of variance (ANOVA) of multiple factors, considering the electrophysiological response to synthetic compounds individually and by R. nigerrimus sex, followed by the Tukey test (α = 0.05) to compare treatment means. The statistical analyses were conducted with R software, version 4.3.1, and the Rcmdr package version 2.8.0 (Fox and Bouchet-Valat 2022).

Results

Behavioral bioassays with flowers and extracts

In the bioassays, adult of R. nigerrimus males were more attracted by flowers of both varieties FT-Cristalina-RCH (G = 209.92, df = 1, p < 0.001) and Flores (G = 2014.50, df = 1, p < 0.001), compared to the control. Likewise, R. nigerrimus females preferred the flowers of the variety Flores (G = 152.59, df = 1, p < 0.001) and FT-Cristalina-RCH (G = 153.70, df = 1, p < 0.001) over the control. However, when the flowers were tested simultaneously there was no significant difference with males (G = 1.11, df = 1, p = 0.29), but there was a significant difference with females (G = 4.24, df = 1, p < 0.05) (Fig. 1a, b).

Fig. 1
figure 1

Mean response (± SE) of R. nigerrimus a males and b females to soybean flowers (G. max L.) variety FT-Cristalina-RCH, Flores, and both flower varieties in bioassays conducted in a Y-tube olfactometer. The asterisk (*p < 0.05; **p < 0.01; ***p < 0.001) indicates significant differences between evaluated treatments (G test). C control, NR non-responding, NS non-significant

Full size image

R. nigerrimus males preferred the volatile extracts from the varieties FT-Cristalina-RCH (G = 5.77, df = 1, p < 0.01) and Flores (G = 15.55, df = 1, p < 0.001) over the control. Females showed the same response pattern, preferring the volatile extracts from both varieties (G = 11.77, df = 1, p < 0.001 in FT-Cristalina-RCH; G = 13.70, df = 1, p < 0.001 in Flores) over the controls. Between males and females, there were no significant differences when we tested the two volatile extracts simultaneously (G = 1.00, df = 1, p = 0.31 in males and G = 1.22, df = 1, p = 0.26 in females) (Fig. 2a, b).

Fig. 2
figure 2

Mean response (± SE) of R. nigerrimus a males and b females to volatile extracts from var. FT-Cristalina-RCH and Flores soybean flowers (G. max L.) and from both varieties in bioassays conducted in a Y-tube olfactometer. The asterisks (*p < 0.05; **p < 0.01; ***p < 0.001) indicate significant differences between the evaluated treatments (G test). C control, NR non-responding, NS non-significant

Full size image

Identification of volatile compounds

The analysis using gas chromatography coupled with mass spectrometry (GC–MS) showed that the composition of the volatiles emitted by the flowers of the two varieties includes at least 11 compounds: α-copaene, 1-octen-3-ol, 2-ethyl-1-hexanol, α-pinene, limonene, undecane, nonanal, 1-octen-3-one, octyl hexanoate, trans-α-bergamotene and calamanene (Table 1). The compounds 1-octen-3-one (H = 6.81, df = 1, p < 0.01), 2-ethyl-1-hexanol (H = 3.93, df = 1, p < 0.05), α-pinene (H = 3.93 df = 1, p < 0.05), 1-octen-3-one (H = 5.77 df = 1, p < 0.01), 2-ethyl-1-hexanol (H = 6.81 df = 1 p < 0.001) and limonene (H = 5.77 df = 1, p < 0.01) vary quantitatively between the two varieties of soybean flowers (Figs. 3, 4).

Table 1 Volatile compounds identified in soybean flowers of the varieties FT-Cristalina-RCH and Flores (Percentage of peak area ± SE)
Full size table
Fig. 3
figure 3

Percentage of peak areas (± SE) of the volatile compounds emitted by flowers of soybean varieties FT-Cristalina-RCH and Flores. Comparisons at p < 0.05 were significant (Kruskal–Wallis test). NS non-significant

Full size image
Fig. 4
figure 4

Quantities (± SE) of the compounds emitted by flowers of the soybean varieties FT-Cristalina-RCH and Flores. Comparisons at p < 0.05 were significant (Kruskal–Wallis test). NS non-significant

Full size image

Electroantennography (EAG)

The antennal response of R. nigerrimus females was more intense than that of males to each of the evaluated treatments (F = 17.88, df = 12, p < 0.001). The following treatments stand out: volatile extract from both soybean flower varieties, synthetic blends, and the compounds α-pinene, 1-octen-3-ol, limonene, nonanal and α-copaene. In general, we observed stronger antennal response from females than from males (F = 128.27, df = 1, p < 0.001) (Fig. 5).

Fig. 5
figure 5

Electroantennographic (± SE) response of R. nigerrimus males and females to volatile extracts, synthetic blends and individual compounds from soybean flowers. Capital letters indicate differences between treatments in females. Lowercase letters indicate differences between treatments in males (p < 0.05; ANOVA and Tukey test). DCM dichloromethane

Full size image

Bioassays with synthetic compounds

Males were more attracted by α-copaene (G = 10.60, df = 1, p < 0.001), 1-octen-3-ol (G = 7.98, df = 1, p < 0.01), and α-pinene (G = 10.68, df = 1, p < 0.001) than by the controls (Fig. 6a), while females were attracted by the compounds α-pinene (G = 35.97, df = 1, p < 0.001), 1-octen-3-ol (G = 4.52, df = 1, p < 0.05) and α-copaene (G = 31.07, df = 1, p < 0.001) (Fig. 6b).

Fig. 6
figure 6

Mean response (± SE) of R. nigerrimus to individual synthetic compounds, a males and b females. The asterisks (*p < 0.05; **p < 0.01; ***p < 0.001) indicate significant differences between evaluated treatments (G test). C control, NR non-responding, NS non-significant

Full size image

Male and female weevils significantly preferred the eight compounds blends of the varieties FT-Cristalina-RCH (G = 17.76, df = 1, p < 0.001 in males and G = 13.47, df = 1, p < 0.001 in females) and Flores (G = 17.81, df = 1, p < 0.001 in males and G = 10.96, df = 1, p < 0.001 in females) over the controls in all the bioassays. However, we found no significant differences when we evaluated both synthetic blends (G = 2.00, df = 1, p = 0.15 in males and G = 0.11, df = 1, p = 0.73 in females) (Fig. 7a, b).

Fig. 7
figure 7

Mean response (± SE) of R. nigerrimus a males and b females to the eight compounds blend that simulate the odor of varieties of soybean flowers (G. max L.): FT-Cristalina-RCH, Flores and both flower varieties, in tests conducted in a Y-tube olfactometer. The asterisks (*p < 0.05; **p < 0.01; ***p < 0.001) indicate significant differences between evaluated treatments (G test). C control, NR non-responding, NS non-significant

Full size image

Discussion

In this study we found that the Mexican soybean weevil R. nigerrimus is attracted to volatiles emitted by the flowers of the soybean plant and their aeration extracts. These results are congruent with the knowledge that many phytophagous species and pollinators are attracted by volatiles emitted by flowers (Baldelomar et al. 2018; Bouwmeester et al. 2019; Dudareva et al. 2013; Hetherington-Rauth and Ramírez 2016; Junker et al. 2010; Marín-Loaiza and Céspedes 2007). Like R. nigerrimus, other weevils such as A. pomorum (L.). (Coleoptera: Curculionidae) and its allied species prefer their host plants in flowering stage (Collatz and Dorn 2013; Mozūraitis et al. 2020). Other insects, such as the small green plant bug Apolygus lucorum (Meyer-Dür) (Hemiptera: Miridae), prefer host plants also in the flowering stage (Pan et al. 2015).

The two varieties of flowers evaluated in this study emit the same mixture of volatiles. However, they vary quantitatively. Floral aromas vary not only among species but also within a single plant species, as described by Campbell et al. (2019) and Delle-Vedove et al. (2017). However, there are few studies that center on intraspecific variation of floral aromas in a given geographic site, rather, they focus on comparing populations at different spatial scales and type of pollination (Aceves-Chong et al. 2018; Dötterl et al. 2005; Farré-Armengol et al. 2015; Wang et al. 2019). Our study permits us to appreciate how the volatile profiles of the flowers of the two soybean varieties differed: the compounds 1-octen-3-one and 2-ethyl-1-hexanol differed significantly in proportions, while α-pinene, 1-octen-3-one, limonene and 2-ethyl-1-hexanol differed in quantities. Nevertheless, the two mixtures of volatiles attract R. nigerrimus. However, we do not exclude the possibility that R. nigerrimus uses other factor to find its host as the flower color, size and shape, since when both physical flowers were evaluated simultaneously in bioassays, females preferring the “Flores” variety and when evaluating both volatile extracts did not show preferences. This study must be complemented with additional experiments to corroborate this aspect.

The compounds 1-octen-3-ol, α-copaene and 2-ethyl-1-hexanol were identified as those present in higher proportion in the FT-Cristalina-RCH variety (24.40, 27.19 and 18.03%, respectively), while in the variety Flores the major compounds were 1-octen-3-ol, α-copaene and 1-octen-3-one (25.02, 28.00, 16.62%, respectively). These compounds were also reported by Espadas-Pinacho et al. (2021) in volatiles from soybean plants and pods, but at a lower proportion. Likewise, Boue et al. (2003) reported the presence of 1-octen-3-ol, 2-ethyl-1-hexanol and nonanal in soybean pods, while Cortés (2016) reported 1-octen-3-ol as one of the volatiles consistently released by soybean pods, leaves and seedlings. Nevertheless, one limitation of this work is that flowers compounds were collected using detached flowers and not with attached flowers, because the soybean flowers size and number in a plant make difficult to collect their volatiles using dynamic aeration in field condition.

Rhyssomatus nigerrimus responds electrophysiologically to flower volatiles. In this study, we found that R. nigerrimus females exhibit stronger antennal response than males to the different treatments. Likewise, Ceballos et al. (2015) and Adhikari et al. (2002) reported that female Bruchus pisorum L. and Callosobruchus maculatus (Fabricius) had a stronger electrophysiological response to the extracts from pea (Pisum sativum L.) and cowpea seeds (Vigna unguiculata L.), respectively. It is likely that this response obeys the female insect’s need to find an appropriate host plant and to obtain the proteins necessary for oogenesis (Paukku and Kotiaho 2008).

Generally, α-copaene and α-pinene are dominant terpenoids in the mixtures of floral volatiles. They are common compounds emitted by plants and have preponderant roles in flower-insect and plant–insect interactions (Ramya et al. 2020). However, in our case, the only compound found in larger quantity was α-copaene, while α-pinene had interesting results during the bioassays with individual synthetic compounds; we observed that male and female R. nigerrimus were attracted to both compounds. Regarding 1-octen-3-ol, there is evidence that it is an attractant for several insect species and is of paramount importance in plant–insect and plant-plant interactions (Chen et al 2019; Ramoni et al. 2001). This was confirmed during the bioassays in which both sexes of R. nigerrimus were attracted. Espadas-Pinacho et al. (2021) found volatiles induced by the damage caused by R. nigerrimus when it feeds on the soybean plant, this could be similar with respect to the weevil-soybean flower interaction, where the flowers probably emit induced volatiles when the weevil feeds of them, however this hypothesis needs to be tested.

In conclusion, we have demonstrated, through bioassays, that the volatiles emitted by flowers of two soybean varieties attract R. nigerrimus. We have also demonstrated that R. nigerrimus uses the sense of smell by means of its antennae to detect these volatiles. A blend of eight compounds attracted R. nigerrimus under laboratory conditions. These results provide the basis for obtaining an attractant that can be used in the field for management of R. nigerrimus.