Introduction

The use of improved crop varieties is one of the key practices towards an efficient integrated pest management programme against herbivorous pests and pathogens (Acquaah, 2004; Liebman & Davis, 2009). One focus of plant breeding programmes is generally to obtain hybrids that can withstand extreme conditions of cold or drought or that have resistance to pathogenic organisms or insect pests (Enders & Begcy, 2021; Kumar et al., 2020; Kutka, 2011; Lee et al., 2015).

Maize (Zea mays L.) is a versatile multi-purpose crop, used as a feed globally and important food crop, especially in sub-Saharan Africa and Latin America (Erenstein et al., 2022). It is also one of the most important cereal crops in Nigeria grown for human consumption, animal feed and several industrial uses (Abdulraham & Kolawole, 2006; Iken & Amusa, 2004; Odeyemi et al., 2020). Nigeria is the largest producer of maize in Africa with a production of 10.2 million metric tons on 4.8 million hectares of land, representing 0.42% of the total global production (FAO, 2016; Kamara et al., 2020). In maize, open-pollinated (OP) varieties are traditionally used varieties that inter-pollinate freely during seed production, resulting in heterogeneous varieties.

The OP varieties have broad genetic bases selected by the environment and farmers over many generations, which help to maintain moderate stress resistance and yield characteristics (Dávila-Flores et al., 2013; Lima et al., 2022). Thus, OP varieties are valuable genetic resources for breeding programmes (Costa et al., 2018, 2020; Dos Santos et al., 2020; Lima et al., 2022). In contrast to OP varieties, hybrid maize varieties are cultivated varieties that result from the fertilization of one maize plant by another genetically unrelated plant through controlled cross-pollination (FAO, 2016). The quality of hybrid maize depends greatly on methods of field production, both in adherence to quality assurance standards and implementation of appropriate agronomic management (Bedő & Barnabás, 2013; Karim et al., 2018; MacRobert et al., 2014). Maize hybrid seeds provide farmers with varieties that have improved specific traits, such as high yield potential and unique trait combinations to counter diseases and adverse growing conditions (MacRobert et al., 2014), for example, increased resistance to insect attack (Barry et al., 1992; de Lange et al., 2014) and pathogens (Wang et al., 2008).

The fall armyworm, Spodoptera frugiperda (J.E. Smith) is a major phytophagous pest of crops endemic to tropical and subtropical regions of the Americas (Sparks, 1979; Johnson, 1987; Nagoshi et al., 2012). It is a migrant pest with a wide host range causing great economic loss whenever present. The pest can feed on over 350 host plants belonging to 76 families, with the Poaceae being the most preferred host (Montezano et al., 2018). Since S. frugiperda invaded Africa (Goergen et al., 2016), it has been reported in more than 45 African countries, affecting maize production (Kasoma et al., 2021; Edosa and Dinka, 2021) with over 50% of maize field losses (Abro et al., 2021; De Groote et al., 2020). In Nigeria, S. frugiperda has established itself in three agroecological zones, causing extensive damage to maize farms, particularly in the humid forest agroecological zone in southwestern Nigeria (Ojumoola & Omoloye, 2023), causing between 50–80% damage in many maize fields (Odeyemi et al., 2020).

One potential approach to S. frugiperda management is to explore the role of maize varieties in influencing the attraction and acceptance behaviour of S. frugiperda. Host plants play an important role in the chemical ecology and behaviour of S. frugiperda because the insects respond to chemical cues for oviposition and larval feeding (Guo et al., 2021; Sotelo-Cardona et al., 2021; Zhang et al., 2023). In general, different maize varieties, whether hybrid or OP, vary in their biochemical compositions, which make them more or less suitable for S. frugiperda (Yang et al., 2023). For example, some maize varieties produce higher levels of secondary metabolites such as terpenoids (Prasanna et al., 2021), which have been shown to have insecticidal properties, while other varieties produce compounds, including, (Z)-3-hexenyl acetate, β-linalool, indole, (E)-β-farnesene and sesquiterpenes ylangene/( +)-cycloisosativene that elicited antennal responses of S. frugiperda females (Pinto-Zevallos et al., 2016). In addition, physical characteristics of host plants, such as trichomes, wax amount, thickness and toughness of leaves and secondary toxic metabolites influence host-plant selection behaviour (Gatehouse, 2002). Differential preference of S. frugiperda was reported in maize cultivars due to differences in cuticular lipids (Yang et al., 1993a), presence of wax materials on the leaf surface (Yang et al., 1993b) and anti-feedant and anti-repellent properties (Tiwari, 2022).

Several studies have investigated the oviposition preference of S. frugiperda to identify which host plant species are either resistant or susceptible to the insect pest (Ba et al., 2020; Guo et al., 2021; Sisay et al., 2023; Tiwari, 2022). Only a few studies evaluated the oviposition preference of S. frugiperda on different maize varieties (He et al., 2021; Zhang et al., 2023). The use of maize varieties that are less preferred by S. frugiperda could reduce the level of infestation in the field. Whether S. frugiperda females show any oviposition preference towards specific hybrid or OP maize varieties is largely unknown. In addition, little is known about the larval attraction of S. frugiperda to different maize plant varieties, particularly in Nigeria (Odeyemi et al., 2020).

To evaluate whether S. frugiperda shows differential attraction and oviposition behaviours towards different maize varieties, we investigated the oviposition preference of the locally collected S. frugiperda on six maize varieties, three of which were hybrid and three were OP varieties. To determine whether the number of egg masses laid by S. frugiperda female increased or decreased over the oviposition period, we checked for variability in the timing of egg-laying on the maize varieties, by conducting oviposition assay over three days in both multiple-choice and no-choice experiments. Finally, we determined whether the S. frugiperda larvae were attracted differentially towards the odors from the different maize plant varieties.

Materials and methods

Study location

The experiment was conducted in the Entomology Research Laboratory and Insect Chemical Ecology Laboratory, Department of Crop Protection and Environmental Biology, University of Ibadan, Nigeria, under ambient conditions of 27 ± 1 °C and 65 ± 5% RH. Greenhouse conditions were 26 ± 2 °C, 70 ± 5% RH and a photoperiod of 14:10 (L:D).

Source of maize seeds and planting

Three hybrid maize varieties (referred to in italics in the remainder of this manuscript), DEKAIB, 30Y87 and P3966W, and three OP varieties (referred to in bold), LMFP, SWAN1 and V9928, were used for experiments. DEKALB was obtained from the International Institute of Tropical Agriculture; V9928, LMFP and SWAN1 were obtained from the Institute of Agriculture Research and Training, Moor Plantation, Apata Ibadan, while 30Y87 and P3966W were obtained from Corteva Agriscience. These varieties were selected because they are cultivated by Nigerian farmers, but so far have never been tested for attraction and oviposition by S. frugiperda, and because the varieties have similar life cycle durations of between, 75–85 days. Two seeds from each variety were sown in a 10 kg pot filled with heat-sterilized loamy soil. The potted plants were grown and maintained in the green house and each variety was planted in six pots. To have a sufficient number of plants for the experiments, the plants were sown every week. Seven days after each sowing, maize plants in each pot were thinned to one. For the experiments, 14-day-old potted maize plants were used, which is a critical period of infestation and foliar feeding by the pest (Kamweru et al., 2022).

Insect collection and rearing

Spodoptera frugiperda larvae were collected from naturally infested maize farms at Sasha-Ajibode (latitude 7o 281 37.70688 E, 30 541 N) and Elekuru-Akinyele areas (70 36125.54092,30 491 N) of Ibadan, Southwestern Nigeria and reared at the Entomology Research Laboratory, University of Ibadan, Nigeria under ambient conditions of 27 ± 1 °C and 65 ± 5% RH and a photoperiod of 14:10 (L:D). The larvae were separated into a transparent sauce cup (EEZEE, Nigeria; 4 cm × 3 cm; 40 mL) with one larva per cup, which were fed daily with fresh maize leaves of the SWAN 1 variety and covered with a lid until pupation. The pupae were placed into clean vials lined with moist tissue paper until adult emergence. The adult insects were kept in the same vials and fed with a 10% sugar solution (Akinbuluma et al., 2024; Marri et al., 2023).

For mating to occur, 2–3 day old virgin adults (Luo-Yan et al., 2022) were paired in mating cups (AVT Plastics, 500 mL, one pair/cup) with a sauce cup (4 cm × 3 cm; 40 mL) filled with cotton wool soaked in a 10% sugar solution (Akinbuluma et al., 2024; Marri et al., 2023) and covered with a muslin cloth. Eggs were collected daily and newly emerged larvae were individually placed in cups and fed with fresh maize leaves of the SWAN 1 variety (approx. 1 g/cup) for 7–12 days. For the experiments, 2–4 day-old mated adult females and the 7–12 day-old larvae of the new generations were used.

Oviposition assays

Oviposition preference and performance of S. frugiperda females were determined in multiple-choice and no-choice experiments inside cages enclosed in the same green house as described above, using intact plants from the six maize varieties. In the multiple choice test, six pots were properly labelled and placed in a mesh cage (40 × 35 × 50 cm), whereby each pot contained one 14-day-old plant of one maize variety. The potted plants were arranged inside the cage in a completely randomized design (CRD). One 2–4 day-old mated female was released into the cage for 72 h. The multiple choice test was conducted in four replicates. Each replicate contained a different batch of the six plants (of the same age) and was set up separately at one-week interval. The number of egg masses on the maize plants inside the cage was counted and recorded every 24 h for a period of 72 h.

In the no-choice experiment, we placed one pot containing one 14 day-old plant of each variety within similar mesh cages (40 × 35 × 50 cm) together with one 2–4 day-old mated female. The no-choice experiment thus had six treatments with 4 replications for each variety. The number of egg masses on the maize plants was counted and recorded every 24 h for a period of 72 h. Egg masses were carefully counted and recorded from outside the cage such that female insects were not disturbed. All data were collected between 13:00 – 16:00 daily and the bioassays were conducted between March and June 2022.

Responses of Spodoptera frugiperda larvae to intact plants

To determine the orientation responses of S. frugiperda larvae to intact maize plants, we chose to test the two most preferred varieties and two least preferred varieties based on the oviposition preference of S. frugiperda females from the experiment above. The experiment was conducted in a Y-tube olfactometer, as described in Akinbuluma and Chinaka (2023) with some modifications, and summarized here. The experimental arena consisted of a horizontal Pyrex glass Y-tube (10 mm i.d; stem 85 mm; arms 75 mm at a 60° angle to the stem). Air from a field pump was passed through activated charcoal and humidified with double distilled water. The airflow was split into two halves. One half was passed through a glass chamber with a tightly sealed lid, enclosing a potted plant (test) and into one arm of the olfactometer, while the other half of the airflow was passed through an empty glass chamber (control) at the same regulated flow rate of 60 mL/min. A vacuum line was connected to a mini pump powered by a rechargeable battery pulled air through the two arms of the Y-tube. Fourth instar S. frugiperda larvae were placed in the bioassay room for at least 12 h before the experiment to acclimatize them to the room conditions. The larvae were also starved for 3 h before the experiment by removing the larvae from their feed container to a small empty rearing cup (1 larva/cup). Twelve larvae were individually assayed in random order on the plant varieties such that 1 larva was assessed 4 times on the four maize varieties. Larval choice was recorded after spending 3 min in any of the arms of the Y-tube. Insects that failed to choose an arm within 10 min were recorded as non-responders and were not included in the analysis.

After testing on each maize variety, the olfactometer set-up, glass chambers and connections were wiped with 70% ethanol and dried to avoid odor contamination between consecutive bioassays and the odour source positions were also exchanged. After an interval of 15 min, the next larva was tested. Bioassays were always conducted between 13:00 and 18:00 h. This study was conducted between May and July 2022.

Statistics and data analysis

The multiple-choice and no-choice oviposition experiments were analyzed separately. The number of egg masses was modelled with generalized linear models (GLM, McCullagh & Nelder, 1989) using Poisson regression, because our count data clearly showed a Poisson distribution. Both models contained maize variety and duration of oviposition as well as their interaction as explanatory variables. In addition, the position of the egg masses in the multiple choice experiment was determined using a separate model, containing position, hours and their interaction. Finally, we investigated if there were differences between the Hybrid varieties (DEKAIB, 30Y87 and P3966W) and OP varieties (LMFP, SWAN1 and V9928) with a model that contained the varietal type and hour, as well as their interaction as explanatory variables (Table 1). The Aikaike Information Criterum was used to evaluate the models, but we did not change the initial model specifications because the results were clear and well interpretable. Posthoc comparisons and testing of the trends of egg laying over time were done on marginal means (Lenth, 2023) and the result was displayed with the compact letter display (Piepho, 2004), with p = 0.05 as significance level and Tukey adjustment for multiple testing. The olfactometer experiment data were analyzed using chi-square tests. All statistical analyses were performed in R (R Core Team, 2022).

Table 1 Overview of the explanatory factors in models used. In all models, the response was the number of egg masses laid by female Spodoptera frugiperda
Full size table

Results

Oviposition preference of adult S. frugiperda female

When we checked the number of egg masses laid by S. frugiperda female on the different maize varieties in the multiple choice experiment, we found a significant difference between the maize varieties (p < 0.05, n = 72, df = 64), but no significant interaction between hours of oviposition and maize variety.

Significantly fewer egg masses were laid over the 72 h on P3966W and LMFP but not on DEKAIB, 30Y87, SWAN 1 and V9928 (Fig. 1A). However, after 24 h of oviposition, the number of egg masses found on DEKAIB was significantly lower than those of P3966W, LMFP and V9928 varieties (Fig. 1A). Similarly, in the no-choice experiment, we found a significant difference between the maize varieties but no interaction with hours of oviposition (p < 0.05, n = 72, df = 64). Across 72 h oviposition period, we observed significantly fewer egg masses (p < 0.05) of S. frugiperda on maize varieties, LMFP and SWAN 1, but not on DEKAIB, 30Y87 and P3966W (Fig. 1B). However, there was no significant difference in the number of egg masses laid across all maize varieties after 24 h of oviposition (Fig. 1B).

Fig. 1
figure 1

Mean number of egg masses (± SEM, n = 4) of Spodoptera frugiperda female. Line plots show number of egg masses on DEKAIB, 30Y87, P3966W (hybrid maize varieties), LMFP, SWAN 1 and V9928 (open-pollinated varieties) over 72 h period on maize varieties in (A) multiple choice experiment and (B) no-choice experiments. ⃰ = significant difference between mean egg masses over 72 h of oviposition; ns = no significant differences between egg masses between the three time points (24, 48 and 72 h) of oviposition. Different letters under the error bars in a and b show that mean egg numbers between varieties are significantly different after 24 h in multiple-choice and no-choice experiments, respectively. C) Cumulative number of egg masses after 72 h of oviposition in multiple-choice experiment and (D) cumulative number of egg masses in no-choice experiments. Boxes represent the lower and upper quartiles of egg masses on each maize, whiskers indicate the minimum and maximum values, the middle line represents the median. Different letters above the bars in c and d indicate significant differences among maize varieties (p < 0.05)

Full size image

When we combined the egg masses over time for each maize variety in the multiple choice experiment (Fig. 1B), we found that DEKAIB and 30Y87 had significantly lower egg mass numbers than varieties, LMFP, SWAN 1 and V9928 (p < 0.05). However, similar numbers of egg mass were found on 30Y87 and P3966W (p > 0.05). Similarly, in the no-choice experiment, we found significantly lower egg mass numbers on DEKAIB and 30Y87 than on LMFP maize varieties (p < 0.05) (Fig. 1D).

When combining all three hybrid varieties and the OP varieties, egg mass counts only significantly differed between 24 and 72 h in the multiple choice (Fig. 2A), with no significant difference in the number of eggs laid over 72 h in the no-choice experiment (Fig. 2A and B). In the OP varieties, we found significantly lower egg masses on the hybrid varieties after 72 h of oviposition than after 24 h both in the multiple-choice and the no-choice test (p < 0.05). After 24 h of oviposition, significantly fewer egg masses were laid by S. frugiperda female on hybrid maize than on OP varieties in both experiments (p < 0.05) (Fig. 2A and B). When the cumulative egg mass counts were compared, egg masses on the hybrid varieties were significantly lower than those on the OP varieties in both multiple choice and no-choice experiments (p < 0.05) (Fig. 2 C, D).

Fig. 2
figure 2

Mean number of egg masses (± SEM, n = 4) of Spodoptera frugiperda female laid on two types of maize varieties. Line plots show the number of egg masses on hybrid and OP maize varieties over a 72 h period in (A) multiple-choice experiment and (B) no-choice experiments. Different letters on line show significant difference between mean egg masses laid over time. ⃰ In between the error bars in A and B show that mean egg numbers are significantly different after 24 h between hybrid and OP maize varieties in multiple-choice (A) and no-choice experiment (B), respectively. (C) Mean cumulative number of egg masses over 72 h oviposition period in the multiple-choice experiment and (D) in the no-choice experiments. Boxes represent the lower and upper quartiles of egg masses on each maize type; whiskers indicate the minimum and maximum values, the middle line represents the median. Different letters above the bars in C and D indicate significant differences between hybrid and OP varieties (p < 0.05)

Full size image

Response of Spodoptera frugiperda larvae to intact maize varieties

Olfactometer experiment with intact plants of the four maize varieties showed that significantly more S. frugiperda larvae (χ2 = 8.33; p < 0.05) chose the LMFP maize variety over clean air (Fig. 3). Larvae did not show a preference for the other three tested maize varieties relative to clean air (χ2 = 0.33; p > 0.05), 30Y87 (χ2 = 0.33; p > 0.05) and DEKAIB (χ2 = 0.33; p > 0.05).

Fig. 3
figure 3

Orientation responses of Spodoptera frugiperda larvae to odours from intact plants and clean air. Bar charts show the preferences of S. frugiperda larvae for odours from intact four maize varieties and clean air presented in a Y-tube olfactometer. Hybrid maize varieties (DEKAIB, 30Y87) are in italics, Open-pollinated maize varieties (V9928, LMFP) are in bold. Significant means were separated using the Chi-square (χ2) test (p < 0.05, n = 12), ns = not significant

Full size image

Discussion

In this study, we investigated the oviposition and attraction of S. frugiperda to different maize varieties, belonging to hybrid and OP varieties, and found i) variation in the number of egg masses laid on maize varieties at different time periods, ii) ovipositional preferences of S. frugiperda females for SWAN 1, V9928 and LMFP compared to DEKAIB and 30Y87 in both multiple-choice and no-choice experiments, iii) significant attraction responses of S. frugiperda larvae towards LMFP maize variety compared to clean air, while this was not the case for V9928, DEKAIB and 30Y87. Overall, our results show that DEKAIB and 30Y87 were less accepted and less attractive to S. frugiperda than the other maize varieties.

Our results showed significant variation in the number of egg masses laid by S. frugiperda females on the different maize varieties, both in the multiple-choice and no-choice tests. The consistently significantly lower number of egg masses on DEKAIB suggests that the variety was the least acceptable for oviposition compared to all other tested varieties. Differential oviposition preference of S. frugiperda female for some maize varieties has been shown by others as well (He et al., 2021; Zhang et al., 2023) and could be due to differences in plant nutrients in maize host plants, which is important for the development of offspring (Tiwari, 2022; Yang et al., 2023).

Attack on maize by S. frugiperda can occur from the germination phase to the reproductive phase of the plants, depending on variety and cultural practices (Marcos et al., 2023). However, in our study we used 14-day old plants, because Anjorin et al. (2022) tested infestation rates of S. frugiperda on maize plants at 7, 14 and 21 days and found that plants at 14 days were most susceptible to infestation by S. frugiperda.

For a proper timing of management of S. frugiperda eggs on the maize host plants, it is important to know when egg masses are laid and how this changes over time. Spodoptera frugiperda females have been found to lay reduced numbers of egg batches over time (Russianzi et al., 2021). Therefore, we counted the number of egg masses over three days. Among the varieties tested, we found significant reductions in the number of egg masses laid over time. For example, in the no-choice test, significantly more egg batches were laid on the LFMP and SWAN1 maize varieties after 24 h than after 72 h. Also, in the first 24 h, significantly less egg batches were laid on the hybrid maize varieties (Fig. 2). We also showed that significantly less egg masses were found on the combined hybrid maize varieties (DEKAIB, 30Y87 and P3966W) than on the OP varieties (LMFP, V9928 and SWAN1). Even when considered separately, the least number of egg masses were laid on the hybrid varieties compared to the OP varieties (Fig. 1). Our results thus indicate that hybrid maize varieties are less favourable to S. frugiperda for oviposition than OP varieties. An earlier report also showed that S. frugiperda did not infest hybrid maize varieties as much as OP varieties in both the lowland and high-altitude land in Kenya (Mutyambai et al., 2022). However, Zhang et al. (2023) found significant oviposition preference of S. frugiperda females for some hybrid maize varieties compared to OP varieties. Thus, differential oviposition cannot always be explained by maize pollination type. Nevertheless, to reduce the infestation of maize by S. frugiperda in Nigeria, our results suggest that farmers should use hybrid maize varieties instead of OP varieties for establishing maize fields.

We would like to point out that in all our oviposition experiments, we found more S. frugiperda egg masses on the cage surfaces than the maize varieties (Supplementary Fig. SI). This is a common phenomenon that other studies also found (Barros et al., 2010; Guo et al., 2021; Meagher et al., 2011; Rojas et al., 2003; Sotelo-Cardona et al., 2021; Volp et al., 2022). This may suggest that some maize plants’ volatiles have repellent effects on female insects, preventing them from ovipositing on the plants. Rojas et al. (2003) and Barros et al. (2010) reported that S. frugiperda females oviposited more on corrugated surfaces rather than on surfaces treated with some host plant extracts. Thus, our study corroborates the finding that S. frugiperda females lay more eggs on oviposition cage walls away from nearby plants than on plants, which could indicate the repellent effects of some plant-derived volatiles (Sisay et al., 2023).

The olfactometer bioassay with S. frugiperda larvae showed only a larval preference towards the LMFP maize variety relative to clean air, and not to the other varieties. Possibly, this variety releases volatiles that are attractive to larvae. Although lepidopteran larvae possess limited mobility, they can still move between plants by way of so-called ‘ballooning’ (e.g. Sokame et al., 2020), and also orient towards or away from host plants (Liu et al., 2023). Background volatile compounds have been found to modify larval behaviour in S. frugiperda on some maize varieties (Yactayo-Chang et al., 2021). Interestingly, S. frugiperda larvae were not significantly attracted to the hybrid DEKAIB and 30Y87, on which S. frugiperda females oviposited the least.

Conclusion

Our study showed that maize varieties elicit differential oviposition and larval responses in S. frugiperda. Possibly, variation in maize host plant volatiles may partly account for these different responses. Identification of plant volatiles that attract S. frugiperda females could help to reduce the population. Since the use of plant varieties can affect the extent of infestation and damage by insect pests, varieties with less preference for oviposition by the insect will suffer less infestation and damage compared to those with full acceptance for oviposition by the insect. Our study suggests that planting DEKAIB and 30Y87, which are hybrid maize varieties could reduce the level of Spodoptera frugiperda infestation in Nigerian fields. Exploring these avenues further will help to develop integrated pest management (IPM) strategies for this pest.