Abstract

Fall Armyworm (FAW), Spodoptera frugiperda (Lepidoptera: Noctuidae) , is a significant pest causing substantial economic losses in invaded regions, particularly in Sub-Saharan Africa. The prevalent strategy for managing FAW involves insecticide applications, ranging from synthetic to botanical and microbial agents. However, the ecological and economic impacts of these interventions often remain unassessed. This study scrutinizes the ecological and economic viability of two insecticidal treatments: one based on emamectin benzoate and another comprising a combination of Pieris rapae granulovirus (PrGV) and Bacillus thuringiensis subsp. kurstaki (Btk) with varied application timings and frequencies. Both insecticide types were effective in reducing FAW larval populations and the associated crop damage. From an ecological standpoint, the PrGV|Btk treatment enhanced parasitism rates, especially when applied only at the early stage of the crop. However, this approach did not significantly lower crop damage compared to a "targeted" strategy, where insecticide application was contingent upon observed injury levels. Remarkably, the "targeted" strategy led to increased FAW larval parasitism, particularly at the V₇ growth stage of the crop. Economically, the "targeted" insecticide application emerged as both effective and efficient, minimizing the need for multiple sprays and thus recommended for managing FAW infestations while considering cost and ecological balance.

Introduction

In the agricultural tapestry of Sub-Saharan Africa (SSA), maize (Zea mays L.) emerges as an important crop, integral to the sustenance, economic vitality, and cultural heritage of countless commu­ni­ties (Alene et al., 2009; Chivenge et al., 2015; Erenstein et al., 2022). Maize cultivation provides food security for millions, supporting livelihoods of smallholder farmers and contributing significantly to the agricultural economy (Alene et al., 2009). Yet, the path to suc­cessful maize cultiva­tion in SSA is beset with myriad challenges. From soil-related issues to unpredicta­ble climate patterns and the on­slaught of pests (Goergen et al., 2016; Adem et al., 2023; Habte et al., 2023), each factor significantly ham­pers the productivity of maize and, by extension, the food security and economic resilience of SSA. Among the spectrum of biotic threats, the inva­sion by the fall armyworm (FAW), Spodoptera frugi­perda (Lepidoptera: Noctuidae), emerges as a particu­larly daunting challenge.

Originating from the tropical and subtropical regions of the Americas (Spark, 1979), FAW made its unwelcome debut in Af­rica nearly a decade ago (Goergen et al., 2016). Its rapid spread across diverse African land­scapes has been relentless, breaching geographical and climatic boundaries to pose a continuous threat to maize pro­duction across the continent (Koffi et al., 2020a, 2021, 2023a; Niassy et al., 2021). The larvae of FAW are particularly devastating, and causing signif­i­cant damage to maize, which can lead to substantial yield losses (Spark, 1979). The voracious larvae have a wide host range but show a strong prefer­ence for poaceae plants, especially maize which is more vulnera­ble (Signoretti et al., 2012; Montezano et al., 2018). Their ability to cause extensive damage in a short period underscores the urgency of finding effec­tive management strategies.

In response to FAW threat, a diverse range of management strategies has been implemented, from traditional cultural methods to advanced biological and chemical controls (Midega et al., 2018; Akutse et al.,2019; Babendreier et al., 2020; Aniwanou et al., 2021; Otim et al., 2021; Nboyine et al.,2022; Agboyi et al., 2023; Chawanda et al., 2023; Fiaboe et al., 2023a, 2024). Among these strategies, biological control, particularly the recruitment of natural enemies such as parasitoids, stands out as a cornerstone of sustainable pest management (Agbodzavu et al., 2018; Koffi et al., 2020b; 2023a; Abang et al., 2021b; Koffi et al., 2023). These approaches are often advocated for incorporation into Integrated Pest Management (IPM) schemes, which seek to harmonize effective pest mitigation with the preservation of environmental health. Yet, finding a fully sustainable combines remains challenging, due to variation in environmental contexts across ecozones.

A fundamental yet often overlooked element of IPM is the strategic timing and frequency of interventions, which are essential for its success (Tang et al., 2010; McClure et al., 2023). However, the effectiveness of IPM can be greatly improved when pest control measures are synchronized with the life cycle of the pest, which is intrinsically linked to crop phenology (Fiaboe et al., 2023b; Idrees et al., 2023). This approach is not only focused on direct pest control; it also considers the broader ecological context, particularly the trophic interactions between pests, their natural enemies, and the host plants. Herbivores, such as pests, have life cycles that are closely tied to the phenological stages of crops, upon which they feed. In turn, the survival and efficacy of natural enemies, mainly parasitoids, depend on the availability and vulnerability of these herbivores (Ratto et al., 2022; Fiaboe et al., 2023b). By carefully planning pest management activities to coincide with these interconnected life cycles, IPM strategies not only target pests more effectively but also help preserve and support the beneficial organisms that play a vital role in the natural regulation of pest populations.

On the other hand, the economic ramifications of different control strategies against FAW in SSA remain underexplored. Meanwhile, by targeting pest control measures to coincide with key points in the pest and crop life cycles, farmers can optimize the use of resources, reducing the need for frequent or excessive chemical applications (van den Berg et al., 2021; Nboyine et al., 2022). This targeted approach may minimize input costs and maximizes yield potential by ensuring crops are protected at their most vulnerable stages.

This study sought to bridge these gaps by exam­in­ing the efficacy of synthetic emamectin benzoate and microbial bioinsecticides (Pieris rapae Granulosis Vi­rus ^[1] and Bacillus thuringiensis subsp. kurstaki), un­der on-station experiment conditions. The study hypothesized that the timing and frequency of insecticide applications play a critical role in managing FAW larval outbreaks. To test this, the study compared the ecological and economic impacts of various application strategies: early treatment (applying insecticides at the initial growth phase of maize), late treatment (applying at the advanced vegetative stage of maize), preventive treatment (regular applications based on a predetermined schedule), and targeted treatment (applying based on observed damage levels). Additionally, an economic evaluation was carried out to ascertain the cost-benefit ratio of each method, with the hypothesis that an infestation-level-driven approach to insecticide application would strike a balance between effectiveness and economic feasibility. This approach was proposed as a sustainable method for controlling FAW.

Materials and Methods

Study site

The study was conducted at the agricultural re­search facility of the University of Lome, Togo, po­sitioned at 6°22'N latitude and 1°13'E longi­tude, with an elevation of 50 meters. The research site falls within the Coastal Savannah Agro-ecological Zone, which is known for its bimodal distribution of rainfall. The initial rainy phase extends from the mid­dle of March to the end of July, succeeded by a second phase of rainfall that spans from the start of Septem­ber to the middle of November (Fiaboe et al., 2024). Traditionally, maize cul­tivation aligns with two dis­tinct seasons, from April to July and from September to November, benefiting from an average yearly rain­fall ranging between 800 mm and 1100 mm, alongside a mean temperature of 27°C annu­ally. The observed shifts in climate patterns and a continuous need for maize have led to year-round cultivation prac­tices by local farmers, moving away from conventional seasonal cy­cles (Fiaboe et al., 2024). Prior to this study, the exper­imental field was left fallow for a year and then used for maize cultivation. The predominant soil type at the site is classified as ferralsol.

Experimental design and treatments

Two on-station experiments were conducted from April 10 to July 29, 2021, and September 26, 2021, to Jan­uary 14, 2022, respectively. Certified maize seed variety QPM Obatanpa, supplied by the Togolese Insti­tute of Agro­nomic Research (ITRA) was used. This particular variety of maize typically reaches maturity approxi­mately 105 days following plant­ing (Fiaboe et al., 2024). The experi­mental frame­work was struc­tured as a 3x4 factorial de­sign within a randomized block setup, including four replicates.

Planting was orga­nized in a pattern of 0.80 m be­tween rows and 0.20 m between plants, ensuring one plant per hill. Each ex­perimental plot measured 5.6x6 m, with a 3-m buffer alley between both blocks and individual units to pre­vent treatment in­terference. Nutrient management included the application of a complex N₁₅P₁₅K₁₅ fer­ti­lizer, administered at a rate of 250 kg/ha fifteen days fol­lowing seedling emer­gence, sup­plemented by an additional 100 kg of nitro­gen per hec­tare in the form of ammonium nitrate after 30 days. The exper­i­mental plots were maintained fol­lowing standard agricul­tural practices, which encom­passed manual, hu­man-powered weed control.

The insecticidal treatments involved EMACOT 019EC^TM, containing 19.2 g/L of emamectin benzo­ate, and BYPEL 1^®, a blend of Pieris rapae granulo­virus (PrGV) at 10,000 PIB/mg and Bacillus thuringiensis subsp. kurstaki (Btk) at 16,000 IU/mg. Insecticide dos­ages adhered to recommended guidelines, employing a 1 mL/L water so­lution for EMACOT 019 EC^TM and a 1.33 g/L water solu­tion for BYPEL 1^®, ensuring appli­cation volumes of 350 L/ha and 200 L/ha, respec­tively. To ensure optimal effi­cacy, all insecticide appli­cations were carried out in the late afternoon, specifi­cally between 17:00 and 18:00 hrs, using 16-L Ingco backpack sprayers dedicated sepa­rately for the syn­thetic and mi­crobial insecticide treat­ments to prevent cross contamina­tion.

We assessed four distinct application frequen­cies: an "early" treatment involving a single applica­tion at 10 days post-emergence (corresponding to the V₃ growth stage); a "preventive" strategy with appli­ca­tions at 10 days post-emergence followed by addi­tional treatments at the V₅, V₇, and V₉ stages; a "targeted" approach ap­plying insecticide based on vis­ual damage assessments exceeding predefined thresholds (Prasanna et al., 2018); and a "late" in­ter­vention at the V₉ growth stage. Growth stages were determined using the Iowa State University scaling system (Infante et al., 2018), which identi­fies stages from the third collar leaf (V₃) to the nth collar leaf (Vn), along with tasseling (V_T), milk (R₃) stage, and the dry cobs stage.

Infestation and parasitism

Data on infestation were collected, fo­cusing on two primary metrics: the extent of dam­age to each maize plant and the count of FAW larvae present. These ob­serva­tions were rec­orded at four growth stages of the maize plants: V₃, V₅, V₇, and V₉. A de­structive sam­pling method was em­ployed, where five plants from each treatment plot were carefully selected and sev­ered at the base. These plants were then transported to the Entomology and Nematology Laboratory at the School of Agronomy (ESA), University of Lome, for further analysis. The assessment of damage to the plants was conducted using the scoring system de­vel­oped by Davis et al. (1992), which ranges from 0 (in­di­cating no damage) to 9 (indicating the highest level of damage), as detailed in Table S1. Simultane­ously, lar­vae were carefully counted from each plant to gauge the severity of infestation.

For parasitism analysis, observations were ex­tended to include the V_T, R₃, and dry cob stages of maize growth. The larvae collected were individu­ally placed in 100 mL clear plastic cups (Everpack Ghana Ltd, Accra, Ghana) with aera­tion to monitor parasitoid emergence. The cups were maintained at 26±1°C, 75-80% relative humidity, and a 12:12-hour light-dark photoperiod. To track parasitoid emergence, larvae were fed insecticide-free fresh maize leaves daily. The identification of parasitoids emerging from the larvae was conducted by the last two authors of this paper, drawing upon their taxonomical expertise in the species (Koffi et al., 2020b; 2023a).

Yield and economic analysis

Maize cobs were harvested manually from each plant when they reached maturity, 110 days after sowing. After harvesting, the cobs were laid out to dry in the sun for 12 days. After proper drying, the cobs were shelled to separate the grains, which were then weighed. Weight measurements were collected from groups of 10 plants per experimental plot to ensure a representative sample. The weights were scaled up to hectare-level yields, accounting for the plant density used in the experimental design.

The yield values were used in the economic analysis using the following formulas (Mauki et al., 2023) to determine:

The market favored control and PrGV|Btk-treated maize with higher prices compared to those treated with emamectin benzoate, reflecting the premium on ecologically-produced maize over synthetic alterna­tives (Table S2).

Net profit (NP) per hectare was then calculated as:

where TC encompasses the sum of:

• Seed price per hectare
• Manure cost per hectare
• Sowing labor cost per hectare
• Insecticide costs, calculated as the product of the number of applications and the sum of insecticide price, water cost, and labor cost per event.

The economic viability was assessed via Benefit-Cost Ratio (BCR):

A BCR above 1 implies profitability, while below 1 suggests a loss.

Lastly, Return on Investment (ROI) was quantified to determine the investment's profitability, calcu­lated as:

Data analysis

Before performing statistical analyses, data were as­sessed for normality using the Shapiro-Wilk test, and for homogeneity of variances with the Levene's test. Upon satisfying these preliminary conditions, larvae count data were analyzed using a Likelihood Ratio Test (LR test) within a Generalized Linear Model (GLM) framework, incorporating likelihood radio test (LR test) in the log link function. Damage scores, maize grain yield, and economic data were evaluated using a One-Way Analysis of Variance (ANOVA), with "Best Model" serving as the model selection criterion and Mean Squared Error (MSE) as the performance met­ric. Partial Least Squares Discriminant Analysis (PLS-DA) was employed to examine correlations between insecticide application strategies and the collected bi­ological and ecological data.

For all statistical tests, a significance level of P<.05 was established. Tukey's Honestly Significant Dif­ference (HSD) test facilitated post-hoc mean compari­sons, adhering to a significance threshold of P≤.05. Two-sample t-test was used to compare two independent samples, with alternative hypothesis set at Mean 1 − Mean 2 ≠ 0. All analyses were performed using RStudio version 4.1.2 (R Core Team, 2022).

Results

Influence of insecticide treatments on FAW larvae

In the first experiment, FAW larvae counts at the maize V₃ growth stage did not significantly differ among treatment groups (LR-GLM, χ²=0.81, df=4, P>.05; Table 1), a trend that was consistent in the subsequent trial (LR-GLM, χ²=1.20, df=4, P>.05; Table 1). Neither emamectin benzoate nor PrGV|Btk treatments significantly impacted FAW populations at the V₅ stage in both the first (LR-GLM, χ²=9.00, df=4, P>.05; Table 1), and second (LR-GLM, χ²=4.88, df=4, P>.05;Table 1) experiments. However, in the second experiment, a marked decrease in FAW larvae per plant was noted at the V₇ stage, especially in plots receiving preventive or targeted applications of emamectin benzoate (LR-GLM, χ²=14.08, df=4, P<.01; Table 1), and PrGV|Btk (LR-GLM, χ²=25.00, df=4, P<.0001; Table 1) treatments. Similarly, in the second experiments, at the V₉ stage, a significant reduction in FAW larvae was observed in treated plots, particularly those under "preventive" and "targeted" applications of emamectin benzoate (LR-GLM, χ²=22.15, df=4, P<.001; Table 1) and PrGV|Btk (LR-GLM, χ²=11.69, df=4, P<.05; Table 1).

At the V_T stage of maize development, the "Preventive" treatment plots exhibited the lowest counts of FAW larvae, with significant reductions observed in both the initial (LR-GLM, χ²=16.62, df=4, P<.05; Table 1), and subsequent (LR-GLM, χ²=12.80, df=4, P<.05; Table 1) experiments. However, by the R₃ stage, no significasnt differences in larval counts were detected among the various treatment groups in the second trial, regardless of the type of insecticide applied (LR-GLM, χ²=6.82, df=4, P>.05; Table 1). This lack of significant difference persisted in treatments using emamectin benzoate and PrGV|Btk, as evidenced by the FAW larvae counts per maize cob (LR-GLM, χ²=1.93 and χ²=0.85, respectively, df=4, P>.05; Table 1).

Table 1

Crop Stage	Insecticide	Insecticide application regimes					P-value
		Control	Early	Preventive	Targeted	Late
First Experiment
V3	Em. benz.	2.5±0.26a	2.9±0.34a	3.1±0.34a	3.1±0.36a	2.7±0.32a	.851ns
	PrGV|Btk	2.5±0.26a	2.8±0.36a	2.7±0.36a	2.9±0.31a	3.0±0.28a	.948ns
V5	Em. benz.	1.9±0.23a	1.4±0.21a	1.4±0.22a	1.3±0.16a	1.9±0.25a	.285ns
	PrGV|Btk	1.9±0.23a	1.7±0.21a	1.2±0.19a	1.6±0.21a	2.1±0.22a	.174ns
V7	Em. benz.	1.2±0.22a	1.3±0.22a	0.85±0.19a	1.0±0.16a	1.2±0.21a	.619ns
	PrGV|Btk	1.2±0.22a	0.5±0.15a	0.80±0.18a	0.9±0.14a	1.3±0.23a	.072ns
V9	Em. benz.	1.2±0.18a	0.9±0.16a	0.5±0.13a	0.3±0.12a	0.8±0.16a	.089ns
	PrGV|Btk	1.2±0.18a	0.6±0.15a	0.4±0.13a	0.1±0.06a	0.6±0.13a	.069ns
VT	Em. benz.	0.8±0.16a	0.6±0.15a	0.3±0.12a	0.4±0.15a	0.2±0.11a	.053ns
	PrGV|Btk	0.8±0.16c	0.6±0.19bc	0.1±0.06a	0.2±0.11a	0.4±0.10ab	.003**
R3	Em. benz.	0.6±0.11a	0.6±0.15a	0.4±0.13a	0.4±0.13a	0.3±0.12a	.308ns
	PrGV|Btk	0.6±0.11a	0.5±0.18a	0.1±0.08a	0.2±0.12a	0.3±0.10a	.056ns
Cobs	Em. benz.	0.6±0.23a	0.7±0.22a	0.3±0.10a	0.3±0.12a	0.4±0.15a	.505ns
	PrGV|Btk	0.6±0.23a	0.7±0.17a	0.3±0.14a	0.2±0.12a	0.5±0.13a	.314ns
Second Experiment
V3	Em. benz.	2.4±0.31a	2.5±0.31a	2.8±0.31a	2.6±0.41a	2.6±0.43a	.933ns
	PrGV|Btk	2.4±0.31a	2.3±0.36a	2.4±0.33a	2.9±0.34a	2.6±0.43a	.933ns
V5	Em. benz.	2.2±0.27a	1.8±0.20a	1.8±0.26a	1.6±0.27a	2.2±0.21a	.616ns
	PrGV|Btk	2.2±0.27a	2.1±0.22a	1.6±0.24a	1.5±0.21a	2.2±0.21a	.616ns
V7	Em. benz.	1.3±0.19b	1.1±0.23ab	0.5±0.11a	0.5±0.16ab	1.2±0.18ab	.007**
	PrGV|Btk	1.3±0.19b	0.6±0.16b	0.4±0.13a	0.4±0.15a	1.4±0.19b	<.001***
V9	Em. benz.	1.0±0.17bc	1.1±0.21c	0.1±0.07a	0.9±0.22bc	0.6±0.14b	<.001***
	PrGV|Btk	1.0±0.17c	0.5±0.11ab	0.2±0.12a	0.4±0.15ab	0.6±0.15b	.001***
VT	Em. benz.	0.6±0.27a	0.6±0.26a	0.4±0.19a	0.5±0.19a	0.5±0.18a	.867ns
	PrGV|Btk	0.6±0.27bc	0.9±0.27c	0.1±0.07a	0.2±0.11ab	0.6±0.21bc	<.001***
R3	Em. benz.	0.4±0.19a	0.5±0.18a	0.2±0.11a	0.3±0.12a	0.2±0.12a	.449ns
	PrGV|Btk	0.4±0.19a	0.3±0.15a	0.3±0.12a	0.1±0.07a	0.2±0.08a	.544ns
Cobs	Em. benz.	0.4±0.19a	0.5±0.18a	0.2±0.11a	0.3±0.12a	0.2±0.12a	.748ns
	PrGV|Btk	0.4±0.19a	0.3±0.15a	0.3±0.12a	0.1±0.07a	0.2±0.08a	.932ns

This table details the mean counts of Spodoptera frugiperda larvae per maize plant recorded at key growth stages during two on-station experiments. The first experiment was conducted from April 10 to July 29, 2021. Growth stages V3, V5, V7, and V9 correspond to vegetative phases inspected on April 20, April 30, May 10, and May 20, respectively. The V_T stage (tasseling) was assessed on June 19, while the R3 stage (milk) was evaluated on July 9. The second experiment spanned from September 26, 2021, to January 14, 2022. Vegetative stages V3, V5, V7, and V9 were examined on October 6, October 16, October 26, and November 5, 2021, respectively. The tasseling phase (V_T) was observed on November 25, and the milk stage (R3) was inspected on December 15. The "Cobs" category represents the assessment of larval counts on dry maize cobs conducted on January 14, 2022. PrGV denotes Pieris rapae Granulovirus, and Btk represents Bacillus thuringiensis subsp. kurstaki. Displayed values indicate average larval counts alongside the standard error (SE). Data analysis was performed using ANOVA (n=5, selecting the best model based on mean squared error (MSE), with Tukey's HSD test for the differentiation of means. Insecticide application events are marked with brown icons. The statistical significance of the results is denoted by asterisks, with * indicating P < .05, ** indicating P < .01, and *** indicating P < .001; ns indicates no significant difference.

​

Effect of insecticide treatments on maize plant damage

Figure 1

​ ​

Effect of insecticide treatments on Spodoptera frugiperda-related damage in maize at various growth stages during the first on-station experiment. The bars display the mean±SE damage coefficients, which quantify the extent of damage inflicted by FAW larvae on maize plants at different developmental stages. The damage coefficients are normalized by dividing each plant's damage score (Davis et al., 1992) by the highest score recorded at a particular growth stage. The abbreviations Em., PrGV, and Btk denote emamectin benzoate, Pieris rapae Granulovirus, and Bacil­lus thuringiensis subsp. kurstaki, respectively. The analysis was conducted using One-Way Analysis of Variance (ANOVA) with a significance level set at α = 5%. The statistical significance of the results is denoted by asterisks, with * indicating P < .05, ** indicating P .01, and *** indicating P < .001.

​

The observations indicate that the extent of damage inflicted by FAW on maize plants at the V₃ growth stage was uniformly distributed across all treatment plots in both on-station experiments (One-Way ANOVA, F_4,175=0.44 and F_4,175=0.41, respectively, P>.05; Figure 1 & 2). However, the use of emamectin benzoate in the first experiment resulted in significantly reduced damage at the V₅ stage (One-Way ANOVA, F_4,95=4.01, P<.01; Figure 1), an effect not observed with PrGV|Btk treatments (One-Way ANOVA, F_4,175=0.43, P>.05; Figure 1).

Conversely, at the V₇ stage, in both experiments, insecticide applications led to significant differences in damage levels (One-Way ANOVA, F_4,175=33.41 and F_4,175 = 16.84, respectively, P<.0001; Figure 1 & 2), with the lowest damage noted in early, preventive, and targeted treatment plots. This pattern continued at the V₉ stage, where the lowest damage was recorded in preventive treatment plots that underwent four insecticide applications, in both experiments (One-Way ANOVA, F_4,175=37.51 and F_4,175=31.82, respectively, P<.0001; Figure 1 & 2).

Figure 2

​ ​

Effect of Insecticide treatments on Spodoptera frugiperda-related damage in maize at various growth stages during the second on-station experiment. The bars display the mean±SE damage coefficients, which quantify the extent of damage inflicted by FAW larvae on maize plants at different developmental stages. The damage coefficients are normalized by dividing each plant's damage score (Davis et al., 1992) by the highest score recorded at a particular growth stage. The abbreviations Em., PrGV, and Btk denote emamectin benzoate, Pieris rapae Granulovirus, and Bacil­lus thuringiensis subsp. kurstaki, respectively. The analysis was conducted using One-Way Analysis of Variance (ANOVA) with a significance level set at α = 5%. The significance of the results is denoted by asterisks, with * indicating P < .05, ** indicating P < .01, and *** indicating P < .001, highlighting the statistical relevance of the observed differ­ences.

​

Impact of insecticides on FAW parasitoids

The study reported the occurrence of parasitoids such as Chelonus bifoveolatus Szpligeti (Hymenoptera: Braconidae), Chelonus insularis Cresson (Hymenoptera: Braconidae), Coccygidium luteum (Brullé) (Hymenoptera: Braconidae), and Cotesia sp. (Hymenoptera: Braconidae) in FAW larvae that were incubated for observation (Table S2). The rates of parasitism unveiled significant variation at different growth stages of maize plants (Figure 3). showed the highest parasitism rates during the initial stages of the crop (V₃ and V₅; Figure 3). Remarkably, in the first experiment, plots that received early insecticide treatments demonstrated the highest rates of parasitism, particularly peaking at the V₉ and V_T stages, irrespective of the type of insecticide used (Figure 3). Similarly, in both experiments, plots with targeted insecticide treatments, especially when PrGV|Btk was applied, showed a parasitism rate peak between the V₇ and V₉ stages. For plots where emamectin benzoate was applied preventively, a notable peak in parasitism rates was observed at the V_T stage of maize growth.

Figure 3

​ ​

Effect of insecticide treatments on parasitism rates of FAW larvae. This figure illustrates the changes in parasitism rates of FAW larvae among various treatment plots throughout the maize plant growth stages. The lines trace the parasitism dynamics, reflecting how different insecticide applications influence the interaction between FAW larvae and their parasitoids over time. For specifics on when insecticides were applied for each treatment group, see Figure S1. The abbreviations PrGV and Btk stand for Pieris rapae Granulovirus and Bacillus thuringiensis subsp. kurstaki respectively, indicating the types of bioinsecticides evaluated in this study.

​

PLS-DA on insecticide strategy variables for FAW management

Partial Least Squares Discriminant Analysis (PLS-DA) is used here to illustrate the biological and ecological variables associated with different insecticide application timings in the context of managing FAW infestation. The plot combines the first (wc1) and second (wc2) latent variables, which together account for 34% of the variance explained by the PLS-DA model (Figure 4).

The results suggest that "Late" and "Control" insecticide application treatments have similar profiles (Figure 4). Conversely, "Preventive" and "Targeted" treatments are associated with another set of variables, with both treatment strategies appearing to be distinct from the "Early" treatment. Remarkably, the biplot reveals that variable such as parasitism of FAW larvae at V₇ and “cobs" stages are closely aligned with the "Targeted" treatment strategy, while maize leaf damage at V₅, V₇, and V₉, and larval incidence at V₅ and V₇ maize stage was associated with the control treatment (Figure 4).

Figure 4

​ ​

PLS-DA Biplot of Insecticide Application Timings on FAW Management. This biplot illustrates the associations between different insecticide application strategies ('Late', 'Control', 'Preventive', 'Targeted', and 'Early') and key biological and ecological variables in the context of FAW control. The axes represent the first (wc1) and second (wc2) latent variables of the PLS-DA model, explaining 34% of the total variance

​

Economic efficiency of insecticide control of FAW

According to the economic analyses, the highest gross revenues were consistently observed in the "Preven­tive" and "Targeted," followed by "Early," and "Late" treatments, both initially and in subsequent experi­ments (One-Way ANOVA, F_4,35=38.95 and F_4,35=242.08, respectively, P<.0001; Figure 5A& B; Table S3). However, no statistical difference was found between the "targeted" and "preventive" appli­cations concerning gross revenues in either experi­ment (t-Test, t=0.613 and t=-0.169, df=6, P>.05; Figure 5A & B).

Interestingly, during the initial experiment, both "targeted" and "Preventive" insecticide application models demonstrated significantly higher revenues for PrGV|Btk treatment over emamectin benzoate (t -Test, t=-5.33 and t=-7.57, respectively, df=6, P<.001; Figure 5A). Similarly, in the second experiment, higher revenues were obtained from PrGV|Btk treat­ment plots when applied preventively (t-Test, t=-8.62, df=6, P<.001; Figure 5B), or targeted (t-Test, t=-69.53, df=6, P<.0001; Figure 5B).

Figure 5

​ ​

Effects of insecticide on the maize grain yield. Bars depict the mean±SE maize grain yield. The bars represent the average maize grain yield (±SE) across different treatment plots, with yields extrapolated from measurements of 10 plants per plot to tons per hectare (t/ha). The abbreviations PrGV and Btk represent Pieris rapae Granulovirus and Baccilus thuringiensis subsp. kurstaki respectively, indicating the specific bioinsecticides evaluated. Yield data were analyzed using One-Way Analysis of Variance (ANOVA), with a 5% significance threshold (α = .05). The effect of each insecticide treatment on yield, relative to the untreated control, is highlighted by asterisks: * for P < .05, ** for P .01, and *** for P < .001, signifying varying levels of statistical significance. The notation 'ns' indicates no significant difference from the control.

​

Comparable trends were noted in the BCR values, with all ratios surpassing 1 (Figure 6C&D). Specifi­cally, both "Targeted" and "Preventive" applications consistently demonstrated significantly higher BCRs compared to other treatments, including "Early," "Late," and the control, across both experiments (One-Way ANOVA, F_4,35=24.96 and F_4,35=22.46, respectively, P<.0001; Figure 6C&D; Table S3). Moreover, consist­ently higher values were associated with PrGV|Btk appli­cation (Figure 6C&D, P<.0001). Notably, the BCR of the "Late" application was either similar to or signifi­cantly lower than the control treatment, irre­spective of the insecticides used (One-Way ANOVA, F4,15=92.68 and F4,15=92.68, respectively, P<.0001; Figure 6C; Table S3).

Figure 6

​ ​

Economic analysis of insecticide application timing and frequency. The half boxes represent the interquartile range (25-75% of the data), while the whiskers delineate the outliers. Panels (A) and (B) display the gross revenues from the first and second experiments, respectively. Panels (C) and (D) showcase the benefit-cost ratio (BCR). The data on gross revenues and BCR underwent analysis via One-Way Analysis of Variance (ANOVA) with a significance level set at α = 5%. Additionally, a t-Test was utilized to compare two samples. The small letters positioned above the whiskers denote differences between emamectin benzoate-treated maize plants, whereas capital letters signify differences between PrGV|Btk treated maize plants.

​

Discussion

Our research assessed the effectiveness, eco­logi­cal impact, and economic viability of an emamectin benzoate-based insecticide and a microbial bioinsecticide comprising Pieris rapae Granulovirus and Bacillus thuringiensis subsp. kurstaki (PrGV|Btk) in controlling fall armyworm (FAW) larvae. Our findings demonstrated that both insec­ticide formulations effectively mitigated FAW larval infestation across two distinct on-station exper­imental setups, aligning with previous studies (Deshmukh et al., 2020; Nboyine et al., 2022; Agboyi et al., 2023; Fiaboe et al., 2023a, 2024). While previous research has primarily focused on the bioefficacy of these insecticides against FAW, our study addressed the variation in timing and frequency of application, which has been less explored.

We observed that differences in treatment per­formance became notably apparent from the V₇ growth stage of maize plants. Treatments receiving multiple applications, particularly at the early stage of crop growth, exhibited significant reductions in FAW larval numbers per plant. This emphasizes the im­portance of early insecticide application, as it targets smaller larvae and mitigates the development of larger, more damaging larvae later in the season (van den Berg et al., 2021; McClure et al., 2023).

Our analysis revealed that at least one insecticidal application at the early stage of crop growth resulted in less leaf damage by FAW between V₇ to V₉ growth stages. The "Early," "Preventive," and "Targeted" appli­cations significantly reduced FAW-induced dam­age compared to the control, with the "Preventive" and "Targeted" treatments showing the most substan­tial impact. This difference can be attributed to the fre­quency of applications, with "Preventive" plots bene­fiting from three applications by the V₉ stage and "Targeted" plots receiving two to three applications during both experiments. This underscores the criti­cal role of both timing and frequency of applications in managing FAW damage effectively (Tang et al., 2010; van den Berg et al., 2021; Edde, 2022; McClure et al., 2023).

Initiating insecticide treatments during the early growth phases of maize and coupling them with sub­sequent applications can effectively limit damage within economically acceptable levels. "Targeted" ap­proaches, tailored to minimize damage during critical growth stages, offer a promising alternative to stand­ard preventive measures by reducing the need for fre­quent applications while still effectively controlling FAW infestations. This aligns with findings by van den Berg et al. (2021), who highlighted the importance of minimizing FAW damage from V₅ to V₉ stages in en­hancing yield. While our study showed the efficacy of at least two insecticidal applications from V₃ to V₉ maize stage, the ecological implication of the doing is important.

Regarding parasitism results, our study observed that ceasing insecticide treatments at an early crop stage led to higher parasitization rates of FAW larvae, particularly evident in the first experiment. In the sec­ond trial, the beneficial impact of halting early treat­ments was overshadowed by the initial use of emamectin benzoate. In contrast, the early application of PrGV|Btk maintained moderate parasitism rates. This suggests that the strategic cessation or selection of insecticides can influence the biological control of FAW larvae, with implications for integrated pest management strategies that balance chemical control with biological conservation (Ratto et al., 2022). This aligns with recommendations by Agboyi et al. (2023) who recommended the use of bioinsecticides, includ­ing PrGV|Btk over chemical formulations, including emamectin benzoate for FAW control. Nevertheless, they also caution about the variable effectiveness of PrGV|Btk, noting lower parasitism rates in certain re­gions within their study scope. Conversely, Fiaboe et al. (2024) found no adverse effects on populations of natural enemies when integrating the use of this bio­insecticide with soil amendments applied in divided, smaller quantities, highlighting the potential for syn­ergistic strategies in pest management.

Interestingly, our research also points to elevated parasitism rates at the V₇ to V₉ stages, particularly when damage-targeted insecticide applications, pre­dominantly PrGV|Btk, were employed. Also, our PLS-DA analysis indicated that "Targeted" insecticide appli­cation correlated with increased parasitism at the V₇ stage, underscoring the ecological benefits of this application strategy. Given the observed decrease in infestation levels with “targeted” insecticide appli­cation regimen, it is plausible to consider this ap­proach both ecologically beneficial and conducive to enhanced maize yields. This is corroborated by Chisonga et al. (2023), who assert that minimizing FAW infestations during the V₇ to V₉ growth stages is crucial for optimizing maize yield outcomes.

In our investigation, we observed that both pre­ventive and targeted applications of insecticide led to the highest maize grain yield, regardless of the specific insecticide used. This outcome underscores the signif­icance of implementing at least two insecticidal appli­cations between the V₃-V₉ stage of maize crop growth. To mitigate additional costs, we propose adopting infestation-tracked application strategies. As guided by (Prasanna et al., 2018), scouting for FAW management in maize involves systematic field in­spections to monitor larval activity of the pest and the presence of its egg masses on plants. This method of scouting is recognized as an economical IPM strategy, allowing for the application of insecticides solely when it is deemed necessary, as supported by van den Berg et al. (2021). However, in our study, we found that the targeted application did not differ economi­cally form preventive application.

While targeted applications are conceptually more efficient and aligned with IPM principles, their economic advantage over preventive applications may not always be realized due to the complexities of pest management, the inherent costs of application, and the challenges associated with accurately predict­ing and responding to pest pressures. This under­scores the need for a nuanced understanding of pest ecology, application costs, and the effectiveness of scouting and threshold-based decision-making in the economic evaluation of insecticide application strate­gies in maize cultivation. Moreover, it is important to consider that factors beyond the direct costs of insec­ticide applications could influence the economic com­parison of these strategies.

In our study, we found that the efficiency of tar­geted application though higher for the preventive and targeted treatments, differ significantly between insecticides. Targeted PrGV|Btk application resulted in highest benefice-cost than emamectin benzoate. The findings from our study reveal a nuanced land­scape in the efficacy and economic viability of differ­ent insecticide treatments for maize, particularly when comparing targeted applications of PrGV|Btk (a biological control agent) and emamectin benzoate (a synthetic insecticide). The superior cost-benefit ratio observed with targeted applications of PrGV|Btk over emamectin benzoate can be attributed to a combina­tion of factors related to yield impacts and market pricing for crops treated with synthetic versus biolog­ical insecticides. The differential efficiency between these insecticides under targeted application regimes underscores the importance of selecting the appropri­ate pest management strategy based on the specific pest pressure, crop stage, and the mode of action of the insecticide. Biological control agents like PrGV|Btk are often favored in IPM strategies due to their ecolog­ical outcomes, reduced environmental impact, and the lower risk of developing pest resistance compared to synthetic chemicals (Torres & Bueno, 2018; Akutse et al., 2020).

The higher benefit-cost ratio for PrGV|Btk may also reflect market preferences for maize produced with biological control methods, which can command a premium price due to increasing consumer demand for sustainably produced food. This price differential can significantly impact the economic analysis, mak­ing biological treatments more appealing from a fi­nancial perspective, despite potentially higher up­front costs or lower efficacy in some cases (Somasundram et al., 2016; Meemken & Qaim, 2018).

Moreover, the yield impact of using biological versus synthetic insecticides is a critical factor. If PrGV|Btk effectively controls the pest with minimal negative impact on the maize crop, the yield pre­served coupled with a premium market price for bio­logically treated produce can lead to a higher overall economic return. In contrast, while synthetic insecti­cides like emamectin benzoate may offer potent for controlling FAW, as observed in this study, potential negative perceptions and market penalties for syn­thetic chemical use could offset these benefits (Pumarega et al., 2017).

These findings highlight the complexity of choos­ing between biological and synthetic pest control methods. The decision should consider not only the immediate efficacy of the pest control but also the broader implications for crop yield, market pricing, and consumer preferences. This holistic approach en­sures that the selected pest management strategy aligns with both agronomic and economic goals, par­ticularly in the context of sustainable agriculture prac­tices.

Conclusion

These findings highlight the complexity of choosing be-tween biological and synthetic pest control methods. The decision should consider not only the immediate efficacy of the pest control but also the broader implications for crop yield, market pricing, and consumer preferences. The holistic approach ensures that the selected pest management strategy aligns with both agronomic, ecological and economic goals, particularly in the context of sustainable agriculture practices.

Declarations

Ethical statement

The study was conducted outside national parks and protected areas. In the Republic of Togo, neither the insect pest Spodoptera frugiperda nor its natural ene­mies are classified as endangered or protected under specific conservation regulations.

Funding

The authors received no specific funding for this work.

Conflict of interest

The authors affirm that there are no competing interests associated with this work.

Data availability

The authors will provide the raw data that supports the conclusions of this article, readily and without unnecessary reservation.

Acknowledgements

We express our sincere thanks to the technicians at the Entomology and Nematology Lab, School of Agronomy at the University of Lomé for their assistance in various agronomic activities during the study. Our gratitude also extends to all internal and external reviewers of our manuscript.

Author Information

Kokou Rodrigue Fiaboe

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.
• International Centre of Insect Physiology and Ecology (icipe), P.O. Box 30772-00100, Nairobi, Kenya.
• Department of Zoology and Entomology, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa.

Contribution: Conceptualization, Project administration, Investigation, Formal analysis, Visualization, Methodology, Writing original draft, Writing - Review and Editing.

For correspondence: rfiaboe@yahoo.com

Competing interests: No competing interests declared.

0000-0001-9239-194X

Faicedois Abalo

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.

Contribution: Investigation, Methodology, Writing original draft.

Competing interests: No competing interests declared.

0009-0004-7176-8903

Kodjo Médard Abalo

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.

Contribution: Investigation, Methodology, Writing original draft.

Competing interests: No competing interests declared.

0009-0006-7760-1631

Emmanuel Peter

• International Centre of Insect Physiology and Ecology (icipe), P.O. Box 30772-00100, Nairobi, Kenya.
• Department of Zoology and Entomology, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa.
• Department of Agronomy, Federal University Gashua, P.M.B 1005, Yobe State, Nigeria.

Contribution: Writing - Review and Editing.

Competing interests: No competing interests declared.

0000-0003-2660-3830

Agnamto Ossara Agnamba

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.

Contribution: Conceptualization, Project administration, Writing - Review and Editing.

Competing interests: No competing interests declared.

0000-0002-3226-4823

Afoulélou Aboulaye

• Department of Agricultural Economics, University of Nairobi, P.O. Box 29053-00625, Nairobi, Kenya.

Contribution: Data Curation, Formal Analysis, Writing - Review and Editing.

Competing interests: No competing interests declared.

0000-0001-9570-8826

Djima Koffi

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.
• Division of Agricultural Entomology, Georg-August-Universität Göttingen, Germany.

Contribution: Conceptualization, Project administration, Writing - Review and Editing.

Competing interests: No competing interests declared.

0000-0002-5698-2862

Komi Agboka

• Ecole Supérieure d’Agronomie (ESA), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.

• West African Science Service Centre on Climate Change and Adapted Land Use (WASCAL), Université de Lomé, P.O. Box 1515-01, Lomé, Togo.

Contribution: Conceptualization, Project administration, Supervision, Writing - Review and Editing.

Competing interests: No competing interests declared.

0000-0002-5920-3908

Reviewing Editor

Mawufe Komi Agbodzavu

• International Institute of Tropical Agriculture (IITA), Kinshasa, CD.

0000-0001-5435-1250