Key message

  • High use of insecticides against fall armyworm in West Africa, and performance of natural enemies.

  • The use of insecticides has affected population densities and performances of these species.

  • This study draws attention to moderate use of insecticides to improve natural control of this pest.

Introduction

The equilibrium among populations of living organisms is due to the interdependent relationships among plants, herbivores, carnivores, parasites, and pathogens that regulate and balance the bio-ecosystem (Szwabiński et al. 2010). Unfortunately, this balance is constantly challenged by human activities that cause routine disruptions (Ruppert et al. 2018). Agricultural production is one of the human activities that throw off the biodiversity balance, and increased agricultural pest populations can lead to food and economic losses (Zamagni 2012; Sala et al. 2013). To avoid these losses, various crop protection methods have been developed against phytopathogens and crop pests from mammals, birds, reptiles, and arthropods (Berny et al. 2010; Roger et al. 2014). The use of pesticides has been the most common method of pest control. Unfortunately, some pest species have developed different resistance mechanisms to various pesticide families (Anderson et al. 2018; Tay and Gordon 2019), while other species escape and migrate to new regions. In the absence of consistent control measures or indigenous natural enemies, a successful pest invasion is usually followed by rapid population increase and movement, resulting in severe crop damage and potentially serious yield and economic losses (De Barro et al. 2015; Haile et al. 2021).

A currently known invasive pest is the fall armyworm (FAW), Spodoptera frugiperda (J. E. Smith) (Lepidoptera: Noctuidae), an insect that has spread from its native occurrence in the neotropical Americas to many parts of the Eastern Hemisphere. This polyphagous insect has been reported on over 350 host plants, including many crops of economic importance (Montezano et al. 2018). Unfortunately, since 2016, it has been detected in sub-Saharan Africa (Goergen et al. 2016; Nagoshi et al. 2017, 2018, 2022; Koffi et al. 2020a, b), south and east Asia (Nagoshi et al. 2020; Kim et al. 2021), and Oceania and Australia (Bourke and Sar 2020). The spread of this pest has caused severe economic damage to cereal production (mostly maize) and has disrupted global agricultural systems and food security (Koffi et al 2020a, b, 2022). To reduce food and economic losses, insecticides are mainly used by millions of farmers with the support of several governments (Koffi et al. 2020b, 2021). However, indiscriminate use of insecticides not only threatens the human health and environmental protection (O’Dowd et al. 2003) but also disrupts natural biodiversity interdependencies by killing non-target organisms, that includes natural enemies.

Over one hundred species of natural enemies of FAW have been reported worldwide (Molina-Ochoa et al. 2003a; Murúa et al. 2009; Meagher et al. 2016). In Africa, many indigenous entomopathogens, parasitoids, and predators of FAW have been reported (Sisay et al. 2018; Agboyi et al. 2020; Koffi et al. 2020c). However, due to the indiscriminate application of insecticides in the invaded areas, population trends and the potential of natural control from indigenous agents are poorly known. In the west African nation of Togo, infestation of FAW was three times lower from 2018 to 2020, compared to the previous two years following the invasion (Koffi et al. 2020a, 2022). Thus, the aim of this study was to identify and evaluate the spatial distribution and impact of insecticide applications on indigenous entomopathogenic viruses, bacteria, fungi and nematodes, parasitoids, and predators established with FAW populations during the seven years following the invasion of the pest in Togo.

Materials and methods

Sites of collections

Fall armyworm egg masses and larvae were collected during the cropping seasons from April to November of seven consecutive years (2016–2022). The collections of specimens were initiated at the onset of the FAW invasion in Togo. In 2016, selected collection sites covered the area from Lomé (6.176 N) to Kara (9.377 N). From 2017 to 2022, collections covered the entire country from Lomé to Dapaong (10.474 N). The collection sites were randomly selected during survey trips for maize farm inspections by the Entomological Research team from the Ecole Superieure d’Agronomie, Université de Lomé (Fig. 1).

Fig. 1
figure 1

Collection locations with some reference cities, the years of collections and numbers of maize farms inspected from 2016 to 2022

Full size image

Collection and preservation of specimens

Within the selected farms, quadrants were designated at the four cardinal points and middle of each farm. Egg masses and larvae were collected from 100 plants within the quadrants as well the living organisms that were preying on egg masses or larvae. The collected egg masses and larvae were placed individually in rearing boxes as described by Koffi et al. (2020c) and transferred in coolers (Mobicool ME24, 23 L, 12 V, Hannover, Germany) to the Université de Lomé laboratory. Predators were immediately preserved in crystal vials containing 70% ethanol (Koffi et al. 2020c). The egg masses and larvae completed their life cycle during laboratory rearing (25 ± 5 ℃, 78 ± 15% relative humidity, and 12:12 photoperiod) to adulthood of non-infected specimens or to the death of individuals with symptoms of entomopathogens or emergence of parasitoids.

Microbial natural enemies

Dead larvae with evidence of bacterial or viral symptoms were isolated and preserved in the freezer at 4 ℃. Description keys were used to identify bacteria (Eilenberg et al. 2015). Dead larvae presenting green-colored microflora germinations were assigned to undetermined entomopathogenic bacteria. Two categories of viruses were identified using morphology of dead larvae. Stunted larvae with production of virus-filled vesicles and milky-white discoloration were classified as Ascoviruses (Federici et al. 2008). Larvae with whitish-gray discoloration and a swollen body with a ruptured integument leading to liquefaction, were assigned to the Baculoviruses (Haase et al. 2015; Valicente 2019).

Symptoms of infections from entomopathogenic fungi were germination of hyphae from dead larvae, although a few samples were morphologically identified by scientists at the USDA-ARS in Ithaca, NY, USA. The infected samples were removed from their rearing boxes and individually placed in an empty Petri dish under the same laboratory rearing conditions. They were then individually transferred to a Petri dish containing potato dextrose agar (PDA) previously prepared for fungi germinations and isolations. The preparations were incubated at laboratory rearing conditions for 10–21 days for identification.

Nematodes, parasitoids, and predators

Nematodes and parasitoids that emerged from FAW eggs, and larvae were preserved in 70% ethanol in crystal vials. Description keys were used to identify the entomopathogenic nematodes (Crosskey 1968; Baker and Capinera 1997; Firake and Behere 2020).

Identification keys (Braet et al. 2012; Koffi et al. 2020c) and characteristics of insects already identified using molecular barcodes from Africa (Agboyi et al. 2020; Durocher-Granger et al. 2021; Otim et al. 2021) were used to identify emerged parasitoids and collected predators (Brindle 1967; Waller et al. 1999; Kwadjo et al. 2012; Nicolas et al. 2015; Girod and Lassalle 2017).

Calculations and statistical analysis

During the inspections, insecticide application information was documented from the farmers to determine the percentage of farms with insecticide applications per year (sprayed farms). However, calculations did not consider farms where the owner was not present to provide information. Specimen collections were separated into two periods, (1) intensive insecticide applications, where more than 75% of the farms made insecticide applications for FAW management (2016 and 2018) and (2) occasional insecticide applications, where less than 25% of the farms made insecticide applications (2018–2022).

After specimen identifications, several variables were calculated according to the following equations (Koffi et al. 2020c):

$$y=\frac{n}{N}*100$$
(1)

or

$$i= \frac{t}{T}$$
(2)

where infection rates (Ir) (y), n = number of larvae infected by a given entomopathogen species and N = total number of collected larvae. For parasitism rates (Pr) (y), n = number of egg masses or larvae parasitized by a given parasitoid species and N = total number of egg masses or larvae collected. For relative abundances (RA) (y), n = number of a given species and N = total number of collected species. For percent of sprayed farms (y), n = number of sprayed farms and N = total inspected farms per year. For index of dispersion (iD) (i), t = ratio of the number of farms hosting a given nematode, bacterium, virus, fungus, parasitoid or predator species, and T = total number of farms inspected for each year

Data were arranged per location and grouped per year. The percentage data were arcsine square root transformed prior to statistical analysis. All calculations and transformations were carried out using Excel before being submitted to a Shapiro test for normality (GenStat Twelfth Edition GenStat Procedure Library Release PL20.1). Normal data were submitted to one-way analysis of variance at 95% confident interval, and the non-normal data to a non-parametric test (Kruskal–Wallis) at 5% significance level. Multiple means obtained from the ANOVA were subjected to a Tukey test for separation, while means comparing the parasitism rates between the sprayed and unsprayed farms were subjected to a t-test. The assessment of correlation between the numbers of collected egg masses and larvae was also calculated using GenStat.

Results

Impact of insecticides on the performance of natural control

The two years following the invasion of FAW in Togo showed higher numbers of egg masses collected per farm than the next several years. Although there were high numbers of egg masses those first two years, none were found to be parasitized. However, the low numbers of egg masses collected since 2018 showed parasitism rates increasing from 14.72% in 2018 to 45.38% in 2022 (Table 1). The same trend was shown with larvae as higher numbers were collected per farm in 2016 and 2017 than the following years. Correlation analysis between collected egg masses and larvae showed no relationship (r = 0.0002, P = 0.869). Larval parasitism rates were very low in 2016 (1.32%) and 2017 (2.53%) but increased to 15.96% in 2018 and 42.23% in 2022 (Table 1).

Table 1 Numbers and parasitism rates of egg masses and larvae across years of collections
Full size table

During the survey, farmers were interviewed regarding insecticide applications and depending on the year, between 52.1 and 74.1% responded. The percentage of sprayed farms was very high in 2017 (91.38% of farms inspected) and 2016 (73.25%), compared with the range of 23.52–11.08% obtained between 2018 and 2022, respectively. Higher egg and larval parasitism rates were recorded in unsprayed farms than sprayed farms every year of the seven-year study (Fig. 2).

Fig. 2
figure 2

Percent of egg masses and larvae parasitized or infected from 2016 to 2022 in Togo

Full size image

Entomopathogenic organisms associated with FAW

Entomopathogenic agents associated with FAW larvae included a nematode, Ovomermis sinensis Chen, Jian, & Ren (Nematoda: Mermithidae), unidentified species in the bacteria family of Enterobacteriaceae and Enterococcus sp., unidentified viruses belonging to Ascoviruses and Baculoviruses, and fungal species Isaria sp. (Hypocreales: Clavicipitaceae) and Metarhizium rileyi (Farl.) Kepler, Rehner, & Humber (Hypocreales: Cordycipitaceae). The nematode, O. sinensis had low infection rates, relative abundance and index of dispersion until 2018, compared to the following years (Table 2). The infection rate of bacteria species belonging to Enterobacteriaceae was not constant during the period of collections, but its relative abundances were high in 2016 and 2017. The indexes of dispersion were constant over the study period. Infection rates of Enterococcus sp. increased from 2018 onward, while its relative abundance decreased, and its indexes of dispersion remained constant during the study (Table 2). Larvae infected by Ascovirus were collected from 2018 to 2022, and Baculovirus was collected from 2017 to 2022. However, the two viruses and fungal groups had constant infection rates, relative abundance, and indexes of dispersion throughout the study (Table 2). Although the spatial distribution of entomopathogens increased during the study (Fig. 3), the population densities of each species were similar.

Table 2 Entomopathogenic associated with FAW from 2016 to 2022 in Togo and their infection rates, relative abundances, and index of dispersions
Full size table
Fig. 3
figure 3

Sites of collections of entomopathogenic agents—O. sinensis (yellow), Enterobacteriaceae (dark yellow), Enterococcus spp. (orange), Ascoviruses (red), Baculoviruses (purple), Isaria spp. (violet), and M. rileyi (pink) associated to FAW larvae from 2016 to 2022 in Togo

Full size image

Parasitoids

During this study, 10 species of parasitoids were collected and identified. The egg parasitoid Telenomus remus Nixon (Hymenoptera: Platygastridae) was consistently collected from 2018, with increasing parasitism rates from 15.38 to 45.44% in 2022. This species index of dispersion was higher during the last two years of the collection period (Table 3). One species of egg-larval parasitoid, Chelonus bifoveolatus (Szépligeti) (Hymenoptera: Braconidae), was also collected with low parasitism rates that increased from 0.39% in 2016 to 3.70% in 2021. This species had an increasing index of dispersion and decreasing relative abundance during the study (Table 3). The larval parasitoids included two unidentified species of Tachinidae, five hymenopteran species including Coccygidium luteum (Brullé), Cotesia icipe (Fernandez-Triana and Fiaboe), an unidentified Braconidae, and two unidentified Ichneumonidae, and the larval-pupal parasitoid species Meteoridea testacea (Granger) (Braconidae). The indexes of dispersion of the tachinid species were the same across all years of collection, with small variations found among the parasitism rates and relative abundances (Table 3). Parasitism rates of all the other larval parasitoids increased moderately from 2018 onward, with low variation in relative abundance and index of dispersion (Table 3). Although the spatial distribution of parasitoids increased every year (Fig. 4), the population densities of each species were similar.

Table 3 Egg, egg-larval and larval parasitoids with annually parasitism rates, population densities, relative abundance and index of dispersion
Full size table
Fig. 4
figure 4

Sites of collections of parasitoids—T. remus (yellow), C. bifoveolatus, (dark yellow), C. luteum (orange), C. icipe (red), M. testacea, (purple), braconid (violet), ichneumonid (pink), and tachinids (blue) associated to FAW egg masses and larvae from 2016 to 2022 in Togo

Full size image

Predators

A total of 16 species of arthropod predators were collected attacking FAW during this study. These were five heteropterans, Orius sp. (Anthocoridae), Haematochares obscuripennis Stål, Peprius nodulipes (Signoret), Rhynocoris sp., and Zelus sp. (Reduviidae); four beetles, Calleida sp. (Carabidae), Cheilomenes sulphurea (Olivier), Coccinella sp. and an unidentified coccinellid (Coccinellidae); three earwigs, Euborellia sp., Forficula senegalensis Audinet-Serville, and Forficula sp. (Forficulidae); two ants, Pheidole megacephala F. and Polyrhachis sp. (Formicidae); one lacewing, Chrysoperla sp. (Chrysopidae); and one mantid, Sphodomantis viridis Forsskal (Mantidae).

Except for Orius sp. and Zelus sp., all the heteropteran species were collected from 2018 and had similar relative abundances and indices of dispersion during the collection years (Table 4). The beetle species were collected from 2017 and had similar relative abundance across the years of collection with slightly increasing indices of dispersion (Table 4). Except for Euborellia sp., the earwig species were collected from the onset of the FAW invasion with a slight increase in relative abundances and dispersion in the following years (Table 4). Ants collected were social Formicidae that increased in their locations found during the study. Chrysoperla sp. and Mantis sp. were collected in higher numbers during the first years of the FAW invasion in Togo, and their dispersion increased during the study (Table 4). Although the spatial distribution of the predators expanded during the study (Fig. 5), the population densities of each species were similar.

Table 4 The predator species collected from 2016 to 2022 in Togo with their relative abundances and index of dispersions
Full size table
Fig. 5
figure 5

Sites of collections of predators—Forficula sp. (yellow), F. senegalensis (dark yellow), Euborellia sp. (orange), Coccinella sp. (red), coccinellid (purple), C. sulphurea (violet), Calleida sp. (pink), P. nodulipes (light-pink), Rhynocoris sp. (light-green), Zelus sp. (blue), Orius sp. (dark blue), P. megacephala (green) P. lamellidens (dark-green) H. obscuripennis (hash), Chrysoperla sp. (dark-hash), Mantis sp. (dark) associated to FAW larvae from 2016 to 2022 in Togo

Full size image

Discussion

Many of these natural enemy species have been documented before in several areas of the world, especially the parasitoids and predators (Sisay et al. 2018; Koffi et al. 2020c; Abang et al. 2021; Dassou et al. 2021; Otim et al. 2021). The nematode species, O. sinensis, has been described to infest several noctuid species (Li et al. 2003), including S. frugiperda (Sun et al. 2020). The bacterium found, Enterococcus sp., is a common gut bacterium found in many Lepidoptera species and is most likely non-pathogenic (Voirol et al. 2018; Kenis et al. 2022). Viruses have been studied for many years to be used as microbial insecticides against FAW (Molina-Ochoa et al. 2003b; Guo et al. 2020; Hussain et al. 2021). Fungi, including Isaria spp. and M. rileyi, have also been studied as biopesticides against noctuid larvae for many years (Guo et al. 2020).

The invasion of FAW in Africa was successful as increased population growth and rapid spread were followed by severe damage to maize plants and important yield and economic losses (Koffi 2020a, b, 2022). The response of maize producers and governments to threatened food security was the application of insecticides (Koffi et al. 2021). In Togo, up to 73.25% of maize farms were sprayed with insecticides to reduce heavy infestations of FAW. This increased up to 91.38% in 2017 and was expected to stabilize or to increase in 2018 (Ramírez-Cabral et al 2020). Surprisingly, insecticide applications decreased to 23.52% in 2018 and reached 11.08% in 2022. The decrease in insecticide applications coincided with the increasing numbers and activity of natural control agents which were already being collected during maize production in 2016 and 2017. Fortunately, the infestations of FAW between 2018 and 2020 were three times lower than the previous two years (Koffi et al. 2020a, 2022), which may explain why growers reduced their sprays. This unexpected phenomenon calls into question the efficiency of many insecticides applied against the FAW in Togo. The activity of natural control may be underestimated or other abiotic factors (i.e., rainfall) may be involved in reducing FAW infestation.

Even if natural enemy populations were expected to be low and inefficient during the two years following the invasion of the new pest, the three to four times higher parasitism rates of egg masses and larvae in the unsprayed than sprayed farms demonstrated the negative effects of insecticides on these natural enemy populations. Therefore in Togo, the reduction in the number of insecticide applications since 2018 most likely contributes to the emergence of higher populations of pathogens, parasitoids, and predators and improvement in their performance.

Author contributions

All authors designed the research methods. DK, KA, MKAA, MA, and KT collected the data. DK, KA, and RLM wrote the manuscript. All authors reviewed the manuscript.