The fall armyworm, Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae), is an important pest of many crops including maize (Zea mays L.; Poaceae), sorghum (Sorghum bicolor [L.] Moench; Poaceae), cotton (Gossypium hirsutum L.; Malvaceae), and diverse pasture grasses. It is widely distributed in the Americas (Sparks 1979) and recently has become a pest of concern in Africa (Goergen et al. 2016; Nagoshi et al. 2017). Given the importance of maize in these regions, this pest has become one of the most serious problems on both continents. The yield losses caused by fall armyworm vary and depend on various factors, but can range from 19 to 100% in Brazilian plantations (De Almeida Sarmento et al. 2002). After severe outbreaks in Africa, control of fall armyworm has been based principally on the use of chemical insecticides. However, biological control is a highly desirable management alternative for controlling this pest in the long term. The success of any biological control project depends on appropriate biological, ecological, and population studies of the species involved (Miller 1983).

There have been surveys for natural enemies of fall armyworm in the USA (Hogg et al. 1982; Ashley 1986; Meagher et al. 2016), Mexico and Central America (Castro & Pitre 1989; Molina-Ochoa et al. 2003), and South America (Beserra et al. 2002; Murúa et al. 2009). Although some species exist throughout the Western Hemisphere, most species' distribution generally is defined by their geographic areas (Ashley 1979). Since fall armyworm is new to Africa, it is not known which local natural enemy species will use it as a host, although a recent report identifies parasitoids found in eastern Africa (Sisay et al. 2018). Successful biological control management programs for fall armyworm will be those that incorporate multiple natural enemy species (Gross & Pair 1986; Riggin et al. 1993; Figueiredo et al. 2006; Wyckhuys & O'Neil 2006, 2007).

Materials and Methods

COLLECTION AND REARING OF NATURAL ENEMIES

Predators and parasitoids were collected from the field. Predators that were found preying on the larvae of S. frugiperda were directly collected and preserved in 70% ethanol; parasitoids were larval endoparasitoids which developed inside S. frugiperda larvae. Therefore, S. frugiperda larvae were collected from selected sites on infested maize plants, which were identified by the presence of larval feeding injury in the whorls and newly deposited frass.

Table 1.

Localities with the collected natural enemies of S. frugiperda, Ghana 2017.

Fall armyworm larvae collected from the field were morphologically identified and placed individually in conical transparent rearing containers (8.5 × 5.5 × 11 cm³). Rearing containers were fitted with a black mesh cover at the top to allow ventilation and to prevent escape of larvae. Tissue paper was placed at the base of the containers to absorb moisture produced by the diet (maize leaves) and larvae (transpiration and frass). The larvae were reared in the laboratory under room conditions (27 ± 4 °C, 80 ± 8% RH, and 12:12 h [L:D] photoperiod). Parasitoids that emerged from the collected larvae were recorded every 24 h and placed in 70% ethanol. Natural enemies were identified by G. Goergen, Curator at the Biodiversity International Institute of Tropical Agriculture Benin.

Results

Fall armyworm larvae were collected from all maize fields across Ghana during both cropping seasons. Natural enemies of fall armyworm occurred in 17.0% (18 of 106) of the inspected sites, including 8 sites that were sprayed with insecticides. For this study, a total of 1,062 S. frugiperda larvae were collected and 38 were parasitized, yielding a parasitism rate of 3.6%. There was no difference in parasitism between the major (4.8 ± 2.3%) and minor (3.2 ± 0.9%) cropping systems (F_{1, 18} = 0.61; P = 0.4465). There also was no difference in parasitism among regions, with a range from 0% in the Upper West to 7.8% in the Northern region (F_{9, 10} = 0.58; P = 0.7870) (Table 2).

The collection of fall armyworm during the major cropping season from May to Jul included 29 farms from the 10 regions, but natural enemies of fall armyworm were recorded only in 8, representing 27.6% of the sites. Even though parasitoids were not found in the Greater Accra, Central, Western, Brong Ahafo, and Upper West regions, a total of 306 larvae were collected from the field, and only 12 were parasitized giving a 3.9% parasitism rate. For the 8 sites that contained parasitoids, a parasitism rate of 17.4 ± 3.41% was recorded (Table 3).

The minor cropping season collections were carried out from the end of Aug to Nov on 77 farms from the 10 regions. Larval natural enemies of fall armyworm were recorded in 10 sites representing 13.0% of sampled farms. During this season, 756 S. frugiperda larvae were collected, and 26 larvae were parasitized representing a parasitism rate of 3.4%. Parasitoids were not found in the Upper West, Upper East, and Volta regions, but among sites that contained parasitoids, Agogo Aburkyi, Legon, Kpong, and Sanga were not sprayed with insecticides and had the highest parasitism rates of 60%, 55.6%, 33.3%, and 23.8%, respectively (Table 4).

Seven species of parasitoids were identified: C. bifoveolatus, C. icipe, Coccygidium luteum, M. testacea, and Bracon sp., A. erinaceus, and an undetermined Tachnidae species. The 2 most abundant parasitoids were C. bifoveolatus and Coccygidium luteum, with parasitism rates of 1.04% and 0.85%, respectively, and relative abundance of 29.0% and 23.7%, respectively. Chelonus bifoveolatus was the most dispersed parasitoid, found in 7 of the inspected sites (Table 5). Three species of predators were identified: P. megacephala, H. obscuripennis, and P. nodulipes (Table 1). The predator most abundant and most dispersed nationwide was P. megacephala, with a relative abundance of 46.0% collected from 4 of the inspected sites (Table 5).

In total, larval mortality due to unknown factors was 63.8%, and was not different between the major (65.6%) and minor (61.8%) cropping seasons (F_1,18 = 0.61; P = 0.4460). Larval mortality due to unknown factors ranged from 55.3% in the Eastern Region to 73.8% in the Greater Accra Region (Table 2); however, there was no difference among regions (F_{9, 10} = 1.52; P = 0.2612; Table 2). Only 32.7% of larvae collected from fields completed development in the laboratory, and there was no difference between the major (29.5%) and minor (34.9%) cropping seasons (F_1,18 = 2.05; P = 0.1689). The Greater Accra region had the lowest development rate (20.6%), which was significantly lower than the Eastern (40.8%) and Upper West (42.6%) regions (F_{9, 10} = 4.0; P = 0.0208; Table 2).

Table 2.

Fate of fall armyworm larvae collected during the major and minor maize seasons in different regions of Ghana, 2017.

Discussion

The parasitism rate ranged from 6.25% at the Kwame Nkrumah University of Science and Technology site in the Ashanti Region, to 60% at the Bolga 2 site in the Upper East Region, and Agogo Aburkyi in the Ashanti Region. All 3 sites were not sprayed with chemical insecticides. The variations in parasitism is due to natural and cultural practices that can negatively or positively affect natural enemy populations (Kogan et al. 1999). The average parasitism across the country was lower than the previous findings in the Americas (8.1%, Ordòǹez-Garcia et al. 2015; 13.8%, Molina-Ochoa et al. 2004; 15.5%, Wheeler et al. 1989; 18.3%, Murúa et al. 2009; 28.3%, Meagher et al. 2016; 35%, Rios-Velasco et al. 2011; 39%, Murúa et al. 2009). This low parasitism rate of S. frugiperda is due to the aspect of new pest introductions in Ghana. Biological control of fall armyworm requires mass rearing of introduced parasitoids in the laboratories in Africa with field releases to increase parasitism. We believe that parasitism of fall armyworm will increase as the pest continues to be present. However, high applications of chemical insecticides will negatively affect the natural enemies. Therefore, parasitoids must be preserved (Pair et al. 1986; Molina-Ochoa et al. 2004) by using selective systemic insecticides (Figueiredo et al. 2006).

Chelonus bifoveolatus was the most abundant parasitoid collected from larvae. Chelonus spp. are typical of egg-larval solitary koinobiont endoparasitoids that attack Noctuidae and Pyralidae (Marsh 1978; Virla et al. 1999; Murúa et al. 2009) by ovipositing into host eggs (Pierce & Holloway 1912; Rechav & Orion 1975). The parasitized host larvae exhibit reduced growth rates and weight compared to unparasitized larvae (Ables & Vinson 1981; Ashley et al. 1983). In the Western Hemisphere, Chelonus spp. appear to be the most geographically dispersed parasitoid of fall armyworm (Ashley et al. 1982, 1983; Meagher et al. 2016) and were reported to be present in 12 countries of the Caribbean, and South and Central America (Molina-Ochoa et al. 2003). In many areas, Chelonus spp. were reported to be the most common species collected (Wheeler et al. 1989; Cortez-Mondaca et al. 2010, 2012; Rios-Velasco et al. 2011; Estrada-Virgen et al. 2013). Chelonus bifoveolatus was found in 7 of the 10 regions (Greater Accra, Central, Eastern, Volta, Ashanti, Brong Ahafo, and Northern) and was recorded in all agroecological zones of Ghana (Coastal Savannah, Evergreen, Equatorial Forest, Transition Zone, and Guinea Savannah). These results suggest that this species is adapted to all sub-Saharan Africa agroecological zones. Therefore, this species can be a good agent for biological control of fall armyworm in Africa. However, the most efficient biological control programs for fall armyworm are ones that use and amplify several parasitoid species rather than programs that rely on an individual natural enemy species (Riggin et al. 1993). Therefore, we believe that adding New World parasitoids that are known to attack fall armyworm (classical biological control), plus preserving the parasitoids that are already active in Africa (conservation biological control), will contribute to reducing pest populations. The other active parasitoids included C. icipe, which is the koinobiont endoparasitoid that attacks lepidopteran larvae (Quicke 1997; Whitfield 1997). Cotesia spp. also are reported parasitoids of fall armyworm in the Americas (Ashley 1983; Meagher et al. 2016). These species may parasitize both eggs (Ruberson & Whitfield 1996), and the first and second instar larvae (Loke et al. 1983). Cotesia icipe was reported to be a successful parasitoid of a major maize pest in West Africa and in Mediterranean countries (Kaiser et al. 2017). Coccygidium luteum, as all members of this genus, is an internal koinobiont parasitoid of larval Noctuidae (Sharkey et al. 2009). Bracon sp. is an idiobiont ectoparasitoid that attack larvae with hidden behaviors, such as cereal stem borers and cereal stored products borers (Moolman et al. 2013; Souobou et al. 2015). Meteoridea testacae is a gregarious endoparasitoid that was defined as an egg-larval parasitoid of Lepidoptera (Achterberg 1993). The grass fly, A. erinaceus is widespread in Africa, but its parasitic status report is controversial (Upadhyay et al. 2001). However, it is documented as a parasitoid of sugarcane borers. Tachinidae species are almost all parasitoids. In Africa, they are usually collected from cereal stem borer larvae (Moolman et al. 2013; Chinwada et al. 2014).

Table 3.

Sites within regions in Ghana hosting parasitoids of fall armyworm collected during the major cropping season, 2017.

Table 4.

Sites within regions in Ghana hosting parasitoids of S. frugiperda collected during the minor cropping season, 2017.

Pheidole megacephala, the most abundant and most dispersed predator, prefers humid forest habitats (Hoffmann et al. 1999; Wilson 2003; Burwell et al. 2012). When introduced into a new area, P. megacephala expanded its range and invaded into the forest interiors, where it attacked and displaced other introduced and naturally occurring ant species (Hoffmann 1998; Burwell et al. 2012). But the efficacy of P. megacephala as a biological agent is a challenge due to its generalist behavior and ecological disaster. However, it is a good predator with an efficient nest mate recruitment that enables the species to dominate baits and to retrieve prey too large for single workers to carry (Dejean et al. 2007, 2008). Therefore, P. megacephala must be protected as a complementary biological agent of S. frugiperda in forest zones, as well as the other 2 species of predators: H. obscuripennis and P. nodulipes, which also were found in the Equatorial Forest of the Eastern Region. The 2 Reduviids were found in the collection of true bugs sampled from Lama Forest in southern Benin (Attignon 2004).

Table 5.

Parasitism rate, relative abundance, and dispersion of different parasitoid species of fall armyworm collected from maize fields in Ghana, 2017.