The fall armyworm strain associated with most rice, millet, and pasture infestations in the Western Hemisphere is rare or absent in Ghana and Togo

Abstract

The moth pest fall armyworm, Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae) is now present throughout much of the Eastern Hemisphere where it poses a significant economic threat to a number of crops. Native to the Western Hemisphere, fall armyworm is one of the primary pests of corn in the Americas and periodically causes significant economic damage to sorghum, millet, cotton, rice, and forage grasses. This broad host range is in part the result of two populations historically designated as host strains (C-strain and R-strain) that differ in their host plant preferences. Reports of infestations in Africa have to date mostly been limited to the C-strain preferred crops of corn and sorghum, with little evidence of an R-strain presence. However, this could reflect a bias in monitoring intensity, with the R-strain perhaps being more prevalent in other crop systems that have not been as routinely examined for the pest. Because knowledge of whether and to what extent both strains are present is critical to assessments of crops at immediate risk, we analyzed specimens obtained from a systematic survey of pasture grass and rice fields, habitats typically preferred by the R-strain, done contemporaneously with collections from corn fields in Ghana and Togo. Substantial larval infestations were only observed in corn, while pheromone trap capture numbers were high only in corn and rice habitats. Little to no fall armyworm were found in the pasture setting. Comparisons with a meta-analysis of studies from South America identified differences in the pattern of strain-specific markers typically found in fall armyworm collected from rice habitats between the two hemispheres. Genetic tests of specimens from rice and corn area traps failed to show evidence of differential mating between strains. These results are consistent with the R-strain being rare or even absent in Africa and, at least for the Ghana-Togo area, this R-strain lack does not appear to be due to limitations in pest monitoring. The implications of these results to the crops at risk in Africa and the accuracy of existing molecular markers of strain identity are discussed.

Fig 1. Map showing locations of fall armyworm collections in Ghana and Togo.

Major towns near locations surveyed in 2018 are labeled. Sites are described in Table 1. The map was generated using QGIS version 2.18.2 (Open Source Geospatial Foundation).

Fig 2. Schematic of gene segments used for the genetic analysis.

A. COIB segment of the mitochondrial COI gene showing location of the strain diagnostic mCOI1164 SNP as well as two additional polymorphic sites showing the same strain-specificity. Nucleotides observed at each SNP are listed below arrows and the configurations associated with COI-CS and COI-RS described. B. Segment of the nuclear Tpi gene showing location of the gTpi183 site that is diagnostic of strain identity in Western Hemisphere populations. Site gTpi168 shows a similar polymorphic distribution as gTpi183. The gTpi180 polymorphism is not typically found in the Western Hemisphere but is present in Africa. Nucleotides observed at each site are listed below arrows and the configuration associated with TpiC and TpiR indicated. Because Tpi is Z-linked, males have two copies of the gene and so can be heterozygous for these polymorphisms (TpiH). Site gTpi192 is polymorphic for C or T in both strains (strain nonspecific). Nomenclature follows IUPAC convention where Y = C or T; S = C or G; R = A or G; and D = A, G, or T. Small block arrows denote location of relevant primers used for PCR and DNA sequencing.

Fig 3. Graph of fall armyworm pheromone trap captures in corn, pasture, and rice habitats in Ghana (A) and Togo (B).

Fig 4. Bar graphs showing distribution of the R-strain markers COI-RS (A) and TpiR (B) in rice and corn habitats in Argentina and Brazil.

Data are from three studies, Juarez et al. (2012) [9], Machado et al. (2007) [35], and Murua et al. (2015) [14]. Mean frequencies (± standard deviation) are noted above columns with different letters indicating statistically significant differences using a two-tailed t-test.

Fig 5. Bar graphs showing distribution of the R-strain markers COI-RS (A) and TpiR (B) in rice and corn habitats in Ghana and Togo.

Collections are as described in Table 1. Mean haplotype frequencies (± standard deviation) are noted above columns. Within each graph, frequencies with different lower-case letters are statistically different. Numbers in brackets indicate specimens tested from each collection. For TpiR, the graphs show the frequencies of TpiR hemizygotes or homozygotes (dark fill) and TpiH heterozygotes (diagonal lines). TpiR haplotype frequencies were calculated by combining these classes as described in the Methods.

Fig 6. Bar graph of the inbreeding coefficient, F_IS, calculated for different fall armyworm collections.

The number above each column pair denotes the difference (δ) between the F_IS values, δ = (gTpi183 F_IS)–(gTpi192 F_IS). Numbers in brackets indicate specimens tested from each collection.