MITOCHONDRIAL DNA SEQUENCE VARIATION IN THE AFRICAN ARMYWORM SPODOPTERA EXEMPTA (LEPIDOPTERA: NOCTUIDAE)

Author

Amna Eltayeb Elguzouli, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2013

Degree Name

Master of Science

Department

Zoology

First Advisor

Ibrahim, Kamal

Abstract

AN ABSTRACT OF THE THESIS OF AMNA E. ELGUZOULI, for the Master of Science degree in ZOOLOGY, presented on May 8, 2012, at Southern Illinois University TITLE: Mitochondrial DNA sequence variation in the African Armyworm Spodoptera exempta (Lepidoptera: Noctuidae) MAJOR PROFESSOR: Kamal M. Ibrahim The African Armyworm remains problematic as a serious pest of cereals in sub-Sahara Africa. The population genetic structure of African armyworms was investigated to better understand the species' population dynamics and the extent to which the two major outbreak areas in East Africa and Southern Africa are linked by gene flow. Specifically, I characterized the genetic variation of armyworm samples from Tanzania and South Africa and assessed whether the periodic swarming of armyworms homogenizes populations throughout the species range. The variability of a range of mitochondrial genes was screened. A 413 base pair long fragment of the Cytochrome Oxidase I gene (COI), located within positions 1688 to 2175 of the reference locust mitochondrial genome was found to be an ideal marker. This sequence is part of the Folmer region of the COI gene that is used as a 'barcode' for all animal taxa. Haplotype diversity was found to be comparable to reports in the Spruce Budworm (Choristoneuru fumiferana), which, like the African armyworm, undergoes sporadic population buildup. The sequences were A+T rich particularly at the first codon position confirming previous findings in other insect species. ii A population genetic analysis of the sequence variation revealed that African armyworm populations are genetically structured and do not form a single panmictic population. Despite the limited sampling in this study, a clear picture of significant genetic separation between Tanzanian and South African armyworm populations has been confirmed. This leads to the conclusion that the extent of mixing resulting from the well documented tracking of the Inter Tropical Convergence Zone by outbreak populations does not connect Tanzanian and South Africa armyworm populations. Many rare haplotypes that are one mutation step from one of two common haplotypes were observed; a parsimony based haplotype network supports the hypothesis that these rare haplotype may have arisen during the regional population expansions. However, because the southern and northern regions that were sampled share the same two common haplotypes, neither the phylogenetic trees nor the haplotype networks yielded geographic signal that show where the rare haplotypes may have evolved. The presence of many unique singleton haplotypes, coupled with the pattern of their mismatch distribution, suggests exponential growth of localized armyworm populations. Outbreak populations arise separately in different locations across the species geographic distribution; it appears that the sizes of these populations become large enough for new mutations to produce new variants in each location. The fact that the two commonest haplotypes were present both in South Africa and in Tanzania indicates that some exchange of migrants does take place between the two regions. Alternatively, since the haplotypes that are unique to each region are derived from, and a single mutation step away from, these two common haplotypes, this could indicate that the two common haplotypes were shared in the ancestral populations and that no current gene flow takes iii place. In either case, it is apparent that the periodic swarming of armyworms does not homogenize the entire species range. What are the implications of these findings to the control of armyworm infestations? Armyworm population dynamics in the species' northern range appears to be decoupled from its population dynamics in the south. This implies that this pest's control strategies need to be region specific. However, this study had only two sampling points. It is plausible that a stepping-stone type of link exists in between and that I have sampled the extreme northern and southern ends of the link. A finer scale, multipoint sampling throughout the species geographic range, preferably during multiple years, is required in order to answer this question unambiguously. This would assist in developing an outbreak early warning system and in determining if multi-region control strategy is warranted. I also draw attention to another caveat to the above conclusions: a significant proportion of the divergence between the two sampling localities is attributable to two rare but highly divergent haplotypes. The possibility of these two haplotypes belonging to another, as yet unknown moth species, has been discussed. In addition to the above insights, this study has assessed the utility of the large number of universal mtDNA primers in the study of armyworm population genetics.