Mitochondrial genetic variation reveals phylogeographic structure and cryptic diversity in Trioza erytreae

Abstract

Trioza erytreae is the main vector for 'Candidatus Liberibacter africanus', the causative agent of African Citrus Greening disease. The insect is widespread in Africa, and has recently disseminated to Southwestern Europe. This study aimed at generating reference mitogenome sequences for T. erytreae, as a background for future genetic diversity surveys. Complete mitochondrial sequences of three specimens collected in Ethiopia, Uganda and South Africa were recovered using Ion Torrent technology. The mitogenomes of T. erytreae from Uganda and Ethiopia were highly similar, and distinct from that found in South Africa. The phylogeographic structure of T. erytreae was assessed using genetic clustering and pairwise distances, based on a dataset of public COI sequences recorded as T. erytreae. The dataset revealed ten haplotypes with strong phylogeographic structure in Africa and Europe. Three haplotypes found in Kenya on Clausena anisata belonged to pairs separated by distances as high as 11.2%, and were basal to all other sequences. These results indicate that not all sequences identified as T. erytreae belong to the same species, and that some degree of specificity with different plant hosts is likely to exist. This study provides new baseline information on the diversity of T. erytreae, with potential implications for the epidemiology of African Citrus Greening disease.

Figure 1

Mitogenome organization of Trioza erytreae. Circular map of the mitochondrial genome of Trioza erytreae. Protein-coding, transfer RNAs and ribosomal genes are shown with standard abbreviations. The arrows indicate the direction of the genes.

Figure 2

Mitogenome comparisons. Comparison of the mitochondrial sequences of three Trioza erytreae specimens collected in Ethiopia (TE-ETH), Uganda (TE-UG) and South Africa (TE-SA), based on the total complement of 13 protein-coding genes. The number of differences is given as (A) percentage of single nucleotide polymorphisms, and (B) percentage of non-synonymous substitutions, relatively to the size of each gene.

Figure 3

Mitochondrial phylogeny of Triozidae. Maximum likelihood tree representing the phylogenetic relationships within the family Triozidae, using the complete complement of the 13 mitochondrial protein-coding genes. Aphis gossypii and Schizaphis graminum (Aphididae) were used as outgroups. Values represent nodal support calculated from 1,000 bootstrap replicates. The length of the branches is proportional to the number of substitutions per site.

Figure 4

Network of Trioza erytreae haplotypes. Median-joining network of cytochrome c oxidase subunit 1 (COI) gene (607 bp) of Trioza erytreae (n = 89), showing the relationships between haplotypes according to geographic origin and plant host. The size of the circles is proportional to the number of individuals sharing the same haplotype. The length of the branches is proportional to the number of nucleotide substitutions between haplotypes. Numbers along branches indicate the number of substitutions between haplotypes.

Figure 5

Genetic divergence within Trioza species. Intra-specific pairwise distances (K2P) in 31 species of the genus Trioza (Triozidae), based on cytochrome c oxidase 1 (COI) sequences.

Figure 6

NJ tree of Trioza erytreae haplotypes. Neighbour-joining tree representing the relationships among haplotypes of Trioza erytreae collected from citrus and other plant hosts in Africa and Europe. The tree was constructed using a 571-bp alignment of cytochrome c oxidase subunit 1 (COI) sequences. Nodal support was calculated using 1,000 bootstrap replicates. The length of the branches is proportional to the number of substitutions per site.