Bemisia tabaci whitefly species are some of the world’s most devastating agricultural pests and plant-virus disease vectors. Elucidation of the phylogenetic relationships in the group is the basis for understanding their evolution, biogeography, gene-functions and development of novel control technologies. We report here the discovery of five new Sub-Saharan Africa (SSA) B. tabaci putative species, using the partial mitochondrial cytochrome oxidase 1 gene: SSA9, SSA10, SSA11, SSA12 and SSA13. Two of them, SSA10 and SSA11 clustered with the New World species and shared 84.8‒86.5% sequence identities. SSA10 and SSA11 provide new evidence for a close evolutionary link between the Old and New World species. Re-analysis of the evolutionary history of B. tabaci species group indicates that the new African species (SSA10 and SSA11) diverged from the New World clade c. 25 million years ago. The new putative species enable us to: (i) re-evaluate current models of B. tabaci evolution, (ii) recognise increased diversity within this cryptic species group and (iii) re-estimate divergence dates in evolutionary time.

Over the past three decades, highly increased whitefly (Bemisia tabaci) populations have been observed on the staple food crop cassava in eastern Africa and associated with ensuing viral disease pandemics and food insecurity. Increased whitefly numbers have also been observed in other key agricultural crops and weeds. Factors behind the population surges on different crops and their interrelationships are unclear, although in cassava they have been associated with specific populations within the Bemisia tabaci species complex known to infest cassava crops in Africa. This study carried out an in-depth survey to understand the distribution of B. tabaci populations infesting crops and uncultivated plant hosts in Uganda, a centre of origin for this pest complex. Whitefly samples were collected from 59 identified plant species and 25 unidentified weeds in a countrywide survey. Identities of 870 individual adult whiteflies were determined through mitochondrial cytochrome oxidase 1 sequences (651 bp) in the 3′ barcode region used for B. tabaci systematics. Sixteen B. tabaci and five related whitefly putative species were identified based on > 4.0% nucleotide divergence, of which three are proposed as novel B. tabaci putative species and four as novel closely related whitefly species. The most prevalent whiteflies were classified as B. tabaci MED-ASL (30.5% of samples), sub-Saharan Africa 1 (SSA1, 22.7%) and Bemisia Uganda1 (12.1%). These species were also indicated to be the most polyphagous occurring on 33, 40 and 25 identified plant species, respectively. Multiple (≥ 3) whitefly species occurred on specific crops (bean, eggplant, pumpkin and tomato) and weeds (Sida acuta and Ocimum gratissimum). These plants may have increased potential to act as reservoirs for mixed infections of whitefly-vectored viruses. Management of whitefly pest populations in eastern Africa will require an integration of approaches that consider their degree of polyphagy and a climate that enables the continuous presence of crop and uncultivated plant hosts.

Insect–plant associations and their role in diversification are mostly studied in specialists. Here, we aimed to identify macroevolution patterns in the relationships between generalists and their host plants that have the potential to promote diversification. We focused on the Bemisia tabaci species complex containing more than 35 cryptic species. Mechanisms for explaining this impressive diversification have focused so far on allopatric forces that assume a common, broad, host range. We conducted a literature survey which indicated that species in the complex differ in their host range, with only few showing a truly broad one. We then selected six species, representing different phylogenetic groups and documented host ranges. We tested whether differences in the species expression profiles of detoxification genes are shaped more by their phylogenetic relationships or by their ability to successfully utilize multiple hosts, including novel ones. Performance assays divided the six species into two groups of three, one showing higher performance on various hosts than the other (the lower performance group). The same grouping pattern appeared when the species were clustered according to their expression profiles. Only species placed in the lower performance group showed a tendency to lower the expression of multiple genes. Taken together, these findings bring evidence for the existence of a common detoxification “machinery,” shared between species that can perform well on multiple hosts. We raise the possibility that this “machinery” might have played a passive role in the diversification of the complex, by allowing successful migration to new/novel environments, leading, in some cases, to fragmentation and speciation.

The genetic variability of whitefly (Bemisia tabaci) species, the vectors of cassava mosaic begomoviruses (CMBs) in cassava growing areas of Kenya, Tanzania, and Uganda, was investigated through comparison of partial sequences of the mitochondria cytochrome oxidase I (mtCOI) DNA in 2010/11. Two distinct species were obtained including sub‐Saharan Africa 1 (SSA1), comprising of two sub‐clades (I and II), and a South West Indian Ocean Islands (SWIO) species. Among the SSA1, sub‐clade I sequences shared a similarity of 97.8–99.7% with the published Uganda 1 genotypes, and diverged by 0.3–2.2%. A pairwise comparison of SSA1 sub‐clade II sequences revealed a similarity of 97.2–99.5% with reference southern Africa genotypes, and diverged by 0.5–2.8%. The SSA1 sub‐clade I whiteflies were widely distributed in East Africa (EA). In comparison, the SSA1 sub‐clade II whiteflies were detected for the first time in the EA region, and occurred predominantly in the coast regions of Kenya, southern and coast Tanzania. They occurred in low abundance in the Lake Victoria Basin of Tanzania and were widespread in all four regions in Uganda. The SWIO species had a sequence similarity of 97.2–97.7% with the published Reunion sequence and diverged by 2.3–2.8%. The SWIO whiteflies occurred in coast Kenya only. The sub‐Saharan Africa 2 whitefly species (Ug2) that was associated with the severe CMD pandemic in Uganda was not detected in our study.

Projected climate changes are thought to promote emerging infectious diseases, though to date, evidence linking climate changes and such diseases in plants has not been available. Cassava is perhaps the most important crop in Africa for smallholder farmers. Since the late 1990’s there have been reports from East and Central Africa of pandemics of begomoviruses in cassava linked to high abundances of whitefly species within the Bemisia tabaci complex. We used CLIMEX, a process-oriented climatic niche model, to explore if this pandemic was linked to recent historical climatic changes. The climatic niche model was corroborated with independent observed field abundance of B. tabaci in Uganda over a 13-year time-series, and with the probability of occurrence of B. tabaci over 2 years across the African study area. Throughout a 39-year climate time-series spanning the period during which the pandemics emerged, the modelled climatic conditions for B. tabaci improved significantly in the areas where the pandemics had been reported and were constant or decreased elsewhere. This is the first reported case where observed historical climate changes have been attributed to the increase in abundance of an insect pest, contributing to a crop disease pandemic.

In sub‐Saharan Africa cassava growing areas, two members of the Bemisia tabaci species complex termed sub‐Saharan Africa 1 (SSA1) and SSA2 have been reported as the prevalent whiteflies associated with the spread of viruses that cause cassava mosaic disease (CMD) and cassava brown streak disease (CBSD) pandemics. At the peak of CMD pandemic in the late 1990s, SSA2 was the prevalent whitefly, although its numbers have diminished over the last two decades with the resurgence of SSA1 whiteflies. Three SSA1 subgroups (SG1 to SG3) are the predominant whiteflies in East Africa and vary in distribution and biological properties. Mating compatibility between SSA1 subgroups and SSA2 whiteflies was reported as the possible driver for the resurgence of SSA1 whiteflies. In this study, a combination of both phylogenomic methods and reciprocal crossing experiments were applied to determine species status of SSA1 subgroups and SSA2 whitefly populations. Phylogenomic analyses conducted with 26 548 205 bp whole genome single nucleotide polymorphisms (SNPs) and the full mitogenomes clustered SSA1 subgroups together and separate from SSA2 species. Mating incompatibility between SSA1 subgroups and SSA2 further demonstrated their distinctiveness from each other. Phylogenomic analyses conducted with SNPs and mitogenomes also revealed different genetic relationships among SSA1 subgroups. The former clustered SSA1‐SG1 and SSA1‐SG2 together but separate from SSA1‐SG3, while the latter clustered SSA1‐SG2 and SSA1‐SG3 together but separate from SSA1‐SG1. Mating compatibility was observed between SSA1‐SG1 and SSA1‐SG2, while incompatibility occurred between SSA1‐SG1 and SSA1‐SG3, and SSA1‐SG2 and SSA1‐SG3. Mating results among SSA1 subgroups were coherent with phylogenomics results based on SNPs but not the full mitogenomes. Furthermore, this study revealed that the secondary endosymbiont—Wolbachia—did not mediate reproductive success in the crossing assays carried out. Overall, using genome wide SNPs together with reciprocal crossings assays, this study established accurate genetic relationships among cassava‐colonizing populations, illustrating that SSA1 and SSA2 are distinct species while at least two species occur within SSA1 species.