Mosquitoes Acquired a Taste for Human Blood 1.8 Million Years Ago - When Homo Erectus Arrived

Among roughly 3,500 known mosquito species, only a small fraction have developed a preference for feeding on humans. That preference, rare as it is in evolutionary terms, has shaped the history of infectious disease more profoundly than almost any other behavioral adaptation in the animal kingdom. Species that actively seek human blood are the principal vectors of malaria, dengue, yellow fever, Zika, and dozens of other pathogens that have killed more people throughout history than any war.

Understanding when and why that preference evolved has direct relevance for controlling mosquito-borne disease - and for interpreting the deep history of human-mosquito co-evolution. A new study published in Scientific Reports tackles this question in one of the mosquito groups most relevant to Asian malaria transmission: the Anopheles leucosphyrus, or Leucosphyrus, group, which includes species responsible for transmitting malaria parasites across Southeast Asia.

The Evolutionary Analysis

The research team, led by Upasana Shyamsunder Singh and Catherine Walton, sequenced DNA from 38 mosquitoes representing 11 species within the Leucosphyrus group. The specimens were collected between 1992 and 2020 from sites across Southeast Asia. Using those sequences, computer models, and calibrated estimates of DNA mutation rates, the team reconstructed the evolutionary history of the group - estimating not just how the species are related, but when key events in their diversification occurred.

The central question was when, and in what geographic region, the preference for feeding on humans first appeared within the group. Ancestral Leucosphyrus mosquitoes fed on non-human primates. At some point, that preference shifted toward humans - a change that required multiple genetic changes in the olfactory receptors used to detect body odor, since human scent profiles differ substantially from those of other great apes.

The Timing and Location

The analysis estimated that the preference for human-feeding evolved once within the Leucosphyrus group, between 2.9 and 1.6 million years ago. The geographic location of that transition was Sundaland - the ancient landmass that during periods of lower sea levels encompassed what is now the Malay Peninsula, Borneo, Sumatra, and Java.

That timing overlaps with what the archaeological and paleontological record suggests about when early hominins first reached the region. Homo erectus, the first hominin species to leave Africa and spread across Asia, is estimated to have arrived in Southeast Asia approximately 1.8 million years ago. Modern humans did not reach the region until between 76,000 and 63,000 years ago - hundreds of thousands of years after the mosquito preference for human feeding had already been established.

The implication is that these mosquitoes adapted to humans before modern humans were available to adapt to. The most plausible explanation is that the arrival of H. erectus in Sundaland provided a novel, abundant, and physically accessible food source that selected for mosquitoes capable of detecting and preferring human odor over the primate odors they had previously targeted.

Independent Evidence for Early Hominin Arrival

The hominin fossil record for Southeast Asia is fragmentary. Most of what we know about H. erectus in the region comes from a relatively small number of sites - principally in Java - and the dating of early arrival has been subject to ongoing revision and debate. The mosquito evolutionary analysis provides what the authors describe as independent non-archaeological evidence that hominins were present in Sundaland in substantial numbers around 1.8 million years ago.

This is a novel use of vector biology as a proxy for ancient human population history. If the evolution of human-feeding preference requires a threshold density of human hosts - sufficient to make seeking human odor reliably more rewarding than seeking primate odor - then the timing of that evolutionary shift places a minimum bound on when human populations were large enough to exert that selective pressure in the region.

Comparison With African Malaria Vectors

The Leucosphyrus findings can be compared with published estimates for the African mosquito lineage that gave rise to Anopheles gambiae and Anopheles coluzzii - the primary malaria vectors in sub-Saharan Africa and among the most dangerous disease vectors in the world. That human-feeding preference is estimated to have evolved between 509,000 and 61,000 years ago, a much more recent transition than the Leucosphyrus shift and one that postdates the emergence of anatomically modern humans in Africa.

The Leucosphyrus timeline, by contrast, predates modern humans by more than a million years. If the analysis is correct, the human-feeding preference in Southeast Asian mosquitoes is among the oldest evolutionary adaptations to human hosts known in any disease vector species.

Limitations

Molecular clock estimates are sensitive to the calibration rates used and to the completeness of the phylogenetic sampling. The 38-mosquito sample, while spanning 11 species, represents a subset of the full Leucosphyrus diversity. Specimens collected between 1992 and 2020 from varying field conditions may introduce sampling biases that affect the phylogenetic reconstruction. The date range of 2.9 to 1.6 million years ago reflects genuine uncertainty in the estimate rather than a precise determination.

The proposed mechanism - H. erectus arrival driving selection for human-feeding preference - is consistent with the data but cannot be directly tested from molecular phylogenetics alone. It remains a hypothesis supported by temporal and geographic overlap rather than a demonstrated causal chain.