Free-flying hematophagous insects must have flourished during the warm interglacials in the Pleistocene. Large herbivorous mammals dominated the landscape, and hematovores probably required few specialized adaptations for contacting suitable hosts. However, events following the last glacial period, approximately 10,000 years ago, set in motion changes that put new pressures on blood-feeding insects. Mass extinctions of large land mammals coincided with expanding human populations. Nomadic, hunter-gatherer societies were turning to agriculture, and several wild animals gradually were being tamed for use. The vanishing supply of large mammals probably accelerated insects’ development of more efficient, but not necessarily more specific, mechanisms for finding blood. Modern mosquitoes still seem to rely on varying combinations and levels of general host cues instead of unique chemicals. More highly refined sensory mechanisms were needed for locating small, cryptic hosts.

In general, feeding specializations beyond those suited for a specific habitat have been slow to develop. The depauperate nature of vertebrate populations throughout postglacial times must be one of the main reasons most mosquitoes continue to depend on general cues and feed on a broad range of hosts. Data on facultative autogeny and field engorgement rates support this argument.

Surface water, a requirement of both vertebrates and mosquitoes, may have served as an environmental magnet that regularly brought hosts and mosquitoes into close contact, especially when water and hosts were scarce. Species with highly restricted larval habitats had to abandon blood feeding altogether or adapt to smaller hosts whose home ranges overlapped their own restricted habitat. Small hosts had less blood to share with parasites, and consequently they developed behaviors for effectively limiting blood feeding. Mosquitoes that feed on small hosts generally have lower population densities and behaviors for successfully penetrating and surviving hosts’ attempts to repel them. Examples of such adaptive behavior are presented, and it is proposed that most vectors that successfully feed on humans are derived from rodent- or bird-feeding ancestors.

As humans continued to dominate the environment, those mosquitoes that adapted to human-altered larval habitats were increasingly successful. Humans were often the principal blood source in these habitats. Domestic livestock became important hosts when introduced, but their disappearance during times of food shortage or disease stress made them less reliable than humans. In the future, even more efficient human-feeding species are apt to evolve unless steps are taken to discourage host behaviors that encourage frequent contact. The reduction of vector-borne diseases by promoting human behavioral changes is a strategy that deserves more emphasis.