Carl Zimmer wrote a blog post and an article in the NYTimes about this study on malaria. I have actually read the malaria paper on PLoS before I checked The Loom, as I find this disease to be a source of seemingly endless fascinating stories of creative scientific research.
In this study, three boys - one with malaria and two without - were placed in tents at night. Mosquitoes (lab-raised and parasite-free) were then released into a chamber positioned in between the three tents. Mosquitoes had four choices: to remain in the chamber or to fly through plastic tubes into one of the three tents.
And although none of the boys was having a malarial fever at the time of the experiment, thus eliminating raised temperature as a clue, mosquitoes preferentially chose to fly into the tent of the sick boy. The paper did not look at feverish kids, but I bet that attractiveness rises even more during fever (CO2, heat and some chemicals in sweat are all mosquito attractants). Once the sick kids were treated from malaria and cleared the Plasmodium gametocytes from their system, they lost attractiveness - mosquitoes chose each of the three boys equally (perhaps they even found the ex-sick boy a tad bit less attractive, if anything).
So, the mosquitoes can cue onto something - an odor - that emanates from the sick people and not from the malaria-free people. It is not known what that odor might be.
It is not absolutely clear from the paper, though, at exactly what time of night was the experiment conducted and what levels of gametocytes were present in the boys' blood at the exact time of the mosquito test. And this is an important question as the Plasmodium gametocytes burst out of red blood cells at night at precisely the same time when the mosquitoes are most active.
Gametocytes do not spend much time in blood plasma. They hide from the host's immune system by entering and remaining within the red blood cells. Explosive burst at a precise time of night results in so many individual gametocytes appearing symultaneously in the bloodstream that the immune system is overwhelmed, at least long enough while the mosquitoes are biting.
Plasmodium "knows" exactly when to erupt out of blood cell by tracking melatonin secretion in the host. Incidentally, Plasmodium has a melatonin receptor that is on the membrane and utilizes Calcium as the second messenger (fast non-genomic response), unlike receptors in animals which are nuclear (slow genomic response). In the host, melatonin rises in the evening and the Plasmodium, using its own circadian clock, "calculates" how long to wait until eruption.
In different geographic regions, mosquitoes are active at different times and it appears that Plasmodium has evolved to match the timing of the local mosquito population. What a fascinating story of temporal co-evolution of the parasite, its vector and its host!