Host-parasite associations can evolve to any point along a continuum from extremely lethal to so mutually beneficial that neither participant could survive without the other. Trade-off analyses seek to explain the spots along this continuum to which particular associations will evolve, over millions of years or just a few months. The rate depends on many things: what kind of variation exists among the germs, how long it takes to generate variation that would be useful for the germs, the differences in success among the competing germs, and what the host has and can muster as countermeasures. The outcomes may not be stable.

Of particular importance to the outcomes is the dependence of pathogen transmission on host mobility. If a mode of transmission allows pathogens to reach susceptible hosts even when the infected host is entirely immobilized by illness, then we expect natural selection to favor a ravaging disease. Among diseases transmitted by mosquitoes, for example, even very sick individuals can serve as a source of infection because the mosquitoes take care of the transportation. In fact, sick individuals may be even better as sources of infection because they are less able to swat mosquitoes. As expected from this argument, diseases transmitted by mosquitoes, tsetse flies, lice, and sandflies do tend to be more lethal than diseases that rely on person-to-person transmission. This simple trade-off argument explains why the agents of malaria, yellow fever, and sleeping sickness are so much more incapacitating than agents of respiratory diseases, such as the common cold, which typically cause just sneezes, coughs, and runny noses.

Mosquitoes and other organisms that transport pathogens from person to person are called vectors; the diseases they transport are, logically enough, referred to as vector-borne. As a group, vector-borne diseases are particularly well endowed with killers, including malaria, sleeping sickness, and yellow fever. But even those vector-borne pathogens that are not especially lethal tend to be agonizing and immobilizing. The dengue virus belongs in this category. It is a cousin of the yellow fever and West Nile viruses and is transmitted largely by the same mosquito that transmits yellow fever: Aedes aegypti. The dengue virus kills fewer than one of every hundred people it infects, but the low probability of death is little comfort to the dengue patient, as is clear from an account by the tropical-disease expert Alan Spira, describing his own case of dengue, which he acquired in East Africa: "A headache behind the eyes that throbbed and pounded, with a sensation of pressure like a kettle brewing and boiling. A fever, mild at first, but later intense with sweating, came bundled with ferocious muscle aches. These aches were rooted deep in the calves and back, and felt like being punched from the inside-out." Dengue is often called breakbone fever because the pain gives the patient the impression that bones are slowly being broken.

The dengue virus was probably passed to humans from monkeys many centuries ago. Though generally confined to a band within twenty-five degrees latitude from the equator, it has circled the earth within this zone and is found in India, Southeast Asia, Africa, the Caribbean, and Mexico. Using its mosquito transport, it quickly burns through one village and then travels off to another village, town, or even city, wherever the Aedes mosquitoes—and hence the

viruses— have ready access to their human food. The viruses return to repeat the process when the number of susceptible humans in the group increases sufficiently through new births, new immigrants, or the gradual fading of the immunity conferred by previous infection. Dengue is a terrible experience for the afflicted person, yet the incapacitation serves the dengue virus by making the sufferer a more vulnerable target for mosquitoes.

The dependence of transmission on host mobility also explains why some diarrheal diseases are matters of life or death, whereas others are just an annoyance. People can inadvertently create "cultural vectors," which transport pathogens from immobilized hosts much the way mosquitoes transport malaria and dengue. If water supplies are not adequately protected, the washing of clothes and bedsheets can contaminate the water, thereby infecting hundreds of other people, even if the person who contaminated the materials was entirely immobilized with a case of cholera or dysentery. Waterborne pathogens pay a low transmission price when they exploit a person so intensively that the person is completely immobilized, because they can still be

transmitted from immobile hosts; and they gain a big fitness benefit from exploiting infected hosts because contaminated water can contact many more people than an infected person can. Waterborne pathogens are not just limited to an infected person's friends and acquaintances; anyone who drinks contaminated water is a potential victim.

Comparisons of human diarrheal diseases confirm the central prediction of this line of reasoning: the more waterborne the diarrheal bacterium, the more deadly it is. This association explains why cholera, typhoid, and dysentery are so deadly, and the bacteria that infect the

intestines of rich countries are generally so mild. This evolutionary perspective also explains why travel to countries without adequate protection of drinking water is so dangerous, even when such countries do not have notorious vector-borne diseases such as malaria, dengue, and yellow fever. The diarrheal pathogen that enters the traveler's body through contaminated food or drink may have had a long evolutionary history of transmission that has not depended on mobile people. The traveler feels this legacy much more intensely than the local residents because the residents have already generated an immunity. These residents typically paid the price of initiation early in life when, as babies or toddlers, their lives depended on the defenses their immune systems could muster. Those youngsters who did not pass this test joined the three million or so children who are buried in poor countries each year, children who died from something as simple and violent as diarrhea. The traveler isn't as likely to die from the infection, largely because travelers usually have access to the simple but life-saving doses of antibiotics and rehydration solution. The traveler lives to ponder this firsthand experience of life as it is lived now in poor countries and as it was lived just a few generations ago in rich countries. In the nineteenth century almost all urban centers had the same nasty diarrheal pathogens. By allowing drinking water to be fecally contaminated, the technology of the time fostered the distribution of deadly agents. The children of rich and poor countries alike experienced the same lottery that is being held in poor countries today. One in ten children typically succumbed to disease in areas with unprotected water. Whenever water supplies were protected, the most dangerous protagonists vanished as predictably as actors at the end of a play.