In a Prisoner's Dilemma game the converse of cooperation is defection. Not surprisingly, however, cooperation and defection can viewed instead as existing on a sliding scale, where the impact on a given player, one that is the recipient of an interaction, can range from "very positive" to "very negative". In this case, since Prisoner's Dilemmas are built upon relative (versus absolute) payoff values – that is, they are defined in terms of payoff-value inequalities – the impact on an organism of being cooperated with could even be "very negative" while the corresponding defection impact would instead be "very very negative", with both payoffs in this case as viewed relative to the organism's payoff absent participation in the corresponding social interaction. Cooperative interactions thus need not be strictly about improving fitness above some normal baseline but, instead, can simply about not having as great a negative impact as the alternative behavior.

The Prisoner's Dilemma, despite this potentially unintuitive idea that cooperation can exist simply as a behavior of less-defection, nonetheless is a relatively simple game. This is to say that even though one player's payoff is dependent on the behavior of the other player – which certainly is a complication in comparison with payoffs that are completely independent of the actions of others – a more complicated game would occur were one player's payoff dependent also on their opponent's payoff. In such a situation there may exist a defection or instead cooperation optimum that balances direct gains (i.e., those observed within a Prisoner's Dilemma-like game) with more indirect gains where one's gains are also a function of one's opponent's gains. In other words, it is possible to have situations in which fitness is a function not just one's behavior and the behavior of one's opponents but also a consequence of the impact of one's behavior on one's opponent's fitness.

One "game" with an approximation of such qualities is associated with the concept of inclusive fitness, where the fitness gain of a particular allele, following some interaction with other players, is coupled to the fitness of that allele as found in all individuals that happen to be carrying that allele. In this case, to calculate fitness gains one must consider a combination of the cost to the focus player of performing a given act, such as cooperation, in combination with the benefit to the recipient player associated with interacting with that focus player, with all costs and benefits, in this case, considered from the perspective of individual alleles rather than necessarily individual organisms/players. By way of example, helping out one's family, that is, individuals with whom we tend to share alleles, makes evolutionary sense even if the helpful act itself is personally costly because ultimately an allele may be helping out itself as found in a separate individual or body. Optimization of inclusive fitness thus takes into account a combination of the payoffs obtained by both a player and those obtained by its opponent, with selection as a consequence potentially favoring mutual cooperation rather than unilateral defection (i.e., if both player's share a given allele then an allele gains from receiving cooperation in both directions, from player A to player B as well as from player B to player A). As indicated, mutual cooperation would represent an expected outcome of the interactions among relatives, i.e., individuals who are more likely to share alleles and who also, perhaps as a consequence, are more likely to participate in inherently friendly interactions.

What happens if a game retains this complication of the fitness of one player being dependent, in part, on the fitness of other players, but where payoffs are associated instead with inherently antagonistic interactions, that is, rather than potentially cooperative ones? One result is something that we might describe as a Parasite's Dilemma: A parasite's fitness in particular can vary in part as a function of its host's health, which in turn is controlled by a parasite's virulence. That is, the "Parasite's Dilemma" is a Prisoner's Dilemma-like game, one in which the parasite is one player and the host is the other player. As a complication, however, the host's health can be measured in terms of host fitness, on the one hand, but also in terms of its impact on parasite fitness, with reductions in host fitness potentially resulting in reductions in parasite fitness, even though the parasite, as a parasite, is defined as a symbiont that acts to reduce host fitness.

Though not between relatives and therefore not involving inclusive fitness, nonetheless the fitness associated with parasite alleles in other words can be a function not just of what those alleles do directly for the carrying parasite but also in terms of how those alleles affect the parasite indirectly, i.e., by first affecting the host. For example, and at an extreme, a parasite's fitness often will decline to zero should the parasite kill its host. Or, from the perspective solely of alleles, an allele that encodes too-early termination of its infection of a host, i.e., as by killing the host, may display a fitness disadvantage even if those same alleles otherwise give rise directly to more direct positive fitness benefits (the concept of antagonistic pleiotropy – where an allele is helpful within one context but harmful within another – should at this point be springing to mind).

In this chapter I begin our explorations of the concept of parasite virulence by fleshing out this idea of a Parasite's Dilemma. Note that I am using the word "Parasite" in its broader sense, referring not just to eukaryotic infectious agents, such as protozoa, helminthes, or even ectoparasitic arthropods, but also including, perhaps especially, pathogenic bacteria, fungi, and viruses, i.e., symbionts that negatively impact host organisms. The evolution of a parasite's virulence, that is, can evolve, and an important aspect of that evolution, particularly in terms of how it potentially can lead to lower rather than inevitably higher levels of parasite virulence, is captured by this game of parasite-host interaction that I refer to as a Parasite's Dilemma.


Table: Important Terms and Concepts relevant to Issues of Virulence Evolution.

Antagonistic pleiotropy

Improvement in the utility of one aspect of phenotype in one context that results also in a decline in utility in a different context.
In terms of virulence evolution, antagonistic pleiotropy can take the form of conflicts between life stages, particularly within-host population growth versus transmission between hosts. Alternatively, it can take the form of specialization for specific hosts at the expense of potential to replicate on other hosts. In either case, these conflicts could serve to constrain parasite evolution of greater virulence.

Note, though, that a pleiotropy strictly is more than one phenotype associated with a single gene or allele, and therefore these conflicting phenotypes basically are tradeoffs where greater functionality in one context as associated with one allele – versus alternative alleles that can be found at the same locus – comes at the cost of reduced functionality in a different context, again relative to these other alleles. Therefore, in terms of virulence evolution, an antagonistic pleiotropy is seen given an allele that improves especially one of a parasite's life stages at the expense of a different life stage displayed by that same parasite, such as increased population growth within hosts at the expense of transmissibility between hosts, or instead improvement in an ability to exploit one host type at the expense of an ability to exploit alternative host types.

Interspecific interaction in which one organism gains whereas the other organism neither gains nor loses.
In terms of virulence evolution, a commensalistic relationship is usually taken to be one in which symbionts do not harm the host organism. It is at least theoretically possible, however, for commensalism to instead mean 'no net harm', such that a host is harmed but the commensal organism makes up for this harm by also providing the host with benefits. Generally, and nonetheless, evolution of parasites towards commensalism should be viewed, unless there is reason not to, as evolution towards causing the host less harm, that is, towards reduced anti-host virulence. Such evolution in principle may or may not result in improvements in symbiont fitness. That is, in some symbionts, in some contexts, evolution towards commensalism (lower virulence) may be selectively beneficial whereas for other symbionts and/or in other contexts evolution away from commensalism (greater virulence) instead may be selectively beneficial.

Absence of health.
In evolutionary terms, an absence of health can be quantified in terms of the fitness of the affected organism. That is, greater disease explicitly is associated with greater reductions in fitness. Diseases of course vary in their causes as well as their severity and a subset of diseases are described as infectious, that is, those diseases that are caused by parasites. In this case, it is important to keep in mind that not only can the host organism be affected by reductions in health, and not only can this impact host fitness, but parasites also can be affected by host health. With parasites, however, depending on type as well as circumstances, reductions in host health may serve to increase, decrease, or indeed not modify fitness at all. That is, circumstances can exist where decreasing host health can be beneficial to a parasite but also there can exist circumstances where a parasite benefits by either not or instead less substantially decreasing host health.
Evolution of virulence

  (alt link)
Changes in frequencies of parasite alleles that impact the health of host organisms.
The assumption here is that virulence is at least partly under genetic control by parasites and therefore can be modified upward (more virulence) or downward (less virulence) depending upon how the level of virulence displayed by a parasite affects parasite fitness. These issues can be complicated as well by conflicts between fitness gains over shorter versus longer time scales. Thus, it is entirely possible for changes in parasite fitness to result in fitness gains over shorter time scales but to so at the expense of fitness over longer time scales, or indeed vice versa. This issue is explicitly one of antagonistic pleiotropy. Generally, though, it often is increases in parasite virulence that tend to result in shorter term fitness gains at the expense of longer-term fitness, and it particularly is when these two issues are less in conflict (less pleiotropic antagonism) that greater parasite virulence will tend to evolve.
Game of Chicken

Interaction between two or more individuals in which mutual defection is associated with a worse outcome for both individuals than unilateral cooperation for the cooperator.
In Games of Chicken, unilateral defection is preferable to mutual cooperation which is preferable to unilateral cooperation, and which in turn is preferable to mutual defection. Thus, T > R > S > P. In the classic, car-based Game of Chicken, mutual defection would be a head-on collision. That is, not swerving could be described as defection (D) and swerving as cooperation (C). Unilateral not swerving, i.e., unilateral defection (T) can result in positive enhancement of one's reputation for courage whereas swerving instead can bring shame with associated negative hits to one's reputation for courage, and perhaps a greater negative impact given unilateral swerving (S) versus mutual swerving (R). The head-on collisions that are a consequence of mutual not swerving (P) by contrast can result in death, or at least severe damage to body and vehicle. Compare with Prisoner's Dilemma, i.e., where T > R > P > S, that is, where mutual defection is instead preferable to unilateral cooperation.
Hard selection

Additional levels of mortality experienced by a population that can result in population extinction absent successful adaptation.
In terms of parasite virulence, hard selection is that imposed especially by the parasite's potential to be transmitted to a new host. That is, a parasite that is unable to acquire new hosts will surely go extinct just as withholding key resources from any population such as nutrients or oxygen also can surely result in extinction. Similarly, however, a population may be able to respond to losses of access to new hosts (as a resource), e.g., such as through selection for alleles that serve to counter the hard selection imposed by reductions in new host availability. These selected alleles may serve to change what hosts are being sought, how new hosts are sought, or instead, and particularly if rather than complete loss of hosts there is only a reduction in potentially available new hosts, in changes simply in the timing or other aspects of otherwise normal approaches a given parasite type might employ to acquire new hosts.

Thus, for example, responses to the hard selection of reduced host availability could be a longer period over which an infected host remains infectious, an increase in the ability of infected hosts to come into contact with potential new hosts (such as by inflicting less disease on the infected host through a display of lower virulence), or instead by the parasite displaying greater durability outside of the infected host, that is, to aid in the potential for transmission to a new host to be successful following the movement of a parasite out of an infected host and prior to its acquisition of a new host to infect. Hard selection also – perhaps most notably and with less subtlety – can be imposed by the host's immune system, though it is important to keep in mind that the host combating an infecting parasite is only one component of multiple factors that can affect parasite transmission ability and thereby parasite fitness.

Presence of organism functionality.
In terms of virulence evolution, health can be viewed most productively as equivalent to the 'health' of a parasite's 'environment', and particularly in terms of how that health affects parasite fitness, such as in terms of parasite potential to be transmitted to new hosts. Thus, though health in a medical sense is viewed solely from the perspective of host functionality for host needs, from the perspective of the evolution of virulence, host health instead is viewed from the perspective of host functionality for parasite needs. Key issues with regard to the interface between host health and parasite fitness are duration of host infectiousness and the extent to which the specifics host health – e.g., whether or not the host remains ambulatory or even alive – affects, either negatively or positively, the potential for parasite transmission to new hosts.
Host availability

  (alt link)
Product particularly of density along with susceptibility of target organisms for potential symbiont colonization.
Note that "Product" in the definition is being used in terms of times/multiplication, as in host density × host susceptibility. In addition to host availability, also relevant is the potential for symbionts to span any spatial gaps between hosts, and the time required to span those gaps along with symbiont durability outside of hosts. In other words, all factors associated with the transmission of a symbiont, such as a parasite, that come into play especially following the leaving the host organism it currently occupies play roles that are similar, in terms of parasite ecology and evolution, to the issue of host availability.

Number of pathogens that individual would-be hosts must be exposed to such that on average half will become infected.
In terms of virulence evolution, the concept of ID50 is relevant because it is suggestive either of a requirement for multiple infectious agents to initiate infections, for infections to occur, or instead that on average only a single, individual parasite or pathogen succeeds in infecting at any given time while the odds of infection by any specific, individual parasite or pathogen is low. The existence of an ID50 that is somewhat greater than one thus, in at least some cases, may suggest that many infectious diseases are caused by clonal populations of pathogens, that is, initiated by just a single organism, perhaps despite initial exposure to multiple, individual parasite organisms. Clonality of populations infecting individual host organisms in turn, and importantly towards the evolution of virulence, can increase the potential for the display of mutual cooperation by parasite populations infecting a single host. Such cooperation can result in a display of greater parasite virulence but more generally can be crucial towards the evolution of reduced parasite virulence, i.e., as towards an optimal level of virulence.
Inclusive fitness

Description of the reproductive success of alleles as found in multiple individuals rather than just a single, otherwise unaffiliated, allele-carrying individual.
Relevant here is whether an altruistic trait within a parasite could evolve that has the effect of increasing parasite transmission potential from the same host, especially of clonally related parasites. Thus, for example, a parasite could restrain its within-host replication so as to reduce overall host virulence and thus potentially allow for an increased overall potential for parasite transmission from a host. Such efforts may be less beneficial to an altruism-effecting allele, however, if the parasite population is not clonal and, therefore, not all parasite individuals necessarily carry the altruism-encoding allele. One route towards avoidance of a Tragedy of the Commons, in other words, is for alleles encoding greater exploitation of the commons – or in this case, the host – to have the effect of reducing the fitness of other parasite individuals, over longer time frames, that happen to carry the same alleles.

Such over exploitation may instead be selected for, however, if the rest of a population does not carry the same alleles to the extent that over exploitation can enhance within-host parasite fitness. Of course, if over exploitation and/or expedience also provides a selective benefit to the over exploiters in terms of overall transmission potential (between-host selective benefit) then over exploitation could be selected in terms of its impact on transmission as well. Keep in mind with the latter musings, though, that at some point increasing levels of host exploitation must result in declines in individual parasite fitness, with an extreme being levels of anti-host virulence that are so severe that the host is killed prior to a parasite gaining the potential to be transmitted.

Interspecific interaction in which both species benefit.
Though as a rule parasites are not mutualists, it is entirely possible for microorganisms to be facultative in their parasitism and therefore commensals or mutualists under certain circumstances, or while infecting certain hosts, and parasites under other circumstances or while affecting other hosts. Nonetheless, generalizations aside, most parasites likely are not also mutualists.

Organism that lives in or on other organisms from which it is stealing resources and to which it is causing harm.
The term parasite, unfortunately, has different meanings depending upon circumstances. The narrower meaning describes especially eukaryotic organisms that cause disease (e.g., the parasite, Plasmodium, the cause of malaria). These parasites, narrowly defined, include especially both parasitic protozoa and helminths (i.e., parasitic worms). As the qualifier "parasitic" should suggest, a parasite also can more generally be described simply as a disease-causing organism. Pathogens thus represent one category of parasites, and generally in this chapter it is this second, broader meaning – parasites as disease-causing organisms – that is intended. Lastly, and even more broadly defined, the harm that a parasite causes might not be described necessarily as a "disease" and indeed a symbiont that reduces the fitness of a host organism in any manner (e.g., increasing the weight of the host such that the host is not able to run as rapidly) would be considered to be a parasite.

Nonetheless, the damage or harm done be a symbiont to a host must constitute a net damage for that symbiont to be considered to be a parasite since even mutualistic symbionts at some level may cause "harm" to a host organism, but nevertheless (and by definition for mutualisms) the symbiont also may supply a benefit to the host that makes up for the damage caused. Indeed, even parasites can supply benefits to their hosts, but nevertheless for parasites (again, by definition) the level of benefits supplied to the host organisms should not exceed the level of damage inflicted on the host. Parasites thus damage their hosts more than they help their hosts whereas mutualistic organisms help their hosts more than they damage their hosts.
Parasite load

Measure of the host-associated density as well as per-individual virulence associated with symbiotic organisms.
The higher the per-symbiont virulence in combination with the more symbionts present within or on a host organism, then the higher the parasite load. Of course, for the symbiont to display virulence at all, particularly net virulence, it pretty much by definition is a parasite. Therefore, this is the load of parasites associated with an individual host where the higher the parasite load then the higher the impact that population of parasites collectively have on host fitness. If per parasite virulence were constant, along with host size, then the parasite load would just be a function of just the number of parasites present per individual host. Alternatively, if per host parasite number were constant, again along with host size, then the parasite load would just be a function of just the virulence associated with the individual parasites present per individual host.
Parasite's Dilemma

A problem that can arise when defection by a symbiont against a host can increase the symbiont's fitness in the short term while decreasing the symbiont's fitness over longer time frames.
A Parasite's Dilemma occurs within the context of a potential parasite-associated Tragedy of the Commons where the commons is the host's body and the tragedy is a curtailment of a parasite's potential to be transmitted to a new host, or alternatively as considered equivalently as an n-player Prisoner's Dilemma. Note that the Parasite's Dilemma may be viewed as an example of an antagonistic pleiotropy since gains in the short term, as equivalent to a life stage (e.g., rates of within-host replication), come at the expense of gains in the long term (such as transmission opportunity). The Parasite's Dilemma basically is a conflict between parasite defection and cooperation, particularly as viewed within a parasite population, or individual selection versus group selection, or short-term fitness versus long-term fitness, or simply the benefits of displaying increased virulence versus associated costs.

Disease-causing microorganism, and particularly a disease-causing bacterium or virus.
"Pathogen" is a narrower concept than that of parasite, at least when the latter is defined broadly to include not just eukaryotic organisms but pathogenic bacteria and viruses as well. Nonetheless, it can be helpful to consider the evolution of virulence in terms of pathogen evolution, if only for the sake of intuitive visualization. This intuiveness, however, is only to the extent that one is more familiar with bacterial or viral disease-causing agents versus protozoa or helminths that cause diseases. Note that parasitic fungi are also described as pathogens, so thus more inclusively a pathogen is a diseasing-causing bacterium, virus, or fungus.

Killing of an organism for the sake of assimilating the nutrients making up its body.
Predation can be viewed, like parasitism, as a variation on victim-exploiter interactions, where the victim (the prey) is, as noted, both killed and assimilated by the exploiter organism (the predator). Crucially towards consideration of issues of the evolution of virulence, the link between prey health – particularly as this can change following contact with the predator – and predator fitness can be nonexistent. As a consequence, there may be no evolutionary incentive for the predator to maintain prey health versus incentive for the predator to reduce prey health. As a consequence, predators display what can be viewed as an extreme of virulence, that is, they can benefit even if prey organisms are killed instantaneously upon contact with the predator. A spectrum of levels of overall, that is, net virulence thus can be envisaged with predators display the highest degree of virulence, parasites displaying lower levels of virulence (they have to keep prey organisms alive over some interval following contact), commensals display no net virulence, and mutualists display what might be described as "negative" virulence towards their hosts.
Prisoner's dilemma

Game of cooperation and defection between two or more individuals in which especially unilateral cooperation provides a worse payoff than mutual defection.
In terms of parasites, defection represents especially mechanisms that increase the prevalence of parasites within a host but at the expense of the overall potential for parasites to successfully transmit their progeny to new hosts. This especially can be seen in terms of responses to soft selection where evolution towards expedience can be viewed as defection whereas evolution toward economy can be viewed as cooperation, and especially to the extent that economy is associated with greater potentials for transmission to new hosts by the parasite population as a whole and expedience instead results in a reduced potential for transmission. Thus, coming to dominate a population through expedient growth, a population that otherwise displays less-expedient growth strategies, can result in greater numbers of transmitting progeny within a context where transmission is still somewhat optimal. The payoff here would be T for unilateral defection. Alternatively is the mutual cooperation of mutually non-expedient growth (R). Mutual defection consisting of all or at least majority of expedient-strategy individuals could result in overall lower transmission opportunities (P), and displaying less expedient growth strategies within a population of expedient defectors would be the worst of all (S). Thus, T > R > P > S, which is a Prisoner's Dilemma.
Soft selection

Deterministic evolution that can result in increases in the relative fitness of certain individuals but not the absolute fitness of associated populations.
In terms of virulence evolution, soft selection impacts especially the frequency of alleles within populations of parasites that are associated with individual hosts. The impact of selection acting within hosts on the frequency of alleles – by definition for soft selection – cannot positively impact parasite fitness in terms of general ability to transmit, since that would result in increases in the average fitness of a parasite population, which instead is the province of hard selection. An allele or genotype displaying increased prevalence within a parasite population found within a given host, as a result of this increased prevalence, may have greater opportunity to transmit in comparison with other parasite alleles also found within a given host, but not a greater ability to be transmitted on a per parasite-individual basis. Indeed, this latter ability might actually decline as a function of increases in within-host competitive ability should that greater within-host competitiveness result in parasite virulence that is greater than what otherwise would be optimal.
Tradeoff hypothesis

  (alt link)
Idea that there can be costs to parasites especially in their displaying excessive virulence towards their hosts.
In other words, and particularly, excessive virulence might interfere with parasite transmission opportunity. This may be because the parasite curtails transmission opportunities by killing its host or instead curtails transmission opportunities by interfering with host movement and/or socialization with conspecifics. Countering such selection against excessive parasite virulence would be within-host soft selection as outlined above. Together this issues help to define what I describe here as the "Parasite's Dilemma", that is, potential conflicts between within-host gains in fitness and between-host competitive ability in terms of transmission ability. See too the idea of virulence optimization.
Tragedy of the Commons

Multiplayer game in which cheating behavior has the effect of reducing the ability of an environment to support the wellbeing of all players.
For parasites the host organism can be viewed as the commons and disease the tragedy, though note that the tragedy in this case is degradation of the host environment in terms of parasite functionality rather than necessarily declines in host functionality in terms of host fitness.

Movement of especially symbiotic organisms from one host organism to another.
Given that host organisms are not immortal, survival of a parasite lineage – or symbiont lineage more generally – is dependent on parasite spreading of its kind to new hosts. This issue is even more true to the degree that symbionts, or parasites, negatively impact host mortality. Generally the fitness of an obligate symbiont, as measured over longer time scales, is a function of symbiont transmissibility. This then gives rise to the tradeoff hypothesis of virulence evolution as well as conflicts between soft and hard selection also during virulence evolution, virulence optimization, etc., with natural selection generally seeking over longer time scales those symbiont characteristics which serve to maximize symbiont transmissibility, whether this is a product of optimizing within-host competitiveness or instead of between host per-capita transmissibility.
Transmission opportunity

Measure of potential for a symbiont to successfully leave its current host for another.
Transmission opportunity increases in the short term as more parasites are shed from a host but increases in the long term as a function of both the rate of parasite shedding and the duration of that shedding, with rate of parasite shedding potentially negatively impacting duration (which, if so, would represent an antagonistic pleiotropy and/or a Parasite's Dilemma). Note that whether transmission is successful will also depend upon new-host availability and that new-host availability can vary as a function of a function of infected-host behavior, which in turn can vary as a function of parasite virulence or, indeed, parasite genotype/phenotype more generally.
Transmission rate

Measure of number of symbionts associated with a current host that successfully leave that host for another as a function of time.
Transmission rate generally is a key controlling variable on both parasite fitness and virulence evolution. A parasite that transmits at a rate such that on average each infected or infested host gives rise to less than one additional new host will have a lower parasite fitness than a transmission rate that results instead in more than one successful transmission to a new host. That transmission rate which gives rise on average to one additional new host results in a parasite steady state within a host population, that is, a more or less unchanging parasite density within an environment. Nevertheless, whether that transmission rate itself is relatively high or instead relatively low, on a per unit time basis, will depend on how long a parasite's opportunities for transmission last. Higher transmission rates generally will be required to attain at least a parasite population density steady state given shorter windows of opportunity for transmission (e.g., as may be seen with acute infectious diseases) while lower transmission rates will be required given longer windows of opportunity (e.g., as may be seen with chronic infectious diseases).

The capacity for a parasite/pathogen to cause disease.
As with disease generally, it is permissible to measure degrees of infectious disease in terms of the extent to which disease results in reductions in the Darwinian fitness of afflicted individuals. Virulence thus can be defined, from a more evolutionary perspective, as the capacity for a parasite/pathogen to lower a host's Darwinian fitness. Virulence nonetheless is usually is measured over short intervals rather than cumulatively and this in more proximate rather than ultimate terms. Thus, a disease that is short and potentially lethal would be considered to involve greater virulence than a disease that is relatively mild but chronic. Virulence also can be viewed as either the potential to cause disease or the degree of disease actually caused. Just as with health, virulence additionally can be viewed entirely from the host's perspective or instead from the virus' perspective, or some combination both. Lastly, virulence is In any case, virulence is a measure of host functionality as a function of parasite actions.
Virulence optimization

  (alt link)
Parasite evolution of parasite impact on host fitness towards maximization of parasite fitness.
Virulence here can be viewed simply as a parasite character, that is, a parasite adaptation, with degrees of virulence representing different traits. Virulence optimization thus is simply evolution towards that trait – that is, some level of virulence associated with a parasite – that provides the parasite with maximal fitness. Depending upon the parasite, host, and circumstances, that level of virulence may be greater, lesser, or equivalent to current levels of fitness. Furthermore, according to the tradeoff hypothesis, that level of virulence that is optimal should vary at least in part as a function of its impact on transmission opportunities and rates.