ABSTRACT 

      Why do some parasites kill the host they depend upon while 
others coexist with their host? Two prime factors determine parasitic 
virulence: the manner in which the parasite is transmitted, and the 
evolutionary history of the parasite and its host. Parasites which 
have colonized a new host species tend to be more virulent than 
parasites which have coevolved with their hosts. Parasites which are 
transmitted horizontally tend to be more virulent than those 
transmitted vertically. It has been assumed that parasite-host 
interactions inevitably evolve toward lower virulence. This is 
contradicted by studies in which virulence is conserved or increases 
over time. A model which encompasses the variability of parasite-host 
interactions by synthesizing spatial (transmission) and temporal 
(evolutionary) factors is examined. Lenski and May (1994) and Antia et 
al. (1993) predict the modulation of virulence in parasite-host 
systems by integrating evolutionary and transmissibility factors.

INTRODUCTION 

      Why do certain parasites exhibit high levels of virulence within 
their host populations while others exhibit low virulence? The two 
prime factors most frequently cited (Esch and Fernandez 1993, Toft et 
al. 1991) are evolutionary history and mode of transmission. 
Incongruently evolved parasite-host associations are characterized by 
high virulence, while congruent evolution may result in reduced 
virulence (Toft et al. 1991). Parasites transmitted vertically (from 
parent to offspring) tend to be less virulent than parasites 
transmitted horizontally (between unrelated individuals of the same or 
different species). Studies in which virulence is shown to increase 
during parasite-host interaction, as in Ebert's (1994) experiment with 
Daphnia magna, necessitate a synthesis of traditionally discrete 
factors to predict a coevolutionary outcome. Authors prone to 
habitually use the word decrease before the word virulence are 
encouraged to replace the former with modulate, which emphasizes the 
need for an inclusive, predictive paradigm for parasite-host 
interaction.

      Evolutionary history and mode of transmission will first be 
considered separately, then integrated using an equation discussed
by Antia et al. (1993) and a model proposed by Lenski and May (1994). 
Transmission is a spatial factor, defined by host density and specific 
qualities of host-parasite interaction, which gives direction to the 
modulation of virulence. Evolution is a temporal factor which 
determines the extent of the modulation. The selective pressures of 
the transmission mode act on parasite populations over evolutionary 
time, favoring an equilibrium level of virulence (Lenski and May 
1994).

DOES COEVOLUTION DETERMINE VIRULENCE? 

      Incongruent evolution is the colonization of a new host species 
by a parasite. It is widely reported that such colonizations, when
successful, feature high virulence due to the lack of both evolved 
host defenses and parasitic self-regulation (Esch and Fernandez 1993, 
Toft et al. 1991). Unsuccessful colonizations must frequently occur 
when parasites encounter hosts with adequate defenses. In Africa, 
indigenous ruminants experience low virulence from Trypanosoma brucei 
infection, while introduced ruminants suffer fatal infections (Esch 
and Fernandez 1993). There has been no time for the new host to 
develop immunity, or for the parasite to self-regulate. Virulent 
colonizations may occur regularly in epizootic-enzootic cycles. Sin
Nombre virus, a hemmorhagic fever virus, was epizootic in 1993 after 
the population of its primary enzootic host, Peromyscus maniculatus, 
had exploded, increasing the likelihood of transmission to humans 
(Childs et al. 1995). Sin Nombre exhibited unusually high mortality in 
human populations (Childs et al. 1995), which were being colonized by 
the parasite.

      It is assumed that coevolution of parasite and host will result 
in decreased virulence (Esch and Fernandez 1993, Toft et al. 1991). 
Sin Nombre virus was found to infect 30.4 % of the P. maniculatus 
population, exhibiting little or no virulence in the mice (Childs et 
al. 1995). Similar low levels of virulence have been found in the 
enzootic rodent hosts of Yersinia pestis (Gage et al. 1995). In 
Australia, decreased grades of virulence of myxoma virus have been 
observed in rabbit populations since the virus was introduced in 1951 
(Krebs C. J. 1994). Many of the most widespread parasites exhibit low 
virulence, suggesting that success in parasite suprapopulation range 
and abundance may be the result of reduction in virulence over time.
Hookworms are present in the small intestines of one-fifth of the 
world's human population and rarely induce mortality directly
(Hotez 1995). 

      Evolution toward a higher level of virulence has been regarded 
as an unexplainable anomaly. Parasites which do less harm presumably 
have an advantage throughout a long coevolutionary association with 
their hosts. Ebert's (1994) experiment with the planktonic crustacean 
Daphnia magna and its horizontally transmitted parasite Pleistophora 
intestinalis suggests that coevolution does not determine the 
direction of the modulation of virulence. Virulence decreased with the 
geographic distance between sites of origin where the host and 
parasite were collected (Ebert 1994). Thus, the parasite was 
significantly more virulent in hosts it coexisted with in the wild 
than it was in novel hosts. Many viruses, such as Rabies (Lyssavirus 
spp.), persist in natural populations while maintaining high levels of 
virulence in all potential hosts (Krebs, J. W. 1995). Extinction is 
not an inevitable outcome of increased virulence (Lenski and May 
1994). Increased or conserved virulence during coevolution calls
into question long held assumptions about the effect of coevolution on 
parasitic virulence (Gibbons 1994). Parasitic virulence frequently 
changes over coevolutionary time, but the length of parasite-host 
association does not account for the virulence of the parasite. 
Transmission has been identified as the factor which determines the 
level of parasitic virulence (Read and Harvey 1993). 

TRANSMISSION AND THE DIRECTION OF MODULATION 

      Herre's (1993) experiment with fig wasps (Pegoscapus spp.) and 
nematodes (Parasitodiplogaster spp.) illustrates the effect of 
transmission mode on parasitic virulence. When a single female wasp 
inhabited a fig, all transmission of the parasite was vertical, from 
the female to her offspring. The parasite's fitness was intimately 
tied to the fecundity of the host upon which it had arrived. When a 
fig was inhabited by several foundress wasps, horizontal transmission 
between wasp families was possible. In the figs inhabited by a single 
foundress wasp, Herre found that less virulent species of the nematode 
were successful, while in figs containing multiple foundress wasps, 
more virulent species of the nematode were successful. Greater 
opportunity to find alternate hosts resulted in less penalty for 
lowering host fecundity. More virulent nematodes had an adaptive 
advantage when host density was high and horizontal transmission was 
possible. When host density was low, nematodes which had less effect
on host fecundity ensured that offspring (i.e. future hosts) would be 
available.

      Low virulence is characteristic of many vertical transmission 
cycles. Certain parasites avoid impairing their host's fecundity by
becoming dormant within maternal tissue. Toxocara canis larvae reside 
in muscles and other somatic tissues of bitches until the 42nd to 56th 
day of a 70-day gestation, when they migrate through the placenta, 
entering fetal lungs where they remain until birth (Cheney and Hibler 
1990). A high proportion of puppies are born with roundworm infection, 
which can also be transmitted from bitch to puppy by milk (Cheney and 
Hibler 1990). If host density is low, a highly evolved vertical 
transmission cycle (which exhibits low virulence in the parent) 
ensures the survival of the parasite population.

      High virulence is characteristic of horizontal transmission 
cycles. In Herre's (1993) experiment, more virulent parasites were
favored when host density was high and reduction of host fitness was 
permissible. Certain parasites benefit from reduced host fitness, 
particularly parasites borne by insect vectors (Esch and Fernandez 
1993) and parasites whose intermediate host must be ingested by 
another organism to complete the parasitic life cycle. By immobilizing 
their host, heartworm (Dirofilaria immitis) and malaria (Plasmodium 
spp.) increase the likelihood that mosquitoes will successfully ingest 
microfilaria or gametocytes along with a blood meal. Heartworm 
infestation causes pulmonary hypertension in dogs (Wise 1990), 
resulting in lethargy and eventual collapse (Georgi and Georgi 1990). 
Host immobility increases the opportunities for female mosquitoes to
find and feed upon hosts (Read and Harvey 1993). Infected dogs have 
large numbers of D. immitis microfilaria in their circulatory systems, 
again increasing the likelihood of ingestion by the insect. Many 
infected dogs eventually die from heartworm, but in the process the 
parasite has ensured transmission. Similar debilitating effects have 
been observed in tapeworm-stickleback interaction; infected 
sticklebacks must swim nearer the water's surface due to an increased 
rate of oxygen consumption caused by the parasite (Keymer and Read 
1991). Parasitized sticklebacks are more likely to be seen and eaten 
by birds, the next host in the life cycle.

      Many horizontally transmitted parasites manipulate specific 
aspects of host behavior to facilitate transmission between species.
Host fitness is severely impaired in such interactions. The digenean 
D. spathaceum invades the eyes of sticklebacks, increasing the 
likelihood of successful predation by birds (Milinski 1990). D. 
dendriticum migrate to the brains of infected ants, causing them to 
uncontrollably clamp their jaws onto blades of grass, ensuring 
ingestion by sheep (Esch and Fernandez 1993, Combes 1991). Infection 
of a mammalian brain by rabies (Lyssavirus spp.) alters the host's 
behavior, increasing the chance of conflict with other potential 
hosts, while accumulation of rabies virus in the salivary glands 
ensures that it is spread by bites (Krebs, J. W. et al. 1995). 
Horizontally transmitted parasites which target nervous tissue 
increase transmissibility by modifying the host into a suicidal 
instrument of transmission.

      Transmission factors determining parasitic virulence are the 
spatial element in a spatial-temporal dynamic. Host density directly
determines the virulence of parasites which depend upon a single host 
species (Herre 1993). Virulence may be increased when transmission 
necessitates insect vectors or consumption of the primary host by 
another species. Virulence varies inversely with the distance between 
potential hosts; this distance is magnified when it is measured 
between different species.

THE EQUILIBRIUM MODEL 

      It has been proposed that there is a coevolutionary arms race 
between parasite and host, as the former seeks to circumvent the
defensive adaptations of the latter (Esch and Fernandez 1993). In this 
view, parasitic virulence is the result of a dynamic stalemate between 
host and parasite. This exemplifies the red queen hypothesis, which 
predicts continued stalemate until the eventual extinction of both 
species. Benton (1990) notes that the red queen hypothesis ignores the 
potential for compromise in such a system. Snails (Biomphalaria 
glabrata) resistant to Schistosoma mansoni are at a selective 
disadvantage due to the costs associated with resistance (Esch and 
Fernandez 1993). A high level of virulence persists in the system 
because the snail cannot afford to mount an adequate defense. The arms 
race hypothesis assumes that the host population can successfully
counter increasing parasitic virulence with resistance over an 
extended period of time. Although an arms race may be sustainable in 
some fraction of parasite-host interactions, many hosts (such as B. 
Glabrata) cannot participate indeterminately.

      An alternative explanation for the reduced virulence of 
congruently evolved hosts and parasites is the prudent parasite
hypothesis (Esch and Fernandez 1993), in which parasitic virulence 
decreases in response to host mortality. Parasites which are too 
virulent drive their hosts, and themselves, to extinction. Parasites 
which are less virulent persist in the host population. The prudent 
parasite hypothesis helps to account for the variation in 
coevolutionary outcome by linking host population dynamics with 
virulence, but it fails to describe the individual selective forces 
which modulate virulence over time. The prudent parasite hypothesis 
serves as the theoretical framework in which the factors determining 
parasitic virulence can be synthesized. Antia et al. (1993) and Lenski 
and May (1994) propose a tradeoff between transmissibility and induced 
host mortality which predicts that parasites will evolve toward a 
level of virulence which strikes an equilibrium in the parasite-host 
system. Equilibrium models suggest that P. intestinalis, which evolved 
a higher (yet appropriate) level of virulence in its host (Ebert
1994), is a prudent parasite. Antia et al. (1993) use an equation 
developed by May and Anderson in 1983 to examine the tradeoffs in 
parasite-host interaction: Ro = (BN) / (a + b + v). Ro is the net 
reproductive rate of a parasite, B is the rate parameter for 
transmission, N is host density, a is the rate of parasite induced 
host mortality, b is the rate of parasite-independent host mortality 
and v is the rate of recovery of infected hosts. Parasite populations 
grow when transmission or host density increase, when host mortality
decreases or when hosts recover slowly. Studies have established a 
positive correlation between transmissibility (B) and host mortality 
(a) (Ebert 1994, Antia et al. 1993, Lenski and May 1994). Parasite 
populations which exhibit high transmissibility (i.e. virulence) 
within a host population are simultaneously lowering host density. 
When host density is low, parasites which exhibit high virulence may 
kill their hosts before contact with new hosts occurs. Thus, 
transmissibility is a spatial factor which describes the likelihood of 
contact between hosts and, ultimately, between a parasite and its 
host.

      Lenski and May (1994) propose an evolutionary sequence in which 
parasite populations adapt to the changes they cause in host density 
(Fig. 1). A parasite suprapopulation is likely to include a range of 
genotypes which are expressed in different potential levels of 
virulence (Lenski and May 1994). When host density is high, more 
virulent parasites are successful and host density is reduced. At a 
lower density of hosts, less virulent strains of the parasite are at a 
selective advantage as they increase host survival during infection 
and allow more time for transmission to occur. Also, more virulent 
strains of the parasite are prone to induce mortality in entire 
subsets of the host population, driving themselves to extinction along 
with their hosts. This pattern repeats over time, lowering virulence 
with each adjustment to declining host population size. Extinction of 
the host population is avoided when sufficient variation is present in 
the parasite population (Lenski and May 1994).

      The evolutionary sequence may be reversed to explain evolution 
toward higher virulence when parasitic virulence is below the
equilibrium level. More virulent strains of the parasite outcompete 
less virulent strains when host density is above equilibrium.
Conservation of virulence over time occurs when a stable equilibrium 
is maintained. Conserved virulence may be high (Lenski and May 1994), 
but it reflects stability within a system dictated by a unique set of 
transmission factors. Many parasites must reach a certain population 
size within the host to be successfully transmitted, while in certain 
systems, sacrifice of one host facilitates transmission to the next 
host (i.e. interspecies transmission). The inclusiveness of the 
equilibrium model gives it great potential for accurate predictability 
of a broad range of parasite-host interactions.

CONCLUSION 

      Traditional assumptions about the factors determining parasitic 
strategy have been largely apocryphal, ignoring contradictory evidence 
(Esch and Fernandez 1993). Equilibrium models synthesize the temporal 
(i.e. evolutionary) factors and spatial (i.e. transmission) factors 
characteristic of parasite-host systems. Time is required to modulate 
virulence, while spatial factors such as host density and transmission 
strategy determine the direction of the modulation.

      The development of an inclusive, accurate model has significance 
beyond theoretical biology, given the threat to human populations 
posed by pathogens such as HIV (Gibbons 1994). Mass extinctions such 
as the Cretaceous event may have resulted from parasite-host 
interaction (Bakker 1986), and sexual reproduction (i.e. recombination 
of genes during meiosis) may have evolved to increase resistance to 
parasites (Holmes 1993). Parasitism constitutes an immense, if not 
universal, influence on the evolution of life, with far-reaching 
paleological and phylogenetic implications. A model which synthesizes 
the key factors determining parasitic virulence and can predict the 
entire range of evolutionary outcomes is crucial to our understanding 
of the history and future of species interaction.

---

Literature Cited

Antia, R., Levin, B. R., and May, R. M. 1993. Within-Host Population 
Dynamics and the Evolution and Maintenance of Microparasite Virulence. 
The American Naturalist. 144:457-472. 

Bakker, R. T. 1986. The Twilight of the Dinosaurs. pp. 442-444. In The 
Dinosaur Heresies. Zebra Books. New York. 

Benton, M. J. 1990. Red Queen Hypothesis pp. 119-123 In Paleobiology: 
A Synthesis. Briggs, D. E. G. and Crowther

P. R. Eds. Blackwell Scientific Publications. Oxford. 

Cheney J. M. and Hibler, C. P. 1990. Parasitology and Public Health 
pp. 83-112 In Clinical Textbook for Veterinary Technicians, Second 
Edition. McCurnin, D. M. Ed. W. B. Saunders Company. Philadelphia. 

Childs, J. E., Mills, J. N., and Glass, G. E. 1995. Rodent-borne 
Hemmoragic Fever Viruses: A Special Risk For Mammalogists? Journal of 
Mammalogy. 76:664-680. 

Combes, C. 1991. Ethological Aspects of Parasite Transmission. The 
American Naturalist. 138:866-877. 

Ebert, D. 1994. Virulence and Local Adaptation of a Horizontally 
Transmitted Parasite. Science. 265:1084-1086. 

Esch, G. W. and Fernandez, J. C. 1993. Evolutionary Aspects. pp. 
231-267 In A Functional Biology of Parasitism:

Ecological and Evolutionary Implications. Chapman and Hall. London. 

Gage, K. L., Ostfeld, R. S., and Olson, J. G. 1995. Nonviral 
Vector-borne Zoonoses Associated with Mammals in the United States. 
Journal of Mammalogy. 76:695-715. 

Gibbons, A. 1994. Mistreating a Long-Time Host. Science. 265:1037. 

Georgi, J. R. and Georgi, M. E. 1990. Helminths. pp. 103-224 In 
Parasitology for Veterinarians. W.B. Saunders Company. Philadelphia. 

Herre, E. A. 1993. Population Structure and the Evolution of Virulence 
in Nematiode Parasite of Fig Wasps. Science. 259:1442-1444. 

Holmes, B. 1993. Evolution's neglected superstars. New Scientist. 
140:30-34. 

Hotez, P. J. and Pritchard, D. I. 1995. Hookworm Infection. Scientific 
American. 272 

Keymer, A. E. and Read, A. F. 1991. Behavioral Ecology: The Impact of 
Parasitism pp. 37-61 In Parasite-Host Associations: Coexistence or 
Conflict? Toft, C. A. et al. Eds. Oxford University Press. Oxford. 

Krebs, C. J. 1994. Natural Regulation of Population Size pp. 322-348 
In Ecology: The Experimental Analysis of Distribution and Abundance. 
Harper Collins. New York. 

Krebs, J. W., Wilson, M. L., and Childs, J. E. 1995. Rabies - 
Epidemiology, Prevention, and Future Research. Journal of Mammalogy. 
76:681-694. 

Lenski, R. E. and May R M. 1994. The Evolution of Virulence in 
Parasites and Pathogens: Reconciliation Between Two Competing 
Hypotheses. Journal of Theoretical Biology. 169:253-265. 

Milinski, M. 1990. Parasites and Host Decision-Making pp. 95-116 In 
Parasitism and Host Behaviour. Barnard, C. J. and Behnke, J. M. Eds. 
Taylor & Francis. London. 

Read, A. F. and Harvey, P. H. 1993. The evolution of virulence. 
Nature. 362:500-501. 

Toft, C. A. and Aeschlimann, A. 1991. Introduction: Coexistence or 
Conflict? pp. 1-12 In Parasite-Host Associations: Coexistence or 
Conflict? Oxford University Press. Oxford. 

Wise, L. A. (1990) Small Animal Medical Nursing pp. 226-261 In 
Clinical Textbook for Veterinary Technicians, Second Edition. 
McCurnin, D. M. Ed. W. B. Saunders Company. Philadelphia.