POPULATION ECOLOGY Population Density and Transmission of Virus in Experimental Populations of the Western Tent Caterpillar (Lepidoptera: Lasiocampidae) BEATRIX E. BEISNER 1 AND JUDITH H. MYERS Departments of Zoology and Faculty of Agricultural Sciences, Centre for Biodiversity Research, University of British Columbia, 6270 University Boulevard, Vancouver, BC, Canada V6T 1Z4 Environ. Entomol. 28(6): 1107Ð1113 (1999) ABSTRACT Epizootics of nuclearpolyhedrovirus (NPV) are associated with population ßuctua- tions of western tent caterpillars, Malacosoma californicum pluviale (Dyar), in western Canada. To experimentally quantify the spread of NPV in the western tent caterpillar, we measured the transmission of NPV within and among colonies of larvae introduced to alder trees. We altered both host population size (number of larvae per colony on a tree with 4 colonies) and pathogen density (number of infected individuals introduced to 1 of 4 colonies per tree). At the scale of the tree, the transmission coefÞcient (proportion of new infections per initially infected caterpillar) was inde- pendent of host density, but varied with initial pathogen dose. Low doses of virus had higher transmission coefÞcients than high doses. Transmission of virus between colonies within trees was related to the size of the colonies and virus was more widely spread among the 4 colonies when colonies were large (256 larvae per colony). This could be attributed to variation in the development rate and mobility of caterpillars from small and large colonies. The dynamics of NPV in populations of tent caterpillars will be inßuenced by movement of caterpillars among colonies, but increases in the amount of virus above a threshold level will not lead directly to increased numbers of infected individuals. Individual tent caterpillar moths lay all of their eggs in a single egg mass, and in declining populations, egg masses are smaller. The slower development of small colonies arising from these smaller egg masses may delay the recovery of declining populations, restrict the spread of virus, and inßuence the population dynamics of tent caterpillars. KEY WORDS Malacosoma californicum pluviale, caterpillars, nuclearpolyhedrovirus, virus trans- mission, population cycles, experimental epizootiology MANY FOREST CATERPILLAR species in the Northern Hemisphere undergo dramatic ßuctuations in popu- lation density (Myers 1998), and epizootics of nucle- arpolyhedrovirus (NPV) are commonly associated with the outbreak and decline of population density (Myers 1988a, b, Myers 1990). One of the most im- portant processes in the dynamics of viral infection in caterpillar populations is the transmission of disease among individuals. NPV is transmitted by a susceptible caterpillar ingesting occlusion bodies, which have been released on the death of an infected individual. Traditional models of host-pathogen interactions, in- cluding those of NPV epizootics, assume that viral transmission increases linearly with densities of either susceptible hosts or pathogen (mass action assump- tion) (Kermack and McKendrick 1927; Bailey 1975; Anderson and May 1980, 1981). Insect-virus systems are ideal for investigating host- pathogen interactions, and to date several experimen- tal studies of transmission that have been done using laboratory populations (Knell et al. 1998) or caterpil- lars enclosed in bags or on potted plants (Dwyer 1991, DÕAmico et al. 1996). Experimental and observational evidence of interactions between forest Lepidoptera, the Douglas-Þr tussock moth, Orgyia pseudotsugata (McDunnough), and the gypsy moth, Lymantria dis- par (L.), and their respective polyhedroviruses, dem- onstrates that horizontal transmission is related to host density in a nonlinear fashion (Dwyer 1991, Dwyer and Elkinton 1993, DÕAmico et al. 1996). Transmission is more efÞcient at low host density. Theoretical stud- ies have tried to understand how a nonlinear relation- ship between the transmission parameter and host or pathogen density alters predictions for epizootic de- velopment and the stability of host-pathogen interac- tions (Anderson 1979; Post et al. 1983; Liu et al. 1986, 1987; Hochberg 1991b; Briggs and Godfray 1995). These studies show that the functional formulation of this relationship has signiÞcant quantitative and qual- itative effects on the predicted host-pathogen dynam- ics. Several mechanisms could cause nonlinear trans- mission rates. Limited mobility of hosts or a clumped distribution of viral occlusion bodies could restrict the encounter rates of hosts and pathogens, making the system less mixed than expected from mass action. If movement of forest caterpillars changes with density, the form of the nonlinear transmission could also vary. 1 beisner@zoology.ubc.ca, myers@zoology.ubc.ca 0046-225X/99/1107Ð1113$02.00/0  1999 Entomological Society of America