Modelling dwarf mistletoe at three scales: life history, ballistics and contagion

The epidemiology of dwarf mistletoe (Arceuthobium) is simulated for the reproduction, dispersal, and spatial patterns of these plant pathogens on conifer trees. A conceptual model for mistletoe spread and intensification is coded as sets of related subprograms that link to either of two individual-tree growth models (FVS and TASS) used by managers to develop silvicultural and land management plans. This dwarf mistletoe model is based on knowledge of mistletoe biology and forest practices acquired through a series of workshops, programming exercises, and continuing research and development. Key components of mistletoe epidemiology are identified as life history, ballistics, and contagion. An infestation is quantified at the tree-level by a standard measure of mistletoe intensity, the dwarf mistletoe rating (DMR). Life history describes the progression of mistletoe populations from new infections to seed-producing plants and includes biocontrol and mortality of the mistletoe. The model tracks mistletoe populations as changes in DMR rather than individual plants. Life history is represented as changes in pools for various developmental stages; and rates of change are modified by time, light, and other environmental factors (including hyperparasites). Dwarf mistletoes disperse by explosive discharge of small seeds followed by ballistic flight that displaces seeds horizontally to a maximum distance of about 14m. The model represents dispersal as probabilistic, spatially explicit, ballistic trajectories for each host tree in a simulated stand. The spacing of trees and mistletoe within infested stands exhibits a range of patterns as regular, random, or clumped; the rates of spread to new hosts and intensification within infested hosts are influenced by crown and canopy distributions derived from descriptors of stem clumping and mistletoe contagion. Spatial arrangements of trees in the model are determined from stand-level statistics that characterize groups of trees at the scale of a 14 m radius neighborhood, the maximum distance for ballistic dispersal. The number of trees in a simulated neighborhood is a function of the variance to mean ratio for tree density in the stand. The autocorrelation of trees of more similar DMR is used to simulate aggregation of infected trees into infestation patches. Model behavior for sensitivity to key relationships and fit to observed stands is demonstrated using data for a dense western hemlock stand and two initially similar, open-canopy ponderosa pine stands either treated for mistletoe or left untreated. The model provides a practical tool for assessing the long-term, cumulative effects of disease and management in mistletoe-infested stands. (c) 2006 Elsevier B.V. All rights reserved.