Measurement and simulation of dwarf mistletoe infection of second-growth western hemlock on southern Vancouver Island

Abstract

Infection of western hemlock (Tsugaheterophylla (Raf.) Sarg.) residual overstory and second-growth understory by hemlock dwarf mistletoe (Arceuthobiumtsugense (Rosendahl) G.N. Jones) was analyzed in seven plots on southern Vancouver Island. The number of infections in residual trees ranged from 373 to 4058 and in second-growth trees from 3 to 455. The number in second-growth trees was significantly correlated positively with dbh and tree height and inversely with height/dbh ratio. The percentage of crown length infected varied significantly among plots and was significantly correlated with the number of infections. Vertical distribution of infections in tree crowns did not conform to test distributions; distribution by infection age approximated the Poisson. The range in infection age varied according to height in green crown and length of green crown infected. Mortality of infections varied significantly among plots and was greatest in lower slope sites and least in a drier upper slope site. Proportion of dead infections was a function of height in crown and total length of green crown infected. Data suggest that infection level in second-growth trees was proportional to the number of residuals and was inversely related to percent nonhost species, stand density, and growth rate. The number of infections predicted by a simulation model averaged 107% of the number recorded in residual trees and 128% in second-growth trees. Curves of predicted vertical distribution of infections in crowns were of the same shape as those recorded in plot sample trees. The predicted average age-class distribution of infections approximated Poisson distribution within the same probability range as recorded in plot sample trees. The predicted average dbh of residual and second-growth trees were 90 and 97%, respectively, of the recorded values; predicted average heights were 115 and 101%, respectively, of recorded values. Infection predicted by the model over a range of hypothetical stand, site, and infection conditions was used to elucidate major epidemiological factors that might bear on silvieultural control. Results confirmed the observed effects of number of residuals, stand composition, growth rate, and stand density on infections.