Radial Growth Losses in Preferred and Avoided Tree Species During Gypsy Moth Outbreaks Robin Naidooand Martin J. Lechowicz, Department of Biology, McGill University, 1205 Dr. Penfield Ave.,Montrdal,Qudbec, Canada H3A lB1. ABSTRACT: We investigated the effect of gypsy moth larval density on radial growth of preferred and avoided trees: northern red oakand white ash, respectively. Individual trees were censused forgypsy moth larvae from 1979 to 1992 at a site where several outbreaks occurred. Annual growth rings were measured from 1950 to 1992 on increment cores taken from these same trees, as well as from trees at a nearby site that had not experienced any outbreaks. Regression models of growth at the outbreak site on growth at the nonoutbreak site were developed to isolate the influence of gypsy moth defoliation frorh other facto rs such as climate. These were then used to generate expected values for radial growth inthe absence of gypsy moth atthe outbreak site. During the first year of the first gypsy moth outbreak, there was a mean reduction in radial growth of46% in red oak, a loss similar towhat has been reported inother studies. Growth ofwhite ash was much higher than predicted in2 yr during and subsequent to the first outbreak. Yearly larval counts from 1979 to 1992 onred oak at the outbreak site were negatively correlated with oak radial growth after correcting for climate, suggesting that nonoutbreak levels ofgypsy moth may reduce radial growth more than has previously been thought. Larval counts on ash were unco rrelated with ash growth after correcting for climate. North. J. Appl. For. 16(1):11-18. The gypsy moth (Lymantria dispar L.) is an introduced forest insect that has become the most serious defoliator of hardwood trees in northeastern NorthAmerica (Montgomery andWallher 1988).As a "sustained eruption" insect (sensu Berryman 1987), outbreaks can cover large areasand can persist in a particular area for several years before thepopu- lation returns toendemic levels. During these prolonged and widespread outbreaks, defoliation by larvae can have severe socioeconomic (e.g., Liebholdet al. 1995) and ecological (e.g., Campbell and Sloan 1977) consequences. Despite this, theeffect of gypsy moth defoliation onradialgrowth of host trees in northeastern North America has notbeen adequately investigated. Several studies have compared theaverage radialgrowth over a certain timeperiod before gypsy moth defoliation to average growth after defoliation (Minott and Guild 1925, Baker 1941,Kulman1971,Brown etal. 1979, Wargo 1981, NOTE: Robin Naidoo is the corresponding authorand his currentad- dressis Sustainable ForestManagement Network, G-208 Biological Sciences Bldg., University of Alberta, Edmonton AB T6G 2E9--Phone: (403)492-7048; Fax:(403)492-6547; E-mail:robin.naidoo @ ualberta.ca. The authors thank Ben6it C6t6, Danielle Charton, and Yves Bergeron for the use of their microscope. Charles Mercier provided stellar assis- tance in the field. We thank Kurt Gottschalk for his comments on an earlierversion of the manuscript. Threeanonymous reviewers provided insightful comments on an earlier version of the manuscript. The research was supported by NSERC and FCAR grants to MJL. Finally, we are grateful to McGill University for allowing us to conduct this research at the Mont St-Hilaire Biosphere Reserve. Campbelland Garlo 1982). Reduction of growth in the postdefoliatio nperiod compared to the predefoliation period wasattributed to gypsy mothdefoliation. The weakness in this method is that potentially confounding differences in climatic conditions between periods are not accounted for, makingthe effectsof climateand defoliation on growth difficult to separate. More recent studies using dendroecological techniques (Fritts.andSwetnam1989) have often usedradial growth chronologies of nonhost species, orof nondefoliated trees of the same species, toestimate losses due toinsect defoliation. Anylarge depressions inthe growth chronology ofdefoliated trees relative to undefoliated trees can be attributed to defo- liation, since responses to macroclimatic conditions should be similar if the trees are growingin the samegeneral environment (e.g., Kulman 1971, Swetnam and Lynch1989, Mason et al. 1997). This approach has not been usedto quantify theeffect of gypsy moth defoliation on growth of hardwood trees in northeastern North America. This paper will discuss the use of chronologies of nondefoliated trees in determining the effect of defoliation by a known density of gypsy moth larvae on radial growth. Our method involved theuse of twonearby stands withdifferent histories of outbreaks and consequent defoliation. Onestand was severely defoliated by gypsymoth on two separate occasions during a 13 yr period. The second stand did not suffer heavy defoliation during the initial outbreak atthe site, although itssubsequent defoliation history isunclear. Radial NJAF 16(1) 1999 11 Downloaded from https://academic.oup.com/njaf/article/16/1/11/4788482 by guest on 07 November 2021