Forest stand composition and impacts associated with Agrilus auroguttatus Schaeffer (Coleoptera: Buprestidae) and Agrilus coxalis Waterhouse in oak woodlands

Abstract

From 2009–2011, we assessed the impact of the goldspotted oak borer, Agrilus auroguttatus Schaeffer, or its sibling species, Agrilus coxalis Waterhouse, at locations in southern California (denoted infested: ICA and uninfested: UCA), southeastern Arizona (AZ), and southern Mexico (MX). Our surveys examined forest composition of oak woodlands; the degree of injury and proportion of oaks infested with either A. auroguttatus (ICA and AZ) or A. coxalis (MX); and the progression of aerially mapped oak mortality in San Diego Co. (ICA). By most measures of impact that we evaluated, the effect on oaks by the two Agrilus spp. was relatively low in their native regions, but significantly higher by A. auroguttatus at ICA sites. Larger diameter red oak species have been the preferred hosts of A. auroguttatus in AZ and ICA sites, and red oaks greater than approx. 13 cm in DBH throughout California are likely at risk of injury from this invasive pest. At sites in AZ there was no evidence of infestation by A. auroguttatus on living or dead white oak species, whereas at ICA sites we recorded a minor amount of infestation by A. auroguttatus on living individuals of a white oak species (Quercus engelmannii Greene), but no mortality. In contrast, a white oak from MX sites (Quercus peduncularis) was more frequently infested by A. coxalis than were indigenous red oaks. Across all ICA sites, A. auroguttatus has infested 61% of the live larger diameter oaks and killed 13% of the oak component of the forest (vs. 4% infested and 2% dead in AZ, respectively). At survey plots near the predicted origin of the outbreak in CA, over 90% of the larger diameter red oaks have been infested. Nearly 90% of the dead oaks surveyed across all ICA sites showed evidence of previous injury symptoms from A. auroguttatus. Aerial oak mortality polygons associated with A. auroguttatus have expanded ∼50 km in nine years, but our analysis confirms that the outbreak appears to still be confined to San Diego Co. The distance of oak mortality polygons from the predicted origin of the outbreak explained the most variance in a principal component analysis. The invasive population of A. auroguttatus is a significant conservation and ecological threat to the oak woodlands of California and should be managed accordingly, especially by restricting firewood movement.

Introduction

Stem-infesting Agrilus spp. (Coleoptera: Buprestidae) phloem/xylem borers have played key roles in historic cases of oak decline and mortality in the eastern USA and Europe (Nichols, 1968, Stringer et al., 1989, Hartmann and Blank, 1993, Führer, 1998, Oszako, 1998, Thomas et al., 2002). Damage from most Agrilus spp. is associated with oak trees already in decline. For example, in the eastern USA, the native twolined chestnut borer, Agrilus bilineatus Weber, frequently attacks oaks weakened by high levels of defoliation from Lepidoptera, infection by Armillaria sp. (Fr.) Staude root rot, injury from frost, or drought (Chapman, 1915, Hursh and Haasis, 1931, Knull, 1932, Baker, 1941, Staley, 1965, Dunbar and Stephens, 1975, Wargo, 1977) and is typically regarded as a secondary pest on stressed oaks (Dunn et al., 1986, Haack and Acciavatti, 1992, Muzika et al., 2000). In Europe, the native oak splendor beetle, Agrilus bigutattus (F.), regularly interacts with the impacts of insect defoliators, canker fungi, late winter frosts, or drought to cause oak mortality (Jacquiot, 1976, Gibbs and Greig, 1997, Moraal and Hilszczański, 2000, Vansteenkiste et al., 2004, Hilszczański and Sierpinski, 2007).

Oak mortality in northern California, USA over the past 15 years has been well-documented and attributed to Sudden Oak Death, caused by Phytophthora ramorum S. Werres et al. (Rizzo and Garbelotto, 2003). Because Agrilus spp. diversity is low in California (Fisher, 1928, Furniss and Carolin, 1977), there has been no record of native stem-infesting Agrilus spp. interacting with the oak mortality attributed to P. ramorum (Swiecki and Bernhardt, 2006). However, beginning around 2002, a second center with elevated levels of oak mortality emerged in southern California (San Diego Co.). In this area, which is devoid of P. ramorum, an estimated 22,171 trees died over 212,460 ha (USDA Forest Service and Forest Health Monitoring, 2010). The primary cause of this mortality is the invasive goldspotted oak borer, Agrilus auroguttatus Schaeffer, which is aggressively colonizing and killing coast live oak, Quercus agrifolia Née, California black oak, Quercus kelloggii Newb., and canyon live oak, Quercus chrysolepis Liebm. across all land ownerships in San Diego Co. (Coleman and Seybold, 2008). Engelmann oak, Quercus engelmannii Greene, a white oak, occurs in this area as well and is also injured on rare occasions, but not killed, by A. auroguttatus (Coleman and Seybold, 2011). Coleman and Seybold (2011) hypothesized that A. auroguttatus was introduced inadvertently into southern California between the mid-1990’s and 2000 via movement of infested firewood from southeastern Arizona or northern Mexico. Prior to the discovery of A. auroguttatus, the observed high oak mortality in the area was attributed solely to acute drought that occurred during 2002–2003 (Coleman and Seybold, 2008, Coleman et al., 2011); additional abiotic and biotic factors that contribute unequivocally to this southern California oak mortality have not been reported.

A. auroguttatus was described in 1905 from specimens collected in southeastern Arizona (Schaeffer, 1905, Fisher, 1928), whereas a sibling species, Agrilus coxalis Waterhouse, had been described in 1889 from specimens collected in Oaxaca, Mexico (Coleman and Seybold, 2011). Since the collection of the type specimens, additional collecting of A. auroguttatus in the 20th Century established its native distribution as southeastern Arizona and the southern tip of Baja California Sur, Mexico, though suspected populations in western New Mexico and northern Mexico have yet to be confirmed (Coleman and Seybold, 2011). In contrast, the native range of A. coxalis extends from central Mexico to Guatemala (Coleman and Seybold, 2011). DNA sequence analysis of gene fragments from mitochondrial cytochrome oxidase I and nuclear 28S ribosomal DNA suggest that the population of A. auroguttatus in California is genetically more similar to populations in Arizona than to populations (of A. coxalis) collected in southern Mexico (Coleman et al., 2012). Because of the relatively extreme sequence divergence of both the mitochondrial and nuclear gene fragments, this molecular analysis supports the species status of A. auroguttatus and A. coxalis, which previously was based solely on morphological characters (Coleman et al., 2011, Hespenheide et al., 2011).

No life history or impact data were available for either A. auroguttatus or A. coxalis when A. auroguttatus was first collected in southern California in 2004 (Westcott, 2005) and associated with dying oaks there in 2008 (Coleman and Seybold, 2008). We now know that large numbers of the larvae of A. auroguttatus girdle the main stem and larger branches of oaks by feeding primarily on the surface of the xylem, injuring the cambium and phloem. Infested trees are identified by thinning crowns, D-shaped emergence holes, bark damage from woodpecker foraging, and bark staining (Coleman et al., 2011, Hishinuma et al., 2011). Following several years of repeated and extensive injury, oaks eventually succumb to this larval herbivory.

Although initial surveys in southeastern Arizona revealed that Emory oak, Quercus emoryi Torrey, and silverleaf oak, Quercus hypoleucoides A. Camus, are susceptible to injury caused by A. auroguttatus, only low levels of infestation and tree mortality were observed in this native region of the beetle (Coleman and Seybold, 2011). In southern Mexico, Quercus conzatti Trel. and Quercus peduncularis Nee were confirmed recently as hosts for A. coxalis (Coleman et al., 2012). Thus, in California, the invasive population of A. auroguttatus is associated with host oak species for which it had no prior co-evolutionary relationship in Arizona, and the extent of the potential short- and long-term impacts of A. auroguttatus on California’s oak woodlands is unknown.

The objective of this study was to characterize the forest stand conditions in infested and uninfested areas in California by focusing specifically on three factors: the oak composition; the current levels of infestation and tree mortality; and the degree of tree injury associated with A. auroguttatus in the San Diego Co. region where it has been introduced. These data were compared to the same data recorded from plots in the native region of A. auroguttatus in southeastern Arizona, and from plots in the native region of A. coxalis in southern Mexico. Annual oak mortality aerial survey data were also analyzed to assess site characteristics associated with oak mortality polygons and the rate of spread of San Diego Co. oak mortality from this new invasive species.

This information provides short-term impacts (level of tree injury and mortality) and identifies factors associated with higher risk of A. auroguttatus infestation in southern California. These data can help land managers focus on susceptible host species and time their management activities to prevent and slow tree injury and mortality at high-value sites. This work also establishes baseline conditions in long-term plots, which will be revisited in the future by using the same sampling protocols.