Research Trench Inserts as Long-term Barriers to Root Transmission for Control of Oak Wilt

Wilson, A. D., and Lester, D. Ci. 2002. Trench inserts as long-term barriers to root transmission for control of oak wilt. Plant Dis. 86: 1067-1074. Physical and chemical barriers to root penetration and root grafting across trenches were evaluated for their effectiveness in improving trenches as barriers to root transmission of the oak wilt fungus in live oaks. Four trench insert materials were tested, including water-permeable Typar and Biobarrier, and water-impermeable Geomembranc of two thicknesses. Systemic fungicide treatments of trees immediately outside of trenches also were tested. In the first several years following trench installation, an abundance of small adventitious roots commonly formed from roots sevcrcd by trenching. These roots provided opportunities for initiation of root grafts across trenches in subsequent years. Although trench inserts did not significantly improve trenches during the first 3 years following trench installation, water-permeable inserts did effectively improve the performance of trenches beyond the third posttrenching year, when trenches are normally effective, and extended trench longevity indefinitely. The water-permeable inserts were more effective root barriers because they did not direct root growth from the point of root contact. The water-impermeable materials, however, did tend to direct root growth around these barriers, leading to the development of new root graft connections and associated oak wilt root transmission across the trench. The additional cost of trench inserts above trenching costs was justified in urban and rural homestead sites, where high-value landscape trees required more protection and additional retrenching costs were avoided. Additional keywords: Cer-trtoc~stis egucrcrrunz, cultural control, propiconazole, Quercus ,fic,ri,fbnnis, Quer-cus ~i~~inirrrra, tritlurahn herbicide Oak wilt, caused by Cmtrocystis jir,~trcetrr~~r~~ (T.W. Bretz) J. Hunt, is a major vascular wilt disease that continues to shape the ecology of hardwood forest ecosystems of the eastern United States. Since oak wilt was first discovered in Wisconsin in 1942 (I), it has been considered by many to be the most serious disease of oak (Q~KXXS sp.) in North America (2,16,20). The oak wilt fungus is potentially the most destructive of all forest pathogens because few phytopathogenic microbes have greater capacity to kill their tree hosts with such rapidity (16,23,3X). The impact of the disease on oak forests in the United States Publication no. D-2002-0809-01 R This article is in the public domain and not copyrightable. It may be freely reprinted with customary crediting of the source. The American Phytopathological Society, 2002. has been exacerbated by changes in forest stand composition and forest management practices that have resulted in oak stands with greater proportions of susceptible red oak species ( 16,29). Oak wilt probably was first observed in Texas in the 1930s within the Hill Country or Edwards Plateau region. Unusually high live oak mortality was reported during this period in the Austin area (24,25). The semievergreen live oaks, Qrrercus $& ,fortni.s Small (plateau live oak) and Quercus vir;i’itGnrm Miller (coastal live oak), are considered the most valuable woodland and urban tree species in central Texas (2 1). It appears that populations of the oak wilt fungus gradually developed within susceptible live oak stands over the next 40 years until they reached critical mass in the 1970s (29). The oak wilt epidemic that ensued continues to cause increasingly devastating losses to oak resources in Texas. Cumulative economic losses hitherto have been very conservatively estimated to be in the hundreds of millions of dollars statewide (2930). Such losses arc easily rationalized given that a single large live oak can be worth up to $20,000 to residential property values in metropolitan areas ( 13). Confirmed diagnoses of the disease have been reported in oaks from at least 61 of 254 Texas counties. The practice of mechanically cutting root connections to control root transmission of the oak wilt fungus has been recommended for many years (18). Trenching to sever root connections between healthy trees in advance of the visible expanding edge of infection centers has long been the cornerstone of oak witt suppression efforts both in Texas and in midwestern states with active control programs (29). Since 1988, the Texas Forest Service has administered the Texas Oak Wilt Suppression Project (TOWSP), which has installed over 650,000 m of trench to combat this disease (4). Trenching has been a particularly important tool for dealing with the disease in highly valued live oak stands because root grafts result in exlensively interconnected root systems. This tendency is further compounded by the growth habit of live oaks in forming root sprouts from mother trees that often give rise to large clusters of clonal trees or “motts” with common root systems (12,22,27). These natural growth tendencies increase the predisposition of live oaks to root transmission and have often resulted in dramatic mortality over very large areas. Red oaks such as Texas red oak or Spanish oak (Q. ttr~~lnu Buckley = Q. buckleyi Dorr & Nixon), blackjack oak (Q. ttmrilrttzrlica Milnchh.), and Shumard oak (Q. .shumurdii Buckley) are the most susceptible species to oak wilt in Texas, but disease incidence is greater in live oaks due to their growth-form predispositions and shallow, extensive root systems. The higher incidence of the disease in live oaks also may be attributed to their abundance in both urban and rural forests of central Texas. The predominance of live oaks has resulted from landscapemanagement practices involving fire suppression, preferential thinning, overgrazing, and selective protection of live oaks in this region (2,7,12). A 7-year study was initiated in 1993 to evaluate the efficacy of adding physical and/or chemical barriers to trenches for long-term control of root transmission of C. ,fir~ctcrrrrrrm in live oaks. The primary objectives of this study were to (i) test the usefulness of trench insert materials in preventing root penetrations and the development of new root graft connections Plant Disease / October 2002 1067 across trenches, (ii) evaluate fungicide (.propiconazole) treatments of trees immediately outside of trenches in preventing root transmission of the oak wilt fungus, and (iii) assess the capacity of all these barriers to provide long-term control of oak wilt root transmission. Since the time required for maturation of this study was unknown, interim results providing periodic assessments of barrier performance were reported previously (32-36). MATERIALS AND METHODS Research site and plot installation. Research was conducted on the l,lOO-ha Circle C Ranch Land Development Tract located at the southern limits of Austin, TX, in Travis County. Plots were established in a mature natural stand of live oaks growing near a residential development site with a predominantly rocky, sandy clay-loam soil. Soil depth to bedrock ranged from 1.0 to I .7 m at the test site. Test trees were selected approximately 25 to 30 m beyond the expanding edge of a large oak wilt infection center, previously determined to be a minimum buffer zone (15,29). A roughly linear trench, established 27 July 1993, was cut approximately 1.6 km long and 1.5 m deep with a Vermeer turbo II trencher immediately adjacent to test trees and between test trees and the infection center. Selection of trench depth was based on soil depth, insert availability, and root penetration of soil (27,37). The experimental design consisted of 18 sequential plots approximately 46 to 157 m long containing 12 to 18 test trees each situated along the full length of the trench. Trees within research plots were mapped for spatial calculations using a Criterion 400 survey laser (Laser Technology, Inc., Englewood, CO) and sequential-target mapping algorithms with polar Y plotting methods (28). Seven barrier treatments were applied to separate plots on 13 December 1993 in a completely randomized linear order along the trench with three replicate plots per treatment. The treatments included trenches with one of four trench inserts, no insert (trench only), trench with fungicide treatments of test trees outside of the trench, and no trench as untreated controls (Fig. 1). The three trench-only plots were established as semicircular bubbles around no-trench segments to maintain continuity to the trench barrier. Four trench-insert materials were tested, including water-permeable Typar polypropylene spunbonded fabric at 4 oz. (1 x) weight; Biobarrier or Typar with tritluralin-impregnated I O-mm-diameter, controlled-release hemispherical pellets (54% polyethylene, 18% carbon black, and 28% trifluralin by weight) bonded to polypropylene fabric with uniform 3%cm spacing or 688 pellets per square meter (Reemay Inc., Old Hickory, TN); and water-impermeable polyethylene Rufco Geomembrane liners (Raven Industries, Springfield, OH) of two thicknesses (20 and 30 mil). Trench inserts were placed into trenches in 15.2or 30.5-m lengths, mounted with 15-cm steel or aluminum pins to the wall of the trench on the side closest to the infection center, and additionally supported by backfilling the trench with soil removed during construction of the trench, followed by leveling with a backhoe scoop blade (Fig. 2A to E). Individual live oak trees within fungicide-treated plots received one of four fungicide applications: high-volume bole injections, low-volume bole injections with two types of microinjectors, and soil applications. All four fungicide application methods utilized the microencapsulated (blue) 14.3% EC formulation of propiconazole (Alamo) without xylene. Fungicide treatments were applied 23 to 27 August 1993 in a completely randomized linear sequence within fungicide-treated plots, each containing two to three replicate trees per treatment. All bole-injection methods applied the fungicide under pressure at 1.5 psi through injection ports, one port per 15.4 cm of tree circumference. High volume injections utilized a Turfco model 490 Injector (Turf Industries, Inc., Austin, TX) pressurized with CO? and connected in a continuous series around the bole with tygon tubing and polyethylene Tinjection ports. The ports, inserted into 7mm holes drilled into exposed root flares approximately 10 to 12 cm below the soil surface, were used to apply the fungicide at a rate of 3 ml/liter H,0/6.4 cm tree diameter at breast height (dbh). Microinjections Fig. 1. Experimental d e s i g n a n d l a y o u t o f sewn p l o t t r e a t m e n t s p l u s h e a l t h y a n d i n o c u l a t e d c o n t r o l s i n one o f th ree rep l ica tes ad jacent to an expanding oak w i l t in fec t ion cen te r . Tes t t rees assoc ia ted w i th t rench t rea tments were loca ted jus t ou ts ide o f the t rench on t h e s i d e o p p o s i t e t o t h e i n f e c t i o n c e n t e r . T r e a t m e n t s a p p l i e d t o p l o t s i n 199.1 i n c l u d e d : H C = h e a l t h y c o n t r o l s ; I C = i n o c u l a t e d c o n t r o l s ( c i r c l e d ) ; N T = n o t r e n c h ; T = t r e n c h o n l y ; T + B = t r e n c h + Bioharrier insert; T + F = tl-ench + fungicide; T + G20 = trench + Geomembrane 20 mil insert: T + G30 = trench + Geomemhrane 30 mil insert; and T + T = trench + Typar inser t . Dash l i n e i n d i c a t e s v i s i b l e f r o n t e d g e o f i n f e c t i o n c e n t e r i n 1993. a n d b o l d wrows i n d i c a t e d i r e c t i o n t h e f r o n t w a s m o v i n g . O p e n and closed tree symbols indicate asymptomatic and symptomatic trees in 1993, respectively. while circled, closed tree symbols indicate asymptomatic trees inocula ted in 1994 . 1068 Plant Disease / Vol. 86 No. 10 (strain SHL-TX-36; ATCC 200432), isolated from the adjacent infection center (3(l), and incubated for 1 week without shaking in 0.5% neopeptone-glucose broth (26). The inoculum was ground with a blender for I to 2 min within the broth prior to quantitation and inoculation. Soil excavations of trench inserts. Root growth in relation to trench inserts was assessed the fifth year following challenge inoculations by random spot root excavations in plots of each treatment to sample soil to a maximum depth of 1.5 m on the inside of the trench for each of the four trench insert types. Determinations were made as to whether root growth occurred over the inserts in cases where the barrier material was buried too deeply below the soil surface. Root growth at the bottom of the trench was assessed only in cases where trench breakouts (symptomatic trees found outside of the trench) occurred in sections with trench inserts and when it was suspected that growth of roots may have occurred under the barrier material. Insert materials also were examined for root contacts and root penetrations in each sampled trench section. The effects of insert materials on root growth were scored based on whether root growth was directionally diverted upon contact with insert materials, exhibited dichotomous branching, grew around insert barriers. was inhibited by chemical action, and whether roots had swollen apices. Data collections and analysis. Data from test trees within all research plots were collected and evaluated on an annual basis during a b-year sequential period (1995 to 2000), with the exception of year 5 (I999), when no data were collected. These data were compared against inoculatcd control trees inside of the trench, healthy control trees well outside of the trench, and infected breakout trees that became infected by root transmission beyond test trees outside of the trench. Trench breakouts, percent tree infection, and percent tree mortality were measured. The incidence of trench breakouts was noted per 183 m of barrier within each of three trench segments. The mean distance of symptomatic trees from the trench was recorded when trench breakouts occurred. Crown symptom ratings, percent branch mortality, percent defoliation, and crown light transmission were recorded as indications of disease severity. Crown symptoms were rated using the following scale: 1 = crown dead, totally defoliated, or with only necrotic leaves attached, 2 = thinning crown with leaves having diagnostic oak wilt symptoms, 3 = crowns containing foliage with chlorosis or reduced leaf size, but lacking diagnostic symptoms of oak wilt. and 4 = full, healthy crown with no apparent foliar symptoms. Veinal necrosis is considered the most diagnostic foliar symptom of oak wilt in Texas live oaks. Veinal necrosis is often accompanied by marginal necrosis in later slages of leaf symptom development. Crown light transmission, indicating the percentage of total available sunlight passing through the crown, was calculated from lux units recorded with an Extech light meter (Extech Instruments Corp., Waltham. MA) under the crown relative to direct sunlight. Sapwood water content was measured with a Protimeter Digital Timbermaster moisture probe (Protimeter Inc., Commack, NY). Disease severity percentage values were arcsine transformed prior to analysis. Significant differences among means were determined according to Fisher’s LSD tests after GLM analysis. RESULTS Trench insert and fungicide barrier tests. During the first 2 years, disease progressed slowly toward the trench from inoculated and naturally infected trees in the adjacent infection center. Only inoculated control trees located inside of cofitainment trenches, used to provide additional challenge to test barriers, expressed diagnostic symptoms of oak wilt during the first year after inoculation. Almost 60% of inoculated control trees had oak wilt symptoms, and 44% were dead due to oak wilt after I year (Table 1). Inoculated controls had considerable decline in crown density due to effects of the disease on foliage as well as associated decreases in crown symptom ratings and considerable increases in branch mortality the first year (Fig. 3A and B). An appreciable decrease in crown symptom rating and increase in defoliation and branch mortality also occurred in trees within no-trench plots during the first year, but no diagnostic oak wilt symptoms were as yet detected (Fip. 3A to C). The advancing front of the infection center moved unimpeded via root transmission into plots lacking trenches (no trench controls) during the second year, although only trees in one of these plots exhibited leaf veinal necrosis. Nevertheless, trees in no-trench control plots developed crown injury at levels that approached those observed in inoculated controls during the first year (Fig. 3A to D). Almost a third of all trees within plots without trenches were infected based on considerable defoliation and crown decline, and almost 14% mortality was observed in these trees by the end of the second year (Table 1). Diagnostic symptoms were found in leaves on the ground, but there were limited numbers of symptomatic leaves on living trees due to drought conditions during the evaluation. However, inoculated controls developed oak wilt symptoms much more rapidly and had higher disease incidence and severity than trees that become infected by root transmission. Drought conditions that prevailed throughout the summer months during the second evaluation year exacerbated disease development, especially branch death. The advancing front of the infection center continued to move unimpeded into two no-trench plots in year 3, resulting in increased incidence of infection and tree mortality relative to year 2. The occurrence of diseased trees outside of containment Table 1. Disease progress associated with irench trcatn~ents indicated by the incidence of trench breakou&, percent infection. and mortality of test trees during the first four years following inoculations of challenge trees insidc conlainmcnt harriers) Treuch breakouts” Percent symptomatic Percent mortality