Forest Ecology and Management 160 52002) 177±187

Effects of prescribed ®re on canopy foliar chemistry and suitability for an insect herbivore L.K. Rieskea,*, H.H. Housmanb, M.A. Arthurb aDepartment of Entomology, University of Kentucky, S-225 Ag. Science North, Lexington, KY 40546-0091, USA bDepartment of Forestry, University of Kentucky, T.P. Cooper Building, Lexington, KY 40546-0073, USA Received 8October 2000; accepted 20 January 2001

Abstract

We assessed the effects of a prescribed ®re on the phytochemical characteristics and vigor of overstory chestnut oak, scarlet oak, and red maple, and measured the impact of potential changes on herbivore ®tness. We compared foliar carbohydrates, tannins, nutrients, and ®ber concentrations in foliage from burned and non-burned forest canopies. There were signi®cant differences in most foliar characteristics between tree species. Total non-structural carbohydrate concentrations in scarlet oak and red maple foliage, and calcium levels in red maple foliage, were signi®cantly lower in burned plots, but other phytochemical characteristics were largely unaffected by burning. Tree growth also varied with species. Burning increased chestnut oak relative growth, decreased scarlet oak growth, and had no affect on red maple growth. Scarlet oak and red maple foliage from burned and non-burned forest canopies were assayed for gypsy moth performance. Caterpillars fed foliage from scarlet oak, the preferred host, grew larger and developed more rapidly than did those fed red maple foliage. There were no signi®cant burn treatment differences in caterpillar development within either tree species, suggesting that managers using prescribed ®re to promote oak regeneration need not worry about enhancing forest stand susceptibility to gypsy moth. However, the ®re in this study was of low to moderate intensity; more intense ®res may alter foliar palatability. # 2002 Elsevier Science B.V. All rights reserved.

Keywords: Oak; Quercus; Red maple; Phytochemistry; Gypsy moth; Defoliation

1. Introduction poor light regimes and vegetative competition from more aggressive, invasive species such as red maple Since the implementation of effective ®re suppres- 5Acer rubrum L.) 5Abrams, 1992). Lack of oak regen- sion in the early to mid-1900s, forest stand composi- eration also has signi®cant ecological consequences, tion in eastern North America has shifted to include including impacts on subsequent forest successional greater dominance by shade tolerant species, and age stages, and shifts in wildlife composition and distribu- and species mosaics have declined 5Delcourt and tion. The shift in species composition away from oak Delcourt, 1998). As a result, seedlings in oak 5Quercus dominance in eastern deciduous forests has resulted spp.) dominated forests have become suppressed by in the decline of an extremely valuable hardwood group. Repeated attempts to reduce the decline of oaks and improve regeneration suggest that ®re sup- * Corresponding author. Tel.: ‡1-859-257-7457/1167; fax: ‡1-859-323-1120. pression is an important causal factor 5Gammon et al., E-mail address: [email protected] 5L.K. Rieske). 1960; Beck and Hooper, 1986; Lorimer, 1989). Fire

0378-1127/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0378-1127501)00444-3 178 L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 can reduce competition and increase sunlight, allow- defoliation by the gypsy moth 5Gottschalk, 1993). ing oaks to thrive 5Carvell and Tyron, 1961; Selective harvesting, high-grading, and prescribed Gottschalk, 1985; Muick and Bartolome, 1987). Com- burning are potential methods to enhance stand vigor petitors may effectively resprout after a single ®re, but and manipulate forest species composition, thereby many oak competitors, including red maple, are suc- decreasing stand susceptibility to this and other cessfully suppressed by repeated burning 5Lorimer, invasive arthropod and plant species. Fire may impact 1985; Arthur et al., 1998). As a result of these potential herbivore populations directly by altering habitat, bene®ts of prescribed burning on oak regeneration, abundance, and species composition 5Siemann et al., prescribed ®re is being used by managers, with poten- 1997), or indirectly via cascading effects caused by tial consequences for the overstory oak. alterations in food quality and availability. Indirect Effects of ®re on overstory oaks may include effects of ®re on herbivory may manifest themselves impacts on plant vigor, alterations in susceptibility to through changes in foliar characteristics that impact mortality agents, and ®re damage to residual stems. herbivore success. For example, ®re may elicit Fire may in¯uence plant growth and foliar chemistry accumulation of foliar nutrients, resulting in enhanced by increasing nutrient concentrations in the soil nutritional substrate that could ameliorate induced 5Prieto-Fernandez et al., 1993; Blankenship and plant defenses 5Haukioja et al., 1990). Fire may also Arthur, 1999). A transitory increase in foliar nitrogen, cause plants to direct additional resources to growth potassium, and phosphorus in tree seedlings has been processes, maintaining productivity but reducing observed directly following prescribed ®re in some defensive abilities 5Price, 1991; Hunter and Schultz, systems 5Reich et al., 1990), which could be due to a 1995). Conversely, ®re may deplete foliar nutrients post-®re increase in available soil nutrients, and which and enhance production of defensive compounds. This may manifest itself as increased seedling growth rates is particularly relevant to the gypsy moth, which is 5Adams and Rieske, 2001), or increased growth of responsive to enhanced nutritional substrate 5Roth overstory trees 5Boerner et al., 1988). Defensive et al., 1994; Rieske and Raffa, 1998) and alterations in phenolic compounds may also be affected by the defensive phenolic compounds 5Baldwin and Schultz, increase in soil nutrients 5Hunter and Schultz, 1995) 1983; Schultz, 1988). or increased sunlight 5Dudt and Shure, 1994). Given Regeneration of oak stands and suppression of the potential effects of ®re on oak phytochemistry, invasive native and non-native plants and animals are foliar nutritional suitability and susceptibility to mam- emerging issues in deciduous forest management. The malian and arthropod herbivores may be drastically objectives of our study were to characterize the effects altered. of ®re on foliar chemistry of canopy oaks and red The gypsy moth 5Lymantria dispar L., Lepidoptera: maple, to determine the effects of any ®re-induced Lymantriidae) is an introduced defoliator with out- phytochemical changes on preferred and non-preferred break potential that has become established through- host suitability to an insect folivore, and to assess how out the northeastern US. The successful expansion of these potential changes may impact deciduous forest the geographic range of the gypsy moth into the composition. central hardwood forests of the Appalachian Cumber- land Plateau is assured by an abundance of several oak species, which are the insects' preferred host plants 2. Methods 5USDA Forest Service, 1991; Leibhold et al., 1995). The impacts of gypsy moth outbreaks can be devasta- We assessed the impacts of a prescribed burn on ting, and include shifts in forest stand composition the foliar chemistry and vigor of canopy chestnut oak away from oak dominance 5Campbell and Sloan, 5Q. prinus L.), and scarlet oak 5Q. coccinea Muenchh.), 1977; Gottschalk, 1993; Fajvan and Wood, 1996; both preferred hosts of the gypsy moth, and red maple, Davidson et al., 1999), similar to shifts in stand a non-preferred gypsy moth host. We also assessed the composition due to ®re suppression. suitability of the burned and non-burned scarlet oak Natural resource managers are increasingly turning and red maple foliage for the development of gypsy to silvicultural techniques to mitigate the effects of moth caterpillars. L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 179

2.1. Study sites and sampling shotgun to remove terminal branches on 6 May 1998, corresponding to leaf-out, and on 20 and 28May, 3 This study was conducted in the Red River Gorge and 17 June. On the ®rst, second, fourth, and ®fth Geological Area of the Daniel Boone National Forest sample dates, foliage was designated for phytochem- in eastern Kentucky 5USA), in conjunction with the ical analysis, including foliar non-structural carbohy- prescribed burn management program implemented drates, tannins, nutrients, and ®ber. On the third 528 by the US Forest Service. The Red River Gorge is May) sample date only, a smaller set of foliar samples located in the Cliff Section of the Cumberland Plateau N ˆ 16† from scarlet oak and red maple were dedi- and is characterized by xeric ridge tops supporting a cated to gypsy moth feeding trials. diverse suite of oaks, including scarlet oak, chestnut oak, white oak 5Q. alba L.), and black oak 5Q. velutina 2.2. Phytochemical analysis Lam.), several species of pines 5Pinus rigida Mill., P. echinata Mill., and P. virginiana Mill.), red maple, Leaves from each branch were removed at the base and mesic coves of beech 5Fagus grandifolia Ehrh.), and divided into two groups; one was ¯ash frozen in Magnolia spp., Rhododendron spp., and hemlock liquid nitrogen for analysis of carbohydrates and 5Tsuga canadensis L.) 5Braun, 1950). tannins, and the other was oven dried for nutrient and Our experiment was conducted on a single ridge ®ber analysis. Foliage that was ¯ash frozen was top. Four blocks were randomly located on the ridge, immediately ground into a ®ne powder using a mortar each of which were divided into two 20  20 m2 plots. and pestle, returned to the laboratory and stored at Prescribed ®re was ignited across the entire ridge; one À808C. Tissue was then freeze-dried 5VirTis Free- plot in each block was excluded from the ®re with ®re zemobile 12SL, The VirTis, Gradiner, NY) for lines. Each non-burned plot had an additional 5 m approximately 36 h and stored at À808C prior to wide non-burned buffer to minimize edge effects. analysis of foliar total non-structural carbohydrates Thus each of the four sections contained a burned area and foliar tannins. Foliar carbohydrate analysis was paired with a non-burned control. conducted spectrophotometrically with an anthrone/ Fires were started with a drip torch by ®ring a line thiourea reagent 5Quarmby and Allen, 1989). Foliar from the highest point and the ridges ®rst, and pulling tannin levels were analyzed using a radial diffusion strips downslope into the wind from the ridges. Point- assay 5Hagerman, 1987), with a protein-based agar to source and strip ®ring were used to increase intensity serve as the substrate for tannic acid binding. to acceptable levels if backing and ¯anking ®res were Leaves used for nutrient analysis were excised at the not of suf®cient intensity 5>0.3 m ¯ame length). At leaf base, packed in wet paper towels, stored on ice, the time of the prescribed ®re on 2 April 1998, wind and returned to the laboratory. Foliar leaf area 5LI- speed was 1±3 km/h, relative humidity was 36±42%, 3100 Area Meter, LI-COR, Lincoln, NE) and fresh and air temperature was 9±10.58C. The ®re was a cool weight were immediately measured. After a minimum surface burn, with ¯ame lengths of 30±90 cm which of 6 days drying at 608C, dry weight measurements consumed primarily the previous year's leaf litter were recorded and leaf area ratio 5cm2/g dry wt) was 5D. Richardson, pers. comm.). calculated. Leaves were then ground to a ®ne powder Individual trees selected for sampling were healthy and stored in a desiccator prior to nutrient analysis. and at least 5 m from any burned/non-burned inter- Foliar nutrient concentrations were assessed using a face; most were located more than 10 m from this block digestion method described by Isaac and interface. Canopy chestnut oaks had an average Johnson 51976). Foliar acid-digested ®ber 5ADF) diameter 5dbh) of 27.6 cm, scarlet oaks averaged analysis was conducted by the University of Wiscon- 35.9 cm dbh, and the red maple, which were smaller sin Soil and Plant Analysis Laboratory 5Madison, WI). co-dominants, averaged 16.0 cm dbh. Foliar samples were obtained from two trees of each 2.3. Tree growth species in each plot by treatment combination N ˆ 48†. Foliage was sampled from the south side Using an increment borer, we sampled the radial of the middle third of the canopy using a 12 gauge growth of each of the 48sample trees 5 N ˆ 16 per 180 L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 species) in September 2000. To estimate basal area length of caterpillar stadium 5duration of third instar growth since 1998, the year of the burn, we estimated in days) were calculated as measures of insect the basal area increment for each tree using diameter performance. at breast height 5dbh) and the sum of the last three years of radial growth. To determine whether trees on 2.5. Statistical analysis burned areas had different rates of growth than those on unburned areas, we calculated the relative growth Data for foliar chemistry and insect assays were rate of basal area since burning 5cm cmÀ1 timeÀ1)as analyzed using analysis of variance 5PROC MIXED; follows 5Hunt, 1990): SAS Institute, 1997) on four blocks. We analyzed the dependent variables describing foliar characteristics ln final basal area†Àln initial basal area† ; 5carbohydrate, tannin, nutrient, and ®ber concentra- time 2 À time 1 tions) across all three tree species before analyzing where ®nal basal area was the basal area of the tree in each tree species separately for burn effects. All 2000, initial basal area was the basal area of the tree in pairwise comparisons were made using a least 1997, and time 2 À time 1 was 3 years. signi®cant difference 5LSD) test. Tree growth data were analyzed using a two-way ANOVA with species 2.4. Insect assays and treatment as the main effects. A signi®cant interaction term P ˆ 0:01† led to one-way ANOVA Gypsy moth performance assays were conducted on each species separately. Caterpillar RGR and on scarlet oak and red maple from burned and non- stadium length also were analyzed for species burned plots using leaves collected only from the differences using ANOVA before analyzing for burn 28May 1998sample date. Gypsy moth egg masses effects within species. All signi®cant differences in were obtained from the USDA-ARS laboratory 5Otis caterpillar performance were assessed using Fisher's AFB, MA, USA) and were surface sterilized with protected LSD. a 0.1% sodium hypochlorite solution. Prior to use in assays, larvae were reared on wheat germ based arti®cial diet 5Southland Products, Lake Village, AR) 3. Results in growth chambers with a 15:9 5L:D) photoperiod at 238C. Newly molted third instar caterpillars were 3.1. Phytochemical analysis chosen at random from each of the 15 different egg masses. Caterpillars were starved for 24 h prior to Foliar non-structural carbohydrates. The three use. tree species differed signi®cantly in foliar total non- Caterpillar performance was assessed by allowing structural carbohydrate concentrations 5Table 1; F ˆ third instar larvae to feed for the duration of the 333:30, P < 0:0001), with levels in chestnut oak the stadium. Whole leaves were weighed and placed in highest and levels in red maple the lowest. Overall, ¯orists' water picks in 7 cm  21 cm clear plastic foliar total non-structural carbohydrate levels were rearing boxes. One-third instar caterpillar was placed signi®cantly lower in foliage from burned trees than in each rearing box and monitored at 24 h intervals for foliage from non-burned trees across all three species the duration of the stadium. At each monitoring F ˆ 4:67; P ˆ 0:0473†, with a strongly signi®cant interval, larvae were weighed, and at 2±3 day interval block effect F ˆ 6:40; P ˆ 0:0053†. Foliar carbohy- leaves were replaced to ensure freshness. Five insects drate concentrations also were affected by sample date were used for each of the eight trees of both scarlet oak across all species F ˆ 6:24; P ˆ 0:0009†. and red maple, for a total of 80 assays. At the Chestnut oak foliar total non-structural carbohy- completion of the assay, plant tissue, insect cadavers drate concentrations did not vary with burn treatment and waste material were oven dried at 608C for a 5Table 2) or date. Foliar carbohydrate levels were minimum of 10 days and weighed. Relative growth relatively high early in the season 5Fig. 1A), declining rate 5RGR ˆ‰caterpillar biomass gained mg†Š= during the second and third sampling intervals, and ‰ initial caterpillar dry wt mg†† time day††Š† and gradually increasing in the ®nal interval. L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 181

Table 1 Characteristics 5mean 5S.E.)) of canopy chestnut oak, scarlet oak, and red maple across site treatments 5burned and non-burned) on four sample dates in the Daniel Boone National Forest, Kentucky, 1998a

Species

Foliar component Chestnut oak Scarlet oak Red maple

Carbohydrates 5% dry wt) 2.49 50.07) a 2.09 50.06) b 0.94 50.05) c Tannins 5% dry wt) 0.14 50.01) b 0.34 50.01) a 0.11 50.01) c Calcium 5mg/g dry wt) 4.81 50.19) a 3.47 50.10) c 4.19 50.15) b Magnesium 5mg/g dry wt) 1.80 50.04) a 1.30 50.02) b 1.35 50.03) b Nitrogen 5mg/g dry wt) 24.4850.99) a 16.91 50.75) b 19.46 50.85)b Phosphorous 5mg/g dry wt) 2.37 50.14) a 1.70 50.08) b 1.66 50.08) b Fiber 5acid-digested, % dry wt) 35.67 50.98) a 25.29 50.62) b 27.17 51.19) b Leaf area ratio 5cm2/g dry wt) 203.53 56.49) a 167.92 55.18) b 214.65 56.98) a Relative basal area growth rate 5cm2/cm2 yr) 0.02 50.002) b 0.02 50.002) b 0.03 50.002) a

a Means within rows followed by the same letter are not signi®cantly different.

Scarlet oak foliage from trees on burned plots and then red maple. Overall foliar tannin levels were contained marginally lower levels of carbohydrates marginally lower in foliage from burned trees than than did foliage on trees in non-burned plots across foliage from non-burned trees across all three species all dates 5Table 2; F ˆ 5:36; P ˆ 0:1035). Carbohy- F ˆ 4:03; P ˆ 0:0610†, and there were no signi®cant drates in scarlet oak foliage were marginally affected differences between burn treatments for species by sample date F ˆ 2:89; P ˆ 0:0639†, with a strong analyzed individually 5Table 2). Foliar tannin con- date  burn interaction F ˆ 5:41; P ˆ 0:0079†. Ini- centrations across all species also were affected by tial carbohydrate levels were slightly greater in burned sample date F ˆ 24:11; P ˆ 0:0001†. than non-burned plots 5Fig. 1B; 6 May: P ˆ 0:0966), Foliar tannins from chestnut oak were strongly equivalent for the second 520 May) sample interval, affected by sample date F ˆ 54:43; P ˆ 0:0001†. and greater in non-burned plots by the ®nal two Tannin levels were relatively high early in the season sample dates 53 June: P ˆ 0:0777; 17 June: 5Fig. 1D), declining rapidly in the second and P ˆ 0:0011). subsequent sampling intervals. Tannin concentrations Red maple foliar carbohydrates were strongly in overstory scarlet oak foliage were also strongly affected by burning across sample dates 5Table 2; affected by sample date F ˆ 11:42; P ˆ 0:0001†. F ˆ 26:65; P ˆ 0:0141), with foliage from trees on Tannins in scarlet oak foliage were highest during burned plots containing signi®cantly lower carbohy- the ®rst two sampling intervals and declined steadily drate levels than foliage from trees on non-burned as the season progressed 5Fig. 1E), but were equivalent plots. Sample date also affected red maple foliar on burned and non-burned plots. Similar to the oak carbohydrate levels 5Fig. 1C; F ˆ 13:17; P ˆ 0:0001), species, foliar tannin levels in red maple were affected although there was no difference in foliar carbohy- by sample interval 5Fig. 1F; F ˆ 7:40; P ˆ 0:0014). drates in burned versus non-burned red maple foli- Tannin levels in red maple foliage increased slightly as age for each individual date. Initial carbohydrate the season progressed, whereas foliar tannin levels concentrations were high and declined dramati- declined seasonally in the oaks. Foliage from trees on cally in both burned and non-burned trees by 20 burned plots contained tannin concentrations similar May, increasing slightly during the ®nal two sample to those on non-burned plots at each sampling date. intervals. Foliar nutrients. There were strongly signi®cant Foliar tannins. There were strongly signi®cant dif- species differences in foliar nutrients, with chestnut ferences among the three species in foliar tannin levels oak containing higher levels of foliar calcium, mag- 5Table 1; F ˆ 547:01; P < 0:0001), with the highest nesium, nitrogen, phosphorous, and ®ber than either levels found in scarlet oak, followed by chestnut oak scarlet oak or red maple 5Table 1). The leaf area ratio 182 L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187

Table 2 Characteristics 5mean 5S.E.)) of canopy chestnut oak, scarlet oak, and red maple on burned and non-burned plots across four sample dates in the Daniel Boone National Forest, Kentucky, 1998a

Treatment

Foliar component Species Burned Non-burned F Pr > F

Carbohydrates 5non-structural) 5% dry wt) Chestnut oak 2.4850.036) a 2.50 50.036) a 0.15 0.724 Scarlet oak 1.9850.071) b 2.21 50.071) a 5.36 0.104 Red maple 0.90 50.011) b 0.9850.011) a 26.65 0.014 Tannins 5% dry wt) Chestnut oak 0.13 50.006) a 0.14 50.006) a 4.00 0.139 Scarlet oak 0.34 50.009) a 0.35 50.009) a 0.780.443 Red maple 0.11 50.008) a 0.11 50.008) a 0.58 0.503 Calcium 5mg/g dry wt) Chestnut oak 4.80 50.283) a 4.83 50.257) a 0.08 0.937 Scarlet oak 3.61 50.136) a 3.33 50.143) a 0.76 0.457 Red maple 3.7850.217) b 4.60 50.206) a 2.20 0.041 Magnesium 5mg/g dry wt) Chestnut oak 1.7850.054) a 1.8350.061) a 0.48 0.634 Scarlet oak 1.29 50.032) a 1.31 50.035) a 0.23 0.818 Red maple 1.32 50.046) a 1.37 50.037) a 0.50 0.624 Nitrogen 5mg/g dry wt) Chestnut oak 25.07 51.378) a 23.89 51.444) a 0.83 0.419 Scarlet oak 18.46 51.105) a 20.46 51.282) a 1.39 0.180 Red maple 15.9850.991) a 17.8451.104) a 1.29 0.212 Phosphorous 5mg/g dry wt) Chestnut oak 2.40 50.199) a 2.34 50.201) a 0.51 0.616 Scarlet oak 1.62 50.088) a 1.77 50.123) a 1.30 0.211 Red maple 1.62 50.091) a 1.70 50.122) a 0.66 0.517 Fiber 5acid-digested, % dry wt) Chestnut oak 35.3851.630) a 35.97 51.326) a 0.28 0.787 Scarlet oak 35.46 51.073) a 25.12 50.789) a 0.25 0.811 Red maple 26.55 51.503) a 27.7852.029) a 0.49 0.643 Leaf area ratio 5cm2/g dry wt) Chestnut oak 214.86 510.371) a 192.20 57.445) a 0.13 0.898 Scarlet oak 166.98 56.839) a 168.86 57.877) a 0.07 0.942 Red maple 216.55 59.448) a 212.74 510.408) a 0.33 0.745 Relative basal area growth rate 5cm2/cm2 yr) Chestnut oak 0.021 50.003) a 0.017 50.003) a 3.280.092 Scarlet oak 0.014 50.003) a 0.025 50.003) b 6.780.021 Red maple 0.032 50.003) a 0.02850.003) b 0.72 0.484

a Means within rows followed by the same letter are not signi®cantly different.

5cm2 gÀ1 dry wt) of chestnut oak foliage was 5Fig. 1G±R), but these were unaffected by burning. equivalent to that of red maple foliage, but exceeded Fiber levels and leaf area ratio were unaffected by that of scarlet oak F ˆ 15:200; P < 0:0001†. burning in all three species 5Table 2). Foliar nutrient concentrations in each of the three species were largely unaffected by burning 5Table 2). 3.2. Tree growth With the exception of calcium in red maple foliage, where non-burned foliage contained greater levels There were signi®cant differences in relative basal than burned foliage F ˆ 2:20; P ˆ 0:041†, magne- area growth rate between the three tree species F ˆ sium, nitrogen and phosphorous levels in foliage from 10:427; P ˆ 0:0002†, and a signi®cant species  burn burned trees were equivalent to those of non-burned interaction F ˆ 4:647; P ˆ 0:015†. The relative trees. Sample date was a signi®cant factor affecting basal area growth rate of scarlet oak and chestnut nearly all foliar nutrient levels in all three species oak were signi®cantly lower than that of red maple L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 183

Fig. 1. Seasonal phytochemical characteristics of burned 5±*±) and non-burned 5±&±) canopy chestnut oak, scarlet oak, and red maple in the Daniel Boone National Forest, Kentucky, 1998. Total non-structural carbohydrates 5A±C), tannins 5D±F), calcium 5G±I), magnesium 5J±L), nitrogen 5M±O), and phosphorous 5P±R).

5Table 1). Chestnut oak growth rate was marginally burned plots when compared to non-burned plots greater on burned plots than on non-burned plots F ˆ 6:778; P ˆ 0:0208†. There was no signi®cant 5Table 2; F ˆ 3:280; P ˆ 0:0916). In contrast, scarlet difference in growth rate between burned and non- oak relative growth rate was signi®cantly lower on burned red maple. 184 L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187

Table 3 Performance of third instar gypsy moth caterpillars fed on foliage collected on 28May 1998from canopy scarlet oak and red maple across site treatments 5burned and non-burned) in the Daniel Boone National Forest, Kentuckya

Insect performance parameter Scarlet oak Red maple

Relative growth rate 5mg/mg/day) 0.289 50.021) a 0.124 50.016) b Stadium length 5day) 6.92 50.174) b 10.11 50.415) a

a Means within rows followed by the same letter are not signi®cantly different.

Table 4 Performance of third instar gypsy moth caterpillars on canopy scarlet oak and red maple foliage from burned and non-burned sites in the Daniel Boone National Forest, Kentuckya

Scarlet oak Red maple

Caterpillar performance Burned Non-burned Burned Non-burned

Relative growth rate 5mg/mg day) 0.301 50.032) a 0.27850.027) a 0.13850.022) a 0.107 50.025) a Stadium length 5day) 6.895 50.314) a 6.947 50.143) a 9.850 50.504) a 10.350 50.678) a

a Means within rows within species followed by the same letter are not signi®cantly different.

3.3. Insect assays maple defense. Levels of foliar nutrients 5calcium, magnesium, nitrogen and phosphorous) were highest Caterpillar performance was assessed on scarlet in chestnut oak and intermediate to low in the red oak and red maple foliage. Insects fed scarlet oak maple. Fiber content and leaf area ratio also were foliage had relative growth rates over two times greatest in chestnut oak. greater than did those fed red maple foliage 5Table 3; The phytochemistry of canopy foliage was largely F ˆ 38:974; P < 0:0001). Caterpillars fed scarlet oak unaffected by burning. With the exception of decrea- foliage also developed more rapidly and had a shorter sed non-structural carbohydrates in scarlet oak and stadium length than did those fed red maple foliage red maple foliage, and calcium in red maple foliage, 5Table 3; F ˆ 49:988; P < 0:0001). There were no our prescribed ®re did not impact levels of the foliar signi®cant differences in caterpillar development components we measured. Tannin levels were not within either tree species based on burn treatment impacted by burning in any species. Surprisingly we 5Table 4). saw no differences in foliar nutrient levels due to burning. Reich et al. 51990) found signi®cant increases in foliar nutrients, including nitrogen, phosphorous, 4. Discussion and potassium, following a prescribed ®re. However, that studies was conducted with seedling material, The differences in canopy foliar chemistry among whereas our analysis was conducted on foliage sampled the three species were expected. The two oak species from the canopy of older trees. contained fairly high levels of foliar total non- Tree growth, as measured by relative growth rate of structural carbohydrates compared to red maple. basal area, was greatest in red maple, and equivalent Scarlet oak contained over two times the tannin between the two oak species. Red maple growth concentration of chestnut oak; red maple had the exceeded that of both oak species on burned and lowest tannin concentration. Red maple foliage is unburned sites, with no growth response to burning. rich in defensive alkaloids 5Barbosa et al., 1990), However, burning can easily damage even large red and tannins, which are not a major component of red maple trees because of its thin bark 5Burns and maple foliage, most likely play a minor role in red Honkala, 1990). A more severe burn would very likely L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 185 have damaged red maple trees, resulting in reduced was absent. An additional explanation for the lack of a growth rates. Chestnut oak relative growth rate burn effect may be that mid- to overstory trees are responded somewhat positively to burning, whereas units too large to be affected by late-winter prescribed scarlet oak growth response to burning was strongly burning, and the buffering capacity of the tree may negative. This contrasts with the increased growth exceed our ability to detect any differences. rates of both white and chestnut oak found by Boerner These results have useful implications with respect et al. 51988) after prescribed burning in the New Jersey to the use of prescribed ®re in managing deciduous Pine Barrens. Lower growth rate of scarlet oak after forest systems to maximize the oak component. Our burning may re¯ect the concurrent effects of recent data are compatible with other studies demonstrating droughts; the 1999 growing season had subnormal the compromise in red maple performance due to precipitation, with a Palmer drought index of À1.76 single and repeated ®res 5Lorimer, 1985; Arthur et al., for April 1998±March 1999 5University of Kentucky 1998). Our data also suggest that foliar palatability to Agricultural Weather Center, 1999), and scarlet oak on some herbivorous insects, including outbreak species these and other nearby sites, burned and unburned, such as the gypsy moth, may be unaffected by burning, show greater occurrence of crown die-back than and that managers making decisions regarding pre- chestnut oak 5Arthur, unpublished data). scriptions need not worry about enhancing forest stand The differences in gypsy moth caterpillar growth susceptibility to outbreak defoliators with light, late- and development between the scarlet oak foliage and winter burning. red maple foliage were expected. Scarlet oak, a pre- ferred host plant, grew larger insects which developed more quickly than did red maple, a non-preferred host Acknowledgements 5Leibhold et al., 1995). Accelerated larval develop- ment allows for avoidance of natural enemies and other We thank Greg Abernathy, Aaron Adams, Beth mortality factors, and is generally considered a Blankenship, Emily Bremer, Rebecca Brown, Lyle competitive advantage in insects 5Price, 1997). Addi- Buss, Peter Hadjiev, James Hamilton, Millie Hamil- tionally, larger caterpillars tend to develop into larger ton, Suzanne Murray, Ted Sizemore, and Michelle pupae. In the gypsy moth, pupal weights are directly Smith for assistance. The comments of Ken Yeargan correlated with fecundity 5Drooz, 1985), so that larger and two anonymous reviewers greatly strengthened insects produce more offspring and are therefore more this manuscript. This work was partially funded by a successful than their smaller counterparts. Challenge Cost-Share Agreement with the Daniel Consistent with the phytochemical characteristics, Boone National Forest, Winchester, KY, and was burning did not signi®cantly affect caterpillar perfor- made possible through collaboration with personnel mance. Although gypsy moth relative growth rates from the Stanton Ranger District. Financial support were greater, and development time less, on burned was also provided through the Kentucky State foliage versus non-burned foliage, these differences Agricultural Experiment Station 5No. 00-08-153). were not signi®cant. It is interesting to note, however, that foliar carbohydrates in burned scarlet oak and red maple tended to be less than non-burned foliage, in References contrast to studies in other herbivore/plant systems which suggest the phagostimulatory nature of foliar Abrams, M., 1992. Fire and the development of oak forests. BioScience 42, 346±353. carbohydrates 5Thorsteinson, 1960; Schoonhoven, Adams, A.S., Rieske, L.K., 2001. The effects of ®re on oak 1968). seedling growth and herbivore interactions. For. Sci. The lack of treatment 5burn) effects in this study Arthur, M.A., Paratley, R.D., Blankenship, B.A., 1998. Single and could be due to a number of factors. It is possible that repeated ®res affect survival and regeneration of woody and the ®re itself was too light to fully combust the litter herbaceous species in an oak±pine forest. J. Torr. Bot. Soc. 125, 225±236. layer, so that the nutrient release was not large enough Baldwin, I.T., Schultz, J.C., 1983. Rapid changes in tree leaf to manifest itself in canopy foliar chemistry. Scorch chemistry induced by damage: evidence for communication heights were considerably less than 1 m, and scarring between plants. Science 221, 277±279. 186 L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187

Barbosa, P., Gross, P., Provan, G.J., Pacheco, D.Y., Stermitz, F.R., Hagerman, A.E., 1987. Radial diffusion method for determining 1990. Allelochemicals in foliage of unfavored tree hosts of the tannin in plant extracts. J. Chem. Ecol. 13, 437±449. gypsy moth, Lymantria dispar L. 1. Alkaloids and other Haukioja, E., Ruoohomaki, K., Senn, J., Suomela, J., Walls, M., components of Liriodendron tulipifera L. 5Magnoliaceae), Acer 1990. Consequences of herbivory in the mountain birch 5Betula rubrum L. 5Aceraceae) and Cornus ¯orida L. 5Cornaceae). J. pubescens ssp. tortuosa): importance of the functional Chem. Ecol. 16, 1719±1730. organization of the tree. Oecologia 82, 238±247. Beck, D.E., Hooper, R.M., 1986. Development of a southern Hunt, R., 1990. Basic Growth Analysis: Plant Growth for Appalachian hardwood stand after clearcutting. South. J. Appl. Beginners. Unwin Hyman, London. For. 10, 168±172. Hunter, M.D., Schultz, J.C., 1995. Fertilization mitigates chemical Blankenship, B., Arthur, M.A., 1999. Soil nutrient and microbial induction and herbivore responses within damaged oak trees. response to prescribed ®re in an oak±pine ecosystem in eastern Ecology 76, 1226±1232. Kentucky. In: Stringer, J.W., Loftis, D.L. 5Eds.), Proceedings of Isaac, R.A., Johnson, W.C., 1976. Determination of total nitrogen the 12th Central Hardwood Forest Conference, February 28± in plant tissue, using a block digestor. J. AOAC 59, 98±100. March 2, 1999, Lexington, KY. US For. Serv. Gen. Tech. Rep. Leibhold, A.M., Gottschalk, K.W., Muzika, R.M., Montgomery, SRS-24, Asheville, NC. M.E., Young, R., O'Day, K., Kelley, B., 1995. Suitability of Boerner, R.E.J., Lord, T.R., Peterson, J.C., 1988. Prescribed North American trees species to gypsy moth: a summary of burning in the oak±pine forest of the New Jersey Pine Barrens: ®eld and laboratory tests. Gen. Tech. Rep. NE-211. US effects on growth and nutrient dynamics of two Quercus Department of Agriculture, Forest Service, Northeastern Forest species. Am. Midl. Nat. 120, 108±119. Experiment Station, Radnor, PA. Braun, E.L., 1950. Deciduous Forests of Eastern North America. Lorimer, C.G., 1985. The role of ®re in the perpetuation of oak Hafner, New York. forests. In: Johnson, J.E. 5Ed.), Challenges in Oak Management Burns, R.M., Honkala, B.H., 1990. Silvics of North America. 2: and Utilization. Univ. Wisc. Coop. Ext. Serv., Madison, WI, Hardwoods. US Department of Agricultural Handbook No. pp. 8±25. 654, Vol. 2. USDA, Washington, DC. Lorimer, C.G., 1989. The oak regeneration problem: new evidence Campbell, R.W., Sloan, R.J., 1977. Forest stand responses to on causes and possible solutions. In: Proceedings of the Annual defoliation by the gypsy moth. For. Sci. Monogr. 19, 34. Symposium of the Hardwood Research Council, Memphis, TN, Carvell, K.L., Tyron, E.H., 1961. The effect of environmental pp. 23±40. factors on the abundance of oak regeneration beneath mature Muick, P.C., Bartolome, J.W., 1987. Factors associated with oak stands. For. Sci. 7, 98±105. regeneration in California. Gen. Tech. Rep. PSW-100. Paci®c Davidson, C.B., Gottschalk, K.W., Johnson, J.E., 1999. Tree Southwest Forest and Range Experiment Station, US Depart- mortality following defoliation by the European gypsy moth ment of Agriculture, Forest Service, Berkeley, CA. 5Lymantria dispar L.) in the United States: a review. For. Sci. Price, P.W., 1991. The plant vigor hypothesis and herbivore attack. 45, 74±84. Oikos 62, 24±251. Delcourt, P.A., Delcourt, H.R., 1998. The in¯uence of prehistoric Price, P.W., 1997. Insect Ecology, 3rd Edition. Wiley, New York, human-set ®res on oak±chestnut forests in the southern Oikos 62, 244±251. Appalachians. Castanea 63, 337±345. Prieto-Fernandez, A., Villar, M.C., Carballas, M., Carballas, T., Drooz, A.T., 1985. Insects of eastern forests. US For. Serv. Misc. 1993. Short-term effects of a wild®re on the nitrogen status and Publ. No. 1426. Washington, DC, 608pp. its mineralization kinetics in an Atlantic forest soil. Soil Biol. Dudt, J.F., Shure, D.J., 1994. The in¯uence of light and nutrients on Biochem. 25, 1657±1664. foliar phenolics and insect herbivory. Ecology 75, 86±98. Quarmby, C., Allen, S.E., 1989. Organic constituents. In: Allen, Fajvan, M.A., Wood, J.M., 1996. Stand structure and development S.E. 5Ed.), Chemical Analysis of Ecological Materials. Wiley, after gypsy moth defoliation in the Appalachian Plateau. For. New York, pp. 160±201. Ecol. Manage. 89, 79±88. Reich, P.B., Abrams, M.D., Ellsworth, D.S., Kruger, E.L., Tabone, Gammon, A.D., Rudolph, V.J., Arend, J.L., 1960. Regeneration T.J., 1990. Fire affects ecophysiology and community dynamics following clearcutting of oak during a seed year. J. For. 58, of central Wisconsin oak forest regeneration. Ecology 71, 711±715. 2179±2190. Gottschalk, K.W., 1985. Effects of shading on growth and Rieske, L.K., Raffa, K.F., 1998. Interactions among insect development of northern red oak, black oak, black cherry, herbivore guilds: in¯uence of thrips bud injury on foliar and red maple seedlings. I. Height, diameter, and root/shoot chemistry and suitability to gypsy moths. J. Chem. Ecol. 24, ratio. In: Dawson, J.O., Majerus, K.A. 5Eds.), Proceedings of 501±523. the Fifth Central Hardwood Forest Conference, Urbana- Roth, S.K., Lindroth, R.L., Montgomery, M.E., 1994. Effect of Champaign, IL, SAF Publ. 85-05, University of Illinois, foliar phenolics and ascorbic acid on performance of the gypsy pp. 189±195. moth 5Lymantria dispar). Biochem. Syst. Ecol. 22, 341±351. Gottschalk, K.W., 1993. Gypsy moth effects on mast production. SAS Institute, 1997. SAS/STAT Software: Changes and Enhance- In: McGee, C.E. 5Ed.), Proceedings of the Southern Appala- ments through Release 6.12. SAS Institute, Cary, NC. chian Mast Management Workshop. University of Tennessee, Schoonhoven, L.M., 1968. Chemosensory bases of host plant Knoxville, August 14±16, 1986. selection. Annu. Rev. Entomol. 13, 115±136. L.K. Rieske et al. / Forest Ecology and Management 160 72002) 177±187 187

Schultz, J.C., 1988. Many factors in¯uence the evolution of herbivore USDA Forest Service, 1991. Gypsy moth: a balanced perspective. diets, but plant chemistry is central. Ecology 69, 896±897. USDA Forest Service Video, Southern Region, Forest Pest Siemann, E., Haarstad, J., Tilman, D., 1997. Short-term and long- Management. term effects of burning of oak savanna arthropods. Am. Midl. University of Kentucky Agricultural Weather Center, 1999. Nat. 137, 349±361. Department of Biosystems and Agricultural Engineering, Thorsteinson, A.J., 1960. Host selection in phytophagous insects. College of Agriculture, Lexington, KY. www: agvx.ca.uky.e- Annu. Rev. Entomol. 5, 193±218. du/all_ky_rain.txt.