Growth responses of yellow-poplar(Liriodendron tulipifera L.) exposed to 5 years of O3 aloneor combined with elevated CO2


Joanne Rebbeck.Fax: +1 740-368-0152; e-mail:


Field-grown yellow-poplar (Liriodendron tulipifera L.) werefumigated from May to October in 1992–96 within open-topchambers to determine the impact of ozone (O3) aloneor combined with elevated carbon dioxide (CO2) on saplinggrowth. Treatments were replicated three times and included: charcoal-filteredair (CF); 1 × ambient ozone (1 × O3);1·5 × ambient ozone (1·5 × O3);1·5 × ambient ozone plus 350 p.p.m.carbon dioxide (1·5 × O3 + CO2)(target of 700 p.p.m. CO2); and open-air chamberlessplot (OA). After five seasons, the total cumulative O3 exposure (SUM00 = sumof hourly O3 concentrations during the study) rangedfrom 145 (CF) to 861 (1·5 × O3) p.p.m. × h (partsper million hour). Ozone had no statistically significant effecton yellow-poplar growth or biomass, even though total root biomasswas reduced by 13% in the 1·5 × O3-exposedsaplings relative to CF controls. Although exposure to 1·5 × O3 + CO2 hada stimulatory effect on yearly basal area growth increment aftertwo seasons, significant increases in shoot and root biomass (∼ 60% increaserelative to all others) were not detected until the fifth season.After five seasons, the yearly basal area growth increment of saplingsexposed to 1·5 × O3 + CO2-air increasedby 41% relative to all others. Based on this multi-yearstudy, it appears that chronic O3 effects on yellow-poplargrowth are limited and slow to manifest, and are consistent withprevious studies that show yellow-poplar growth is not highly responsiveto O3 exposure. In addition, these results show thatenriched CO2 may ameliorate the negative effects of elevatedO3 on yellow-poplar shoot growth and root biomass underfield conditions.


Current projections suggest that forests within the eastern UnitedStates will continue to be exposed to increasing levels of troposphericO3 (Chameides et al.1994) and ambient CO2 (Watson et al.1990). Rural regions of eastern USA typically average summerdaytime levels in excess of 0·05 p.p.m. O3 (Lefohn, Edwards & Adams 1994) with occasionalO3 episodes above 0·100 p.p.m. TroposphericO3 has long been known to be phytotoxic (Middleton1956) and causes the greatest amount of damage on vegetationof any gaseous pollutant by reducing growth and productivity of manyplant species through reductions in photosynthesis, acceleratedleaf senescence and decreased root growth (Bortier,Ceulemans & de Temmerman 2000; Rudorff,Mulchi & Lee 2000). Elevated CO2 concentrationsgenerally increase plant growth through increased carbon assimilation,biomass and leaf area (Ceulemans & Mousseau.1994; Sax, Ellsworth & Heath 1998).It is hypothesized that plants growing in elevated CO2 maybe protected from O3 damage by reducing O3 uptakebecause elevated CO2 often reduces stomatal conductance(Eamus & Jarvis 1989; Mousseau& Saugier 1992). Elevated CO2 may also providesubstrates for detoxification and repair processes against O3 damage (Rudorff et al. 2000).

Empirical information on the combined effects of the O3 andCO2 on woody plants is limited (Allen1990; Olszyk et al. 2000).In some studies, elevated CO2 offsets the negative effectsof O3 in deciduous species (Volin,Reich & Givnish 1998; Broadmeadow &Jackson 2000), whereas in others no ameliorative effectsare detected (Kull et al. 1996).Multi-year field fumigations are needed to accurately assess thechronic response of tree species to the combination of elevatedCO2 and O3. This can be achieved by exposingthe same population of trees to gaseous pollutants during theirdevelopment from seedlings to saplings or mature trees (Lee & Jarvis 1995; Retzlaff,Williams & DeJong 1997; Rey & Jarvis1997). These studies are very costly to run and only a fewhave been undertaken, resulting in a paucity of empirical data onthe response of saplings and mature trees to gaseous pollutants(Karnosky et al. 1999; Isebrands et al. 2001). Thesetypes of studies can serve as an important linkage to extrapolateseedling pollutant response data to larger and more structurallycomplex older trees.

In the current study, we investigated how field-planted yellow-poplar(Liriodendron tulipifera L.) seedlings grown under less thanoptimal conditions (unmanipulated soil moisture and fertility) respondedto chronic levels of elevated O3 either alone or combinedwith elevated CO2 over five seasons. The effects of elevatedO3 on yellow-poplar, an ecologically and economicallyimportant hardwood species in eastern USA forests, have been wellstudied in potted seedlings under environmentally controlled conditions showingboth inhibitions and stimulations in growth and physiological processes(Chappelka, Chevone & Seiler 1988; Jensen & Patton 1990; Roberts1990; Cannon, Roberts & Bargar 1993; Rebbeck & Loats 1997; Loats& Rebbeck 1999). Long considered an O3-sensitivespecies, yellow-poplar is used as a bioindicator of O3 ineastern US forests (Davis & Skelly 1992).Our objectives were to determine: (1) if the responses of field-grownsaplings to multi-year O3 exposures were comparable withthose of seedlings in shorter-term O3 exposures undermore controlled conditions; and (2) if doubling ambient CO2 concentrations wouldameliorate negative O3 growth effects.

Materialsand methods

Seedlingculture and site characteristics

In 1991, a field plantation of yellow-poplar was established atthe Northeastern Research Station's Forestry Sciences Laboratoryin Delaware, Ohio, USA (latitude 40°21′ Nlongitude 83°04′ W), by clearing a 26 mby 73 m area in a 20-year-old-abandoned American elm (Ulmusamericana L) plantation. The glacial till soil was primarilyof the Blunt Series with pockets of the Morley Series. Soil textureswere determined to be primarily clay loams. Soils were chemicallyanalysed prior to planting to a depth of 25 cm and wereconsidered adequate in total N (24·7 µg g−1 NH4-N and8·2 µg g−1 NO3-N).Mean soil pH was 6·39 and CEC was 13·9 meq/100 gdry soil. Extractable ion concentrations were 7·4 µg g−1 P;103·5 µg g−1 K;1909 µg g−1 Ca;and 217·6 µg g−1 Mg.Soil chemical properties were measured annually; no major changesin concentrations were observed throughout the study.

One-year-old bareroot stock originating from seeds ­collectedin south-eastern Tennessee was obtained from a private nursery inwestern Pennsylvania, USA. Approximately 250 uniform bare-rootedseedlings were planted 2·1 m by 2·1 mapart in July 1991 to serve as buffer trees between the plots. InMay 1992 within each chamber or plot, 12 randomly chosen seedlingswere planted 47 cm apart in a circular pattern approximately1·8 m in diameter and 0·6 mfrom the inside edge of a standard 3-m-diameter open-top chamber(OTC) (Heagle, Body & Heck 1973) or equivalentchamberless plot. During the early establishment of seedlings in1992 and 1993, ambient rainfall was occasionally supplemented inorder to minimize severe water stress.

In 1992, methyl [1-[(butylamino)carbonyl]-1H-benzimidazol-2-yl] carbamate [Benomyl50E® (14·7 mg active ingredientL−1); DuPont, Wilmington, DE, USA] wasapplied biweekly from mid-August to mid-September as a prophylacticto control powdery mildew, a serious foliar disease of yellow-poplarin central Ohio that can kill young seedlings. As observed in previousstudies, Benomyl did not protect the seedlings from oxidant injury(Rebbeck 1996a; Loats& Rebbeck 1999). No other applications of pesticideswere made during the study. In 1993 and 1994, sporadic outbreaks ofaphids were observed with a fairly uniform distribution in all plots.Insecticidal soap [Safer® (39 mL L−1);Gardens Alive, Lawrenceburg, IN, USA] was applied biweeklyto suppress aphid populations. Biological control measures weresubsequently adopted and commercially reared ladybugs (Hippodamia sp.)were released biweekly from mid-July until late August during eachseason.

Pollutantexposure and experimental design

The experiment was a randomized block design with three replicatesof the following treatments: (1) charcoal-filtered air (CF); (2)1 × ambient O3 (1 × O3);(3) 1·5× ambient O3 (1·5 × O3);(4) 1·5 × ambient O3 plus350 p.p.m. CO2 above ambient (target CO2 concentrationwas 700 p.p.m) (1·5 × O3 + CO2);and (5) open-air chamberless plot (OA). Due to the prohibitive costsassociated with supplying twice ambient CO2 to threeadditional open-top chambers, we did not characterize the effectsof elevated CO2 alone on yellow-poplar growth. Gaseswere dispensed 24 h per day from mid-May until mid-Octoberin 1992–96. The polyvinyl-chloride film panels were removedeach season when gas exposures were terminated. In spring 1994,standard OTCs (3 m in diameter, 2·4 mhigh) were extended to a height of 4·6 m. Themean volumetric airflow within chambers was 47 m3 min−1 ± 10% andwas provided by a single blower. In spring 1995, standard OTCs werereplaced with larger open-top chambers (4·6 min diameter, 9·2 m high) similar to those developedby Heagle et al. (1989),but doubled in height. Air was blown into the bottom of each chambervia two 2HP, 91·4-cm-diameter fans (Penn Ventilator Inc.,Philadelphia, PA, USA), each housed in a galvanized sheet-metalbox (123 cm wide, 123 cm high, 123 cm long).Unfiltered ambient air entered each chamber after passing throughparticulate filters and connecting plastic ducts at approximately6·8 m3 s−1.

Ozone was generated from vaporized liquid oxygen with an electricspark discharge O3 generator (OREC Model 03V10-0; OREC,Glendale, AZ, USA) and was dispensed into the 1×O3 and1·5 × O3 OTC's throughTeflon tubing when ambient levels exceeded 0·03 p.p.m.(Rebbeck 1996a; Rebbeck& Loats 1997). Gaseous CO2 was vaporized from liquidCO2 (14 000 kg reservoir, MG Industries,Malvern, PA, USA) and dispensed through Teflon tubing into the OTCsand adjusted manually with needle valves (Part 8513D-2-E-4E-1 A,Brooks Instrument Division, Emerson Electric, Hatfield, PA, USA)to maintain target levels. The O3 concentration was regulatedwith mass-flow control valves (Model FC260, Tylan General, San Diego,CA, USA) through a microcomputer to match a set-point value basedon the most recent ambient O3 reading. Each chamber wasmonitored for O3 and CO2, and air temperature (subsetof chambers) every 2 min. Hourly means were automaticallycalculated and stored on a personal computer. Seasonal average O3 andCO2 concentrations were calculated for each chamber/plot.However, seasonal average concentration often does not correlatewith plant injury because it does not include important exposurefactors such as episodic peaks and the cumulative effects of lowO3 concentrations. To better relate plant responses toO3 treatments, cumulative O3 exposures foreach chamber/plot were estimated (Bortier et al.2000). Two cumulative exposure indices were calculated eachgrowing season: SUM00 (p.p.m. × h)that is equal to the sum of all hourly average O3 concentrations;and SUM06 (p.p.m. × h) isequal to the sum of all hourly concentrations above 0·059 p.p.m.Daily single-point and weekly multi-point calibrations of the O3 monitor(Model 49PS; Thermo Electron Instruments, ­Hopkinton, MA,USA) were made with a multi-gas ­calibrator (Model 8500;Monitor Laboratories, San Diego, CA, USA). CO2 levelswere monitored with a Li-Cor Model   6251   infrared   gas   analyser   (Li-Cor   Inc.,   Lincoln, NE,USA) with daily single-point calibrations and zero span checks,and weekly multi-point calibrations using certified CO2 spangases (AGA Industries, Cleveland, OH, USA). Ambient photosyntheticphoton fluence rate (PPFR) over the waveband 400–700 nm(Li-Cor quantum sensor), relative humidity (%RH) and airtemperature [Model XN217 (Hydrometrix RH sensor and Fenwal ­ElectronicsUUT51J1 thermistor) Campbell Scientific, Logun, UT, USA] weremonitored with a Campbell 21X datalogger.

Growthand biomass measurements

Date of budbreak and length of terminal stem dieback at the startof each growing season was recorded for each seedling beginningin 1993 (12 plants × 3 chambers × 5 treatments = 180plants). Stem height and diameter (permanently marked at approximately2·5 cm above soil surface) were measured in Julyand September 1992. In 1993–96, monthly stem height anddiameter measurements were made throughout each exposure season(May–September). Stem height measurements included totalheight (measured from tree base at the ground to tree apex) andyearly height growth increment (length of current year's stem).In late September 1993, the biomass of six yellow-poplar from eachchamber/plot (6 plants × 3 chambers × 5 treatments = 90plants, 90 plants were left in the ground) was determined by destructiveharvest. The root systems were excavated from a fixed volume (25 cm × 25 cm × 25 cm)of soil immediately surrounding the decapitated seedling and wereseparated into taproot and first-order laterals. The soil that had beenremoved was returned to its original location within each plot.In September 1994, three saplings were harvested (3 plants × 3 chambers × 5 treatments = 45plants, 45 plants were left in the ground) from each plot. The roots werenot excavated to minimize damage to the three remaining trees withineach chamber. In late September 1996, above-ground biomass of thethree remaining trees from each plot/chamber was harvested(45 trees). At this final harvest, the stem diameter was measuredon the portion of the main stem that was initiated at the startof the study (1992) and at a height of 50 cm from the ground.Stem basal area at 50 cm from the ground was calculated.In late winter 1997, the roots were excavated from a fixed volume ofsoil (91 cm top width, 71 cm long, 46 cmbottom width) using a commercial tree spade. Roots were separatedinto tap, first- and second-order laterals. In all harvests, thetotal plant height, length of current terminal, length and diameterof each year's height growth increment, basal stem diameter, branchlengths and numbers were measured. Leaf area was estimated usinga calibrated LI-3100 area meter (Li-Cor, Inc.) at every harvest.Following growth measurements, the tissue of each tree was separatedinto leaves, stem and branches, oven-dried at 70 °Cto constant mass [g dry weight (DW)] and weighed.For each sapling, the root : shoot ratio, specificleaf area (cm2 g−1), theratio of total leaf area to lateral roots, and leaf area ratio [LAR = totalleaf area/sapling mass (shoot + rootDW)] was calculated.


Treatment effects were tested using only the chambered plotsand comparisons were made between OA- and 1 × O3-exposedseedlings to assess the effect of chambers on seedling growth. Diameterand height growth for each year (1992–96) were analysedto test for overall treatment effects (CF, 1 × O3,1·5 × O3 and 1·5 × O3 + CO2 plots)with a randomized block design using repeated measures analysis(RMA) (SAS System for Mixed Models; Littell et al. 1996).Repeated measures analysis was used to compare treatment means overthe five growing seasons. The covariance structure of the growthresponse data (height, basal diameter and basal area increments)was determined to be autoregressive, in that measurements takenat adjacent time points were more highly correlated than measurementstaken several time points apart. A mixed model with random [blockand time (year)] and fixed (treatment) effects with anautoregressive covariance structure was used. The three chamberreplicates per treatment were used in the analyses. Least squaremeans ± 1 SE are presented. Apriori single degree of freedom contrasts were used to ascertaindifferences among treatments over time. In addition, multiple degreeof freedom contrasts were used to determine treatment differencesat specific time points. Biomass data were analysed using GeneralLinear Model to test for significant treatment effects using chambermeans (n = 3) (SAS1999). Effects were considered significant if P < 0·10unless otherwise stated.


Pollutantand environmental conditions

Mean monthly ambient air temperatures for June to September 1992–96ranged from 14·1 to 24·7 °C.Air temperatures within OTCs were approximately 1 °Cabove ambient. During the winter months (December–March), minimumair temperature ranged from −32·6 to 10·7 °Cand maximum air temperature ranged from −19·8to 25·1 °C. The coldest temperaturesoccurred in January 1994 during two days of record-breaking lowtemperatures. From June to September PPFR ranged from 21·8to 43 mol m−2 d−1. Monthlyrainfall during each growing season varied year to year from thelong-term 50-year average for the site (Fig. 1). Annual precipitationwas 914, 1024, 677, 1143 and 1125 mm in 1992–96,respectively, in comparison with the long-term average of 993 mm.

Figure 1.

Monthlyrainfall (mm) patterns during each growing season (1992–96)compared with the 50-year average.

Actual O3 levels were as follows: CF ranged from 0·15 × to0·30 × ambient O3;1 × O3 ranged from 0·98 × to1·03 × ambient O3; 1·5 × O3 rangedfrom 1·31·5 × to 1·68 × ambient O3;and 1·5 × O3 + CO2 rangedfrom 1·20 × to 1·58 × ambientO3. Cumulative O3 exposure (p.p.m. × h)was calculated as both SUM00 (sum of every hourly mean)and SUM06 (sum of all hourly O3 values exceeding0·059 p.p.m) for each season (Table 1). In 1992–96,51, 46, 7, 84 and 89%, respectively, of the exposure dayshad daily hourly maximum ambient O3 concentrations whichexceeded 0·050 p.p.m. Daily levels of hourly maximumambient O3 exceeded 0·080 p.p.m. for6·3% of the exposure days in both 1992 and 1993,compared with 0, 30 and 36% in 1994, 1995 and 1996, respectively.Seasonal 24 h mean ambient O3 concentration rangedfrom 0·032 to 0·046 p.p.m.

Table 1.  Seasonal24-h mean O3 concentration and seasonal cumulative O3 exposure a (SUM00 andSUM06) during five seasons of exposure in Delaware, Ohio,USA 1992–96
TreatmentSeasonal 24 hmean ozone concentration (p.p.m)
CF (control)0·24×0·0100·0050·0110·0120·011
1 × O31·00×0·0350·0320·0350·0420·048
1·5 × O31·48×0·0460·0510·0490·0590·078
1·5 × O3 + CO21·42×0·0420·0500·0470·0610·073
TreatmentSeasonal cumulativeozone exposure (p.p.m. × h)
19921993199419951996Total (1992–96)
  1. a Seasonalcumulative ozone exposure: SUM00 equals the sumof each daily 24 h total exposure for the entire exposureseason. SUM06 equals the sum of each daily hourly meanthat exceeded 0·059 p.p.m. O3 duringthe entire exposure season.

  2. b OA,open-ambient air chamberless treatment; CF, charcoal-filtered air;1 ×O3,1 times ambient ozone; 1·5 × O3, 1·5times ambient ozone; 1·5 × O3 + CO2, 1·5times ambient ozone plus 2 times ambient carbon dioxide (targetCO2 concentration was 700 p.p.m). Ozone wasadded to all 1×O3,1·5 × O3 and1·5 × O3 + CO2 chambers.

OA10322·0105 17·3128 33·9120 48·8126 60·0582182·0
CF 27 0·3 16 0·0 39 0·0 36 0·0 27 0·0145 0·3
1 × O310434·9107 23·6125 42·6117 54·6130 73·0583228·7
1·5 × O313681·2171135·4176121·7166131·3212192·2861661·8
1·5 × O3 + CO212467·6166129·4170114·1172136·5198176·7830624

Ambient CO2 concentrations varied diurnally, witha daily mean of 355 p.p.m., ranging from 330 p.p.m.(early to mid-afternoon) to 455 p.p.m. (pre-dawn). In the 1·5 × O3 + CO2 plots,seasonal mean CO2 concentrations were 653, 727, 740,688 and 728 p.p.m. in 1992–96, respectively. CO2 concentrationswithin all other chambered plots did not vary significantly fromthe open-air plots. Diurnal patterns of both O3 and CO2 inthese plots were similar to those found under ambient conditions.


In general, trees broke bud in late April each year. Mean dateof bud burst was 20 April, 23 April, 17 April and 23 April in 1993–96,respectively. Date of bud burst was not affected by exposure to1·5 × O3- or 1·5 × O3 + CO2-air throughoutthe study. On average, all trees broke bud within a week of oneanother each season. In spring 1993, dieback of the main terminalstem averaged 10 cm in CF-, 1 × O3-and 1·5×O3-air compared to 23 cmin 1·5 × O3 + CO2-air(P = 0·012). No significantdifference in the amount of stem dieback was observed in spring1994. In spring 1995 and 1996, visual assessments of stem diebackappeared to be uniform across treatments.

Figure 2a–c summarizestotal stem height, stem diameter and basal area measured each Octoberduring the study. By the end of the fifth growing season, mean saplingstem height was 720 cm, diameter was 94 mm andbasal area was 73 cm2. The chronic effects ofO3 and 1·5 × O3 + CO2 were determinedover the 5 year study by comparing the yearly growth incrementsof yellow-poplar for each season using RMA (Fig. 2d–f, Table 2).Yearly height growth increment was not affected by exposure to 1·5 × O3 or1·5 × O3 + CO2. Yearlydiameter growth increment was impacted by the treatments (P = 0·086)but with no significant treatment by year interactions. ElevatedO3 (1·5 × O3)alone had no significant effect on the yearly stem growth diameterincrement but when combined with 700 p.p.m. CO2 (1·5 × O3 + CO2),the yearly diameter growth increment increased 14% relativeto all others (P = 0·021).Yearly basal area growth increment was impacted by the treatments(P = 0·009) and significanttreatment by year interactions were detected (P = 0·004).Contrasts were tested to determine treatment differences withineach growing season. The effects of 1·5 × O3 onyearly basal area growth increment were not detected. The effectsof 1·5 × O3 + CO2 airwere first detected after two seasons of exposure. The yellow-poplar basalarea growth increment was 33, 24, 28 and 41% greater in1·5 × O3 + CO2-airin comparison with all others in 1993, 1994, 1995 and 1996, respectively.Yearly basal area growth increment of 1·5 × O3 + CO2-treeswas 37, 24, 27 and 37% greater than those trees exposedto 1·5 × O3-alonein 1993, 1994, 1995 and 1996, respectively (Table 2).

Figure 2.

(a) Yearly total stemheight (cm); (b) totalstem diameter (mm); (c) totalbasal area (cm2); (d) yearlyheight growth increment (cm); (e) yearlydiameter growth increment (mm); and (f) yearlybasal area growth increment (cm2) of yellow-poplar seedlingsexposed to O3 alone or O3 plus 700 p.p.m.CO2, measured each September from 1992 to 1996. Leastsquare means ± 1SE are represented (n = 3chamber replicates per treatment). Significant within-year treatmenteffects on basal area growth increment were detected in years 2–5(see Table 2).

Table 2.  Summaryof repeated measures analysis and tests of effects on yearly stemgrowth of yellow-poplar to subambient, ambient O3, elevatedO3 alone or combined with elevated CO2 for 5years (1992–96)
Denominator d.f.
Yearly growth 
Diameter incrementHeightincrementBasal area increment 
  1. a Singled.f. contrasts were made across years unless noted otherwise. Year1 = 1992; Year 2 = 1993;Year 3 = 1994; Year 4 = 1995;Year 5 = 1996. Because significant treatmentby year interactions were only detected for basal area growth inyears 2, 3, 4 and 5, single d.f. treatment contrasts are presentedin (c). Refer to Fig. 2 fortreatment means.

(a) Overall effects      
 Treatment 3 60·0860·3930·009 
 Year 431< 0·001< 0·001< 0·001 
 Treatment by year12310·3190·9490·004 
(b) Contrasts across years a      
 CF versus 1·5 × O3 1 60·3390·7700·803 
 CF versus 1×O3 and 1·5 × O3 1 60·1310·6700·306 
 All versus 1·5 × O3 + CO2 1 60·0210·1410·002 
 1·5 × O3 versus1·5 × O3 + CO2 1 60·0990·3510·005 
 Year 3 versus all other years 132< 0·001 
 Year 4 versus all other years 132< 0·001 
 Year 5 versus all other years 132< 0·001 
(c) Basal area increment – within year contrasts  Year 2Year 3Year 4Year 5
 CF versus 1·5 × O3 1320·5670·6840·2780·379
 All versus 1·5 × O3 + CO2 1320·0290·042< 0·001< 0·001
 1·5 × O3 versus1·5 × O3 + CO2 1320·0150·0910·002< 0·001



Exposure to either 1·5 × O3-or 1·5 × O3 + CO2-airdid not significantly impact leaf area, shoot or root biomass after twoseasons of treatment. Sapling biomass averaged 310·3 g DWfor stems, 142·6 g DW for branches, 237·9 g DWfor leaves, 53·9 g DW for lateral rootsand 76·5 g DW for tap roots. Specificleaf area averaged 169 cm2 g−1 andtotal leaf area averaged 3·96 m2 persapling. The mass of upper canopy terminal leaves for saplings exposedto 1·5 × O3 + CO2-airwas 1·25 times greater than all other treated saplings.The ratio of first-order lateral to tap root biomass was 12–20% higherin saplings exposed to 1·5 × O3 + CO2-aircompared with all other treated saplings (P = 0·024).Saplings exposed to 1·5 × O3-airhad 11% higher leaf area ratios (LAR) than CF-exposed saplings, whereasthose grown in 1·5 × O3 + CO2-airhad 6–18% lower LAR compared with all other saplings(P = 0·020) (Table 3). The ratioof total leaf area per sapling to first-order lateral roots was27% lower in saplings exposed to 1·5 × O3 + CO2-aircompared with saplings exposed to 1·5 × O3 alone(P = 0·013).

Table 3.  Theleaf area ratio a and ratio of total leaf areato lateral roots of yellow-poplar saplings exposed to subambientO3, ambient O3, elevated O3 aloneor combined with elevated CO2 after two (1993) and five(1996) seasons
TreatmentLeaf area ratio(cm2 g−1)Leaf area : lateral root ratio(cm2 g−1)
  • a 

    LeafArea Ratio (LAR) = Total leaf area/tree dry mass.

  • Trees were sampled in mid-SeptemberEach value represents the least square mean ±1SE (n = 3 chamber replicates pertreatment: six trees per chamber in 1993; and three trees per chamberin 1996).

 CF47·6 ± 2·3733·5 ± 66·1
 1 × O352·6 ± 2·4845·4 ± 66·9
 1·5 × O353·2 ± 2·2869·9 ± 63·5
 1·5 × O3 + CO244·7 ± 2·3635·1 ± 64·4
 CF28·7 ± 1·1317·5 ± 29·6
 1 × O332·4 ± 1·8342·7 ± 18·9
 1·5 × O329·9 ± 1·8362·8 ± 18·9
 1·5 × O3 + CO221·4 ± 1·6236·7 ± 16·3
Statistical significance (P-value)


After three seasons, no significant treatment effects on leaf areaor above-ground biomass (branches, stem, or leaves) were detected.Stem dry weight averaged 770 g, branch dry weight averaged466 g, leaf dry weight averaged 465 g and totalleaf area per sapling averaged 8 m2. Roots werenot excavated in this harvest.


A 16-fold increase in standing live plant (shoot + root)biomass was observed from 1993 to 1996. After five seasons of exposure,significant treatment effects on shoot and root biomass were detected (Table 4).Both shoot and root biomass were approximately 60% greaterin saplings exposed to 1·5 × O3 + CO2-airin comparison with all other saplings (Fig. 3).Although statistically significant O3 effects on above-and below-ground biomass were not detected, root (first-order laterals,tap and total) biomass was reduced 13% in saplings exposedto 1·5 × O3-air relativeto CF controls. Root biomass (first- and second-order laterals,tap and total root system) of saplings exposed to 1·5 × O3 + CO2-airwas on average 1·7 times greater than saplings exposedto 1·5 × O3-air (P-valuesranged from 0·02 to 0·04). Stem biomass of saplingsexposed to 1·5 × O3 + CO2 increased53 and 60% in comparison with CF- and 1·5 × O3-exposedsaplings, respectively. No treatment effects on leaf and branchbiomass or root : shoot ratios were detected.

Table 4.  Asummary of the tests of significance for the effects of subambientO3, ambient O3, elevated O3 aloneor combined with elevated CO2 on yellow-poplar growthand biomass measured at the final destructive harvest in September1996
Response variableTreatment effectaP-value
  1. n = 3chamber replicates per treatment. Three trees per chamber were harvestedin September 1996.

  2. a Treatmentsincluded in tests of significance for ‘treatment effect’ includedCF, 1  × O3,1·5 × O3 and1·5 × O3 + CO2 plots.Single d.f. contrasts comparing specific treatments, e.g. CF versus1·5 × O3, 1·5 × O3 versus1·5 × O3 + CO2,etc. were used. Please refer to Fig. 3 formore details.

Growth measurements at 1996 harvest
 Total stem height0·181
 Diameter of stem initiated in 19920·109
 Stem diameter at 50 cm from ground0·026
 Basal area at 50 cm from ground0·028
Above-ground biomass
 Stem + branches0·086
Root biomass
 Total (tap + laterals)0·049
 Plant (shoot + root)0·129
 Root : shoot ratio0·505
Figure 3.

Yellow-poplarbiomass (g dry weight) after five seasons (1996) of exposure toO3 and elevated O3 + CO2-air:(a) shoot – leaves, stem and branches, and (b) root – taproot and lateral root. (n = 3chamber replicates per treatment with three trees per chamber).Significant treatment effects on biomass were detected (see Table 4). Single degreesof freedom contrasts comparing 1·5 × O3 versus1·5 × O3 + CO2 treatmentswere significant for the following parameters: stem (P = 0·04),stem + branches (P = 0·07),lateral roots (P = 0·03),tap roots (P = 0·02)and total roots (P = 0·02).

Treatment effects on total leaf area (average 36·4 m2 per sapling)and specific leaf area (average 187·8 cm2 g−1)were not observed. Treatment effects on LAR and the ratio of totalleaf area to lateral root dry weight were similar to those observedin 1993 (Table 3).Ozone (1 × O3 or 1·5 × O3)did not affect LAR, but saplings grown in 1·5 × O3 + CO2-airhad 25–34% lower LAR in comparison with all othersaplings (P = 0·002).The ratio of total leaf area to first-order lateral roots of saplingsexposed to 1·5 × O3 + CO2-airwas reduced by 35% in comparison with those exposed to1·5 × O3-air (P = 0·001).


Stem growth of saplings exposed to ambient O3 within open-topchambers (1×O3) was 21% greater thanthat of saplings grown in open-air chamberless plots (P = 0·06).No differences in diameter growth, plant biomass or spring terminaldieback were detected between the two groups throughout the study.



This study is one of a few to investigate the cumulative effectsof multiple seasons of O3 exposure on tree seedlings asthey mature (some trees were flowering and producing seed) and increasein structural complexity. Over five seasons of exposure, O3 hadno apparent effect on growth or biomass. However, by the fifth season1·5 × O3 saplings exhibitedreductions of 13% in both lateral and tap root mass relativeto CF-air controls even though decreases were not statisticallysignificant. Growth reductions were limited despite declines inseasonal net photosynthesis of 3–13% in ambientair and 5–26% in 1·5 × O3-air(relative to CF-controls) (Rebbeck 1996b; Rebbeck, Scherzer & Loats 1998), indicatingthat O3 was altering carbon fixation. Because apparentO3-induced growth reductions were not initiated untilthe fifth season of exposure, it appears that yellow-poplar mayhave the ability to compensate for O3-induced decreasesin photosynthesis and sustain root and shoot growth for a considerableperiod of time before a cumulative threshold is reached. However,the exact mode of compensation was not apparent. Compensatory responsesassociated with O3 injury, such as increased leaf productionrates, increased photosynthetic rates of younger foliage, increasedleaf size, increased internode and/or branch lengths werenot observed (Tjoelker & Luxmoore 1992; Päkkönen et al.1996). Both lateral (first- and second-order) and taprootsof saplings appeared to be reduced by elevated O3. Itis quite possible that the magnitude of those reductions in rootmass would have amplified had exposures continued. With many woodyspecies, root biomass is often reduced more than shoots by O3 (Laurence et al. 1994). Reducedroot systems due to O3 may make trees more susceptibleto drought and nutrient stress (Yun & Laurence1999) and increase susceptibility to pathogens and insects.The susceptibility to O3 can also be altered by nitrogensupply (Noormets et al. 2001).However, both soil and foliar nitrogen levels were well within thenormal limits for yellow-poplar (Scherzer, Rebbeck& Boerner 1998); and no decreases in foliar chlorophyllwere observed after the first exposure season (Rebbeck et al.1998).

Our previous work with yellow-poplar supports the findings ofthe current study. In a two-season open-top chamber study, the firstseason growth of potted 1-year-old bareroot yellow-poplar seedlingswas stimulated in the presence of elevated O3 (1·5 × ambient,107 p.p.m. × h) despite a 12% reductionin leaf photosynthetic rate (Rebbeck 1996a; Rebbeck & Loats 1997), and after twoseasons, few growth effects were observed despite reductions of21–42% in leaf photosynthetic rates.

We observed no changes in leaf area, shoot or leaf biomass orstem heights attributable to ambient or elevated O3.It seems plausible to suggest that both pot-grown seedlings andfield-grown saplings of yellow-poplar do not immediately alter allocationpatterns in response to chronic exposures to elevated O3 despitesignificant reductions in leaf carbon fixation. Changes in root : shootratios associated with O3 exposures were not observedin our current study. However, the ratio of total leaf area to first-order lateralroots increased for saplings exposed to 1·5 × O3 whenmeasured after the second and fifth seasons. These increases suggestthat more carbon may have been retained in the leaves for increasedrepair of O3-impacted foliage.

Based on the results of our present study, the response of field-grownsaplings exposed to O3 over 5 years appear to be verysimilar to seedling responses observed in shorter-term exposuresunder more controlled conditions (Chappelka et al.1988; Tjoelker & Luxmoore 1992; Cannon et al. 1993; Rebbeck 1996a; Rebbeck& Loats 1997; Loats & Rebbeck 1999). Isebrands et al. (2001) alsoreported similarities in the sensitivity of Populus tremuloides toO3 between those exposed for one season in open-top chambersand those exposed longer-term in a free air carbon exchange (FACE)system. Demonstrating these similarities in O3 responsesbetween seedlings and saplings should greatly enhance the predictivemodelling effort to scale the impacts of O3 from juvenileto mature trees (Kolb & Matyssek 2001).

O3 + CO2 effects

Although direct inferences as to the effects of enriched CO2-airalone on the growth of yellow-poplar cannot be made from this study,we first observed significant increases (∼ 33%)in basal area growth increment in saplings exposed to 1·5 × O3 + CO2 aftertwo seasons of exposure. After five seasons, the basal area growthincrement increased by ∼ 41%, whereasthe stem and root biomass increased by ∼ 60% inthese saplings relative to all others. During the first three seasonsof this study, seasonal net photosynthesis of saplings grown in1·5 × O3 + CO2-airwas 41–80% higher than trees exposed to 1·5 × O3 alone(Rebbeck et al. 1998). Theseresults suggest that elevated CO2 ameliorated the O3-inducedreductions in carbon fixation and growth.

Stem mass and basal area growth increment was enhanced in 1·5 × O3 + CO2-airwithout any detectable changes in total leaf area, suggesting thatalterations in allocation patterns had occurred. We found that theratio of total leaf area to first-order lateral roots decreasedfor those exposed to elevated O3+ elevatedCO2 relative to CF controls, which was opposite to theresponse observed in saplings exposed to elevated O3-alone.This decrease in the ratio of leaf area to first-order lateral rootssuggests that the addition of elevated CO2 to elevatedO3 shifted carbon below-ground. A similar allocationshift to fine roots in saplings exposed to enriched CO2 alonehas been reported for yellow-poplar (Norby et al.1992) and Betula pendula (Rey &Jarvis 1997). We saw no changes in root : shoot ratiosin 1·5 × O3 + CO2-exposedsaplings. Dickson et al. (2001)also reported limited effects of elevated O3 alone or whencombined with elevated CO2 on root : shootratios. Yellow-poplar has a rapidly growing and extensive root system,and during its juvenile stage has a flexible rooting habit thatis not exhibited in most other species (Renshaw &Doolittle 1958; Beck 1990). Furtherstudy is needed to elucidate the long-term root and allocation responsesto the combined effects of elevated CO2 and O3 onfield-grown tree species.

It has been suggested that bud break, leaf phenology, frost hardinessand other developmental processes of woody species may be alteredby exposure to gaseous pollutants, but few reports are available(Mousseau & Saugier 1992; Lee& Jarvis 1995). El Kohen & Mousseau1994) reported later bud burst and earlier bud set in sweetchestnut exposed to elevated CO2, resulting in a shortergrowing season, and Rebbeck (1996a) reportedthat exposure to twice ambient O3 delayed bud burst ofyellow-poplar and black cherry seedlings. Isebrands et al.(2001) attributed an increased shoot dieback of Populustremuloides exposed in elevated CO2 to a continuationof growth into the autumn frost period, but saw no combined effectof elevated CO2 and O3. We did observe a significantincrease in shoot dieback in 1·5 × O3 + CO2-air,but only in the 1992–93 dormant season. It is possiblethat the buds and stem tissue had not adequately hardened. Thateffect however, was short-lived, and no subsequent treatment effectswere detected. We saw no significant change in bud burst associatedwith elevated O3 alone or combined with elevated CO2,in fact, all trees broke bud between 17 and 23 April each spring.

The stimulation of basal area growth and root mass that we observedon yellow-poplar support the hypothesis that enriched CO2 canameliorate the negative effects of elevated O3. Dickson et al. (2001) reportedthat the addition of elevated CO2 (+ 150 p.p.m.above ambient) to elevated O3 (∼ 97 p.p.m. × hcumulative dose) counteracted the negative O3 responseof an O3-tolerant Populus tremuloides clone buthad no effect on an O3-sensitive clone. In a FACE exposurestudy of these same two aspen clones, Noormets et al.(2001) reported similar clonal responses to elevated CO2 aloneor when combined with elevated O3. Isebrands et al.(2001) reported that after three seasons of exposure, Populustremuloides height, diameter and volume increased in elevatedCO2, decreased in elevated O3, and in elevated CO2 + O3 didnot differ from ambient controls. Significant clonal differencesin response were noted. The growth responses of yellow-poplar inour current study were most like that of the O3-tolerantaspen clone (Dickson et al. 2001; Isebrands et al. 2001; Noormets et al. 2001). Althoughothers have also observed stimulations in growth and biomass ofseedlings and young saplings grown in the mixture of elevated O3 andCO2, this has not been consistently observed in all woodyspecies (Barnes et al. 1995; Kull et al. 1996; Lippert et al. 1996; Palomäki et al. 1996; Kellomäki & Wang 1997; Dickson et al. 1998, 2001).Because there is considerable variation in the magnitude of responseof deciduous species to simultaneous exposures to elevated O3 andCO2 (Eamus & Jarvis 1989; Mousseau & Saugier 1992; Ceulemans, Jiang & Shao 1995a, b; Kull et al. 1996; Dickson et al. 1998, 2001; Karnosky et al. 1999; Isebrands et al. 2001), cautionin the interpretation and comparison of studies must be used. Pollutantexposure levels (e.g. how do responses at 500 p.p.m. CO2 comparewith responses at 700 p.p.m. CO2), genetic homogeneityand cultural conditions can vary considerably from study to study.The ability of enriched CO2 to counteract O3-inducedstress is determined by the CO2 concentration, the plantresponse to the CO2 enrichment, and the magnitude ofthe O3 stress (Rudorff et al.2000). The year-to-year variations in responses observedin our current study reinforces the importance of environmentalfactors such as temperature, rainfall and light on modifying treeresponses to pollutants and further demonstrates the need to studythe integrated response of long-lived woody species to gaseous pollutants.


Our current study represents one of the first to examine the responseof field-grown hardwoods to chronic O3 alone or combinedwith elevated CO2 (∼ 700 p.p.m)with minimal irrigation and no fertilization over multiple years.This work addresses the critical need to better understand if seedlings andolder trees express the same susceptibility or tolerance to gaseouspollutants. We have demonstrated similarities in the pollutant responsebetween field-grown yellow-poplar saplings and pot-grown seedlings(exposed either in open-top chambers or controlled environmentalconditions), and that it can take several years before detectablebiomass impacts are expressed, even for a fast-growing species such asyellow-poplar. The chronic O3 and CO2 responseof individual species needs to be well characterized under field conditionsbefore meaningful predictions of the impacts of future climateson forested ecosystems can be made.


This study is part of a co-operative effort of the North-easternForest Experiment Station, North Central Forest Experiment Station,US Department of Agriculture Forest Service and Michigan TechnologicalUniversity, funded by the Northern Global Change Program. We wouldlike to thank Mary Ann Tate, Arthur Peterson, Jonathan Miller, CarolCalvin, LeRoy Edwards, Carol Arny, Mary Ford, Pamela Jacobs andChad Richards for their highly valued technical assistance. We alsooffer our thanks to Dr David Randall, Dr William Retzlaff and DrMichael Simini for constructive comments on an earlier version ofthis manuscript. The use of trade, firm, or corporation names inthis publication is for the information and convenience of the reader.Such use does not constitute an official endorsement by the US Departmentof Agriculture or the Forest Service of any product or service tothe exclusion of others that may be suitable.

Received 20 December 2001;received inrevisedform 31 May 2002;accepted for publication 11 June 2002