CO 2 Concentrations in Forests along a Topographic Gradient

The University of Notre Dame CO2 Concentrations in Forests along a Topographic Gradient Author(s): W. A. Reiners and R. O. Anderson Reviewed work(s): Source: American Midland Naturalist, Vol. 80, No. 1 (Jul., 1968), pp. 111-117 Published by: The University of Notre Dame Stable URL: http://www.jstor.org/stable/2423607 . Accessed: 03/02/2012 14:02 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The University of Notre Dame is collaborating with JSTOR to digitize, preserve and extend access to American Midland Naturalist. http://www.jstor.org C02 Concentrations inForestsalong a Topographic Gradient W. A. REINERSI and R. 0. ANDERSON Department of Botany, Universityof Minnesota, Minneapolis 55455 ABSTRACT: CO2 concentrationsand air temperatureswere measured at seven levels at 2-hr intervals for 24 hrs at stations along a slope leading froman upland oak forestinto a cedar swamp. Skies were clear and the winds light over the period. Although there was evidence for cold air drainage downslope and concentration of CO2 in low sites during the early evening hours, temperaturestended to become isothermal at all stations along the gradient later in the night while CO-2 concentrations were higher near the ground on the slope but vertically uniformin the swamp. There was little evidence that CO2 concentrations were significantlyhigher through most of the night in the swamp than on the upland. INTRODUCTION CO2 concentrations fluctuatein nature because of variationin ratesof photosynthesis and respiration, and changesin microclimatic conditions.Investigationshave shown that CO2 concentrations are generallyhigherin foreststhan in fieldsor grasslands(Fuller, 1948; Wiant, 1964), and that concentrations increasemarkedlynear the ground (Huber, 1952; Mitscherlichet al., 1963; Sparling and Alt 1966; DeSelm, 1952). Near the ground,concentrations may vary spatiallydependingon sunlightpenetrationor proximity to C02-rich sourcessuch as decayinglogs (Wiant, 1964). Biologicaland microclimatic changes associated with the diurnal cycle produce large changes in CO2 concentrations at all levels, but especially near the ground. Tabulated data presented by Mitscherlich et al. (1963) for a 24-hour period showed average maximum changes of 51 ppm at the 1 m level, 32 ppm at 5 to 5.5 m level, and 25 ppm at various heights within tree crowns. The purpose of this studywas to measure topographic influenceon concentrations of CO2. The study was conducted at Cedar Creek Natural History Area, Anoka Co., Minnesota, 48 km north of Minneapolis. The site was on a sandy peninsula nearly surrounded by a peat-filled basin supporting a white cedar (Thu ja occidentalis L.) swamp forest. Upland vegetation on the peninsula was dominated by northernpin oak (Quercus ellipsoidaluisE. J. Hill). A narrow "lagg" (marginal fen) community about 20 m wide was located along the sand-peat intersectionat the foot of the slope. This community w~as dominated by black ash (Fraxinus nigra Marsh. ), American elm (Ulmus americana L.) and speckled alder (Alnus rugosa (Du Roi) Spreng). 1 Present address: Dept. of Biological Sciences, Dartmouth College, Hanover, New Hampshire 03755. ill 112 THE AMERICAN MIDLAND NATURALIST 80(l) The topo!graphic transectextendingfromthe crestof the peninsula to 20 m withinthe cedar swampwas 78 m long and dropped2.1 m in elevation. Althoughthe gradientwas slight,it was obvious from experiencethat on calm evenings,cold air drained fromthe upland to the lagg an'd swamp. Since cold air drains down fromradiating foliage surfa'cesat night,presumablyconsiderableamountsof CO2 resultingfromleaf respiration, move towardsthe forestfloorwiththis drainage. Also, since large amountsof CO2 are released fromthe forestfloor,it would seem that cold air movingdownslopeand collectingin b,asinswould be particularlyenrichedin CO,2. This study was designedto measure the degree of this enrichmentalong the topographicgradient. Acknowledgments.-This studywas supported by N.S.F. grant No. GB-3636 and the Graduate School of the Universityof Minnesota. Special appreciation is due Dr. William H. Marshall, Director of Cedar Creek Natural History Area, for services and use of the area. METHODS Four 6.1 m poles of electricalconduit were located along the topographic gradient. One was located on the peninsula crest (A), the second midway downslope in the oak forest (B), the third in the lagg (C), and the fourthin the cedar swamp (D). The tops of these "towers"' reached into the lower portions of the oak forest and lagg canopies, and about midway into the cedar canopy. Pulleys were arranged at the top of each of these towers so that a thermocouple and '/4-inchID polyvinyl chloride tubing could be raised to desired levels. Every two,hours for a 24-hr period, CO2 concentrations and air temperatureswere measured at each tower at 0.06, 1.0, 2.0, 3.0, 4.0, 5.0, and 6.0 m levels. Temperatures were measured with 24-gage copper-constantan,shielded thermocouples and recorded by a temperature-compensated Leeds and Northrup potentiometer. CO2 concentrations were measured by drawing air through the tubing to a Beckman model 15-A infrared gas analyzer. The analyzer was recalibrated at least every two hours. Temperatures'were accurate to 0.5 C. The accuracy of'CO2 concentrationmeasurementsvaried with concentration. Approximate accuracies are 2 ppm between 280 and 380 ppm, 10 ppm between 381 and 480 ppm, and 20 ppm at concentratiofns greater than 480 ppm. The first2-hr period began at 0900 hr CST, 25 Aug. 1966, and the last period ended at 0847 hr, 26 Aug. Sunset on 25 Aug. was 1903 hr CST, 'suhrise on 26 Aug. was 0528 hr. Skies were clear over the entire period. Breezes up to 11 km/hr developed during daylight hours but dropped to 3-5 km/hr between' 1900 and 0800 hrs. Air speeds were measured by a cup-anemometer at 18.2 m, 4.5 m above the canopy at Cedar Creek weather station, 0.4 km from the study site. No air movement during night hours was registeredby a similar anemometer located at 1.5 m above the ground in an open field. It is probable that inversion conditions occurred over the night hours. A 1968 REINERS 0 10 20 30 AND ANDERSON: 10 20 10 20 30 io T---L I TEMP 113 CONCENTRATIONS AIR TEMPERATURE (?C) 30 10 20 30 10 20 30 6 ------o * _ 2 CO,2 20 30 * | A~~~~~~~~~. 41 2' 4I w oso to loj~~~~~~~~~~~~~~~~~ 30 13 t 70 2 TIME. ( HOURS, CST ) 06 0 / 10 TE MPE RATURE AIR 20 25 AUG. 1966 30 l0 20 30 l0 300 400 500 600 400 20 30 l0 500 300 400 (C ) 20 30 l0 500 600 400 20 30 l0 20 30 50'0 300 400 500 600 IjI 300 I 400 5 00 I 1 |I? 2 2100 2300 25 AUG. i C02 (PPM) / 0100 TIME (THOURS, I1 PM 0300 0500 CST ) 0700 HRS1 26 AUG. 1966 Fig. 1. CO2 and temperature profilesfor four stations (A, B, C, D) for 12 two-hour periods. Values for the temperature curves are given above the graphs, values for COz curves given below the graphs. 114 THE AMERICAN MIDLAND NATURALIST 80(1) pronouncedinversionwas measuredover St. Cloud, Minnesota,at distancefromCedar Creek to 0500 hr., 26 Aug. 1966. Straight-line St. Cloud is approximately 80 km. RES ULTS TEMPERATURE Temperature and CO2 data are plotted together in Fig. 1. Tem- peraturedata are lackingfor the 1100 and 1500 hr periods. Temperatureprofileswere verynearlyisothermalin the 0900 and 1300 hr were slightly periodsand temperatures higher,ca. 1 to 2 C, at C and D comparedwithupslope positionA and B. At 1700 hr,temperatures were 2 to 3? lowernear the groundthan at 1 m at all stationsand distinctly lowerat all levelsthan at 1300 hr. Temperaturesat all levels tended to decrease downslope. The same were measuredat 1900 hr except that temperatures characteristics were stilllower. Althoughtemperatures continuedto decline at all levels at all stationsfor the remainderof the night,differences betweenstations diminisheduntil by 0100 hr air at all stationswas at nearlyequal temperature.Conditionsat all stationswere isothermalbetween0100 and 0700 hr. By 0700 hr temperatures began to,rise rapidly,withthe C and D stationsapparentlywarmingmore quicklythan the A and B stations. C02 CONCENTRATION CO2 concentrations decreasedslightlyin the A, B, and C stations from0900 to 1300hr. Concentrations werenearlyuniformat all levels exceptforslightly highervalues of I0 to 20 ppm between1 m height and 6 cm. The minimumconcentration the entireperiod throughout was 312 ppm at 4, 5, and 6 m heightsat stationA at 1700 hr. At 1500 hr a distinctincreasein CO2 occurrednear the ground in stationsB, C, and D. This trend continued at 1700 hr but included stationA. At 1900 hr CO2 increasedsignificantly at upper levels at all stationswhile it remainedthe same or even decreasedslightlyat levelsnear the ground. CO2 continuedto increaseat nearlyall levelsthroughthe 0500 hr period. In general,concentrations tendedto becomevertically uniform laterin the night,this tendencybeginningfirstin the C and D stations,thenspreadingto B and A stations.By 0500 hr profilesforall fourstationswere reasonablyuniformverticallybut more regularin C and D. The tendencyfor concentrations to be highernear the groundpersistedlongestat the A and B stations.The average CO2 concentrationin ppm at 0500 hr, the period of generallyhighest values, was 439 for A, 469 for B, 460 for C, and 439 for D. The recordedover the entire24-hrperiodwas 584 highestconcentration ppm at 6 cm, stationA, at 2300 hr. At 0700 hr CO2 began to decline at all stationsbut apparently mostrapidlyat C and D. Conicentrations decreasedmore rapidlyat upper levelsthan lower levels. 1968 REINERS AND ANDERSON: CO2 CONCENTRATIONS 115 DISCUSSION As temperatures dropped between 1300 and 1700 hr, and then rose again between 0500 and 0600 hr, it appeared that changes occurred earliest in the lagg and swamp stations (C, D). While this may have been true, it should be recognized that the measurement periods were two hours long and the stations were measured in a downslope order. Since large changes in microclimate can occur in two hours, it should be expected that the changing characteristicwould be more advanced at the end of the period. The ends of periods coincided with the times downslope stations were measured. A lack of absolute synchronymust be considered in interpretingthese results. As temperatures decreased and CO2 concentrations increased between 1300 and 2400 hr, the levels nearest the ground were both lowest in temperature and richest in CO2. This indicated that cold, CO2-rich air was lying on the ground surface and one would expect this denser air to move downslope, creating deep, cold pools of air in the swamp and lagg. To some extent this seems to, be the case. For example, CO2 concentrations were highest near the ground and higher at 1 m level in D than C at 1500 hr, and temperatureswere steadily lower downslope at 1700 hr. That these gradients represented evidence for downslope air movement is disputable because of the problem of non-synchronousmeasurements discussed above. It is quite clear that by 2100 hr temperatureswere no, lower in the lagg and swamp than on the oak-forestedslope. This was true through the rest of the night. Furthermore,temperatureprofileswere essentiallyisothermal,indicating that on this particular night no coldair pools were peculiar to low-lying communities. The fact that A and B CO2 profilescontinued to show higher concentrationsbetween the ground and 2 m for most of the time between 2300 and 0700 hr, may be evidence of continued downslope movement of slightlydenser, C02-rich air. Yet the CO2 profilesin the lagg and swamp are nearly uniformvertically. If downslope movement of CO2-rich air did occur, it would seem that CO2 concentrations at C and D would either be very high near the ground, or, if the pool were fairlydeep, concentrations would be higher throughout the C and D profiles compared with upslope stations. Neither case was true. This situation may have resulted from one of several factors. Diffusion may have accounted for some loss of CO2 stratificationbut could not have been entirelyresponsible. A slight amount of forced convection was recorded over the night and may also have contributed to mixing. Occasional gusts were reported by De Selm (1952) as having disrupted nighttimestratificationin a beech forest. It is difficult to understand, however, why such light breezes as recorded in this study might have affected the lower, denser, swamp forestmore than the upland forest. A third factor may have been the minimization of downslope air movement after early evening hours because of evenly distributedcold air resulting from the deep inversion. Under these conditions, the 116 THE AMERICAN MIDLAND NATURALIST 80(1) moreprolongedconcentration of CO2 nearthegroundat A and B may have developed fromhigherCO2 evolutionrates fromforestfloors in the oak forestthan in the lagg and swamp and may have been unrelatedto downslopemovement.On thebasis of extensivemeasurement of CO2 evolutionfromforestfloorsin this area, however,it would be expectedthat nighttimeCO2 evolutionrates fromthe oak forestfloorwould be less,not greater,than in the lagg and swamp communities. A fourthpossibility is that a deep pool of C02-rich air may have developedin the swamp,but under the meteorological conditionsof the nightbecame so deep that the top of the pool exceededthe highest measuredlevel,6 m. This hypothesis is wveakened by the factthat if such a deep pool of C02-rich air did develop,and if it did have a level top,the top would have intersected the slope of the peninsulaat some well-defined levelsof stationsA and B. No such intersection is obvious. It is possiblethatthe pool mayhave been much deeperthan 6 m, the top of the towerin the swamp,but if so, it would have entirelyoverriddenthe entirepeninsula and not representeda result of the topographicgradient. The fifthhypothesisis that downslopeair movementmay have occurredbut, ratherthan producinga stagnantpool of cold air in the swamp,thismovementproduceda somewhatcyclicpatternof air flow as describedby Van Arsdel (1965, 1967). Accordingto Van Arsdel,undercalm conditions, cold air flowsdownslopeto swampsbut is pushed upward into the swamp canopy by constantlyunderriding air. This risingair thenmaycirculatea shortdistancealoftdepending on the heightof the shallow inversionceiling,sinkingover upslope sites, flowing down to the soil, and then downslope again, completing the cycle. Such circulation patternshave been observed by Van Arsdel with the use of smoke bombs. At present, this last hypothesis may present the best explanation for the fact that although CO2 profilesat C and D were nearly uniformverticallybetween 2300 and 0700 hr, profilesat A and B showed higher concentrations along the ground. Rising air in the lagg and swamp would produce verticallyuniformCO2 profiles. Unfortunately, temperature profilesat A and B do not indicate colder air along the ground, which would be better evidence for downslope movement and support the hypothesisthat cyclic patterns existed. The distribution of CO2 described in this study is incompletely understood but with better instrumentationmight be better explained. Collection points located well above the canopy, and smoke bomb devices would be helpful in understanding this phenomenon. Automated data collecting and recording devices would be extremely useful, allowing replication of these observations over longer periods and diverse conditions. Whatever the explanation, results from this study did not show an extraordinaryincrease in CO2 concentration in low-lying sites compared with upland sites under clear, calm, rnocturnalconditions. Some 1968 REINERS AND ANDERSON: CO2 CONCENTRATIONS 117 at approximately equal processtendedto maintainCO2 concentrations exceptionvaluesalong thetopographicgradient.For theseconditions, ally abundantCO2 does not appear to be an importantenvironmental exceptforthemorerapid factorforthe swampand lagg communities accumulationof CO2 in earlyeveninghoursand hence,longerperiods of timeundera C02-richatmosphere. REFERENCES DE SELM, H. R. 1952. Carbon dioxide gradients in a beech forest in central Ohio. Ohio J. Sci., 52:187-198. FULLER, H. J. 1948. Carbon dioxide concentration of the atmosphere above Illinois forestand grassland. Amer. Midl. Natur., 39: 247-249. HUBER, B. 1952. Der Einflussder Vegetation auf die Schwankungen des C02Gehaltes der Atmosphaire.Arch. fur Meteorol., Geophys. u. Bioklimatol., Ser. B, Allgem. u. Biol. Klimatol., 4:154-167. MITSCHERLICH, G. VON, K. G. KERN UND E. KUNSTLE. 1963. Untersuchungen iuber den Kohlensaiuregehaltder Waldluft in Plenterwald und Fichten reinbestand. Allgem. Forst u. Jagdztg., 134:281-290. SPARLING, J. H. AND M. ALT. 1966. The establishmentof carbon dioxide concentration gradients in Ontario woodlands. Canadian J. Bot., 44: 321-329. VAN ARSDEL, E. P. 1965. Relationships between night breezes and blister rust spread on Lake States white pines. Lake States Forest Exp. Sta. Res. No,te,LS-60. 4 p. . 1967. The nocturnal diffusionand transportof spores. Phytopathology, 57:1221-1229. WIANT, H. C., JR. 1964. The concentrationof carbon dioxide at some forest micro-sites. J. Forestry,62:817-819. SUBMITTED 14 JUNE 1967 ACCEPTED 13 JULY 1967