ENVIRONMENTAL CONTROLS OVER NET EXCHANGES OF CARBON DIOXIDE FROM CONTRASTING FLORIDA ECOSYSTEMS

Ecological Applications, 9(3), 1999, pp. 936–948 q 1999 by the Ecological Society of America

ENVIRONMENTAL CONTROLS OVER NET EXCHANGES OF CARBON DIOXIDE FROM CONTRASTING FLORIDA ECOSYSTEMS KENNETH L. CLARK,1 HENRY L. GHOLZ,1 JOHN B. MONCRIEFF,2 FORD CROPLEY,2 HENRY W. LOESCHER1 2

AND

1School of Forest Resources and Conservation, University of Florida, Gainesville, Florida 32611 USA Institute of Ecology and Resource Management, University of Edinburgh, Darwin Building, Mayfield Road, Edinburgh EH9 3JU UK

Abstract. Net CO2 exchange estimated using eddy covariance and relaxed eddy accumulation indicated that evergreen pine upland and deciduous cypress wetland ecosystems in north-central Florida had similar apparent light compensation points during the growing season (125 vs. 150 mmol PPFD·m22·s21), but that maximum rates at 1800 mmol PPFD·m22·s21 at the cypress ecosystem were only 59% of those at the pine ecosystem (8.9 vs. 15.2 mmol CO2·m22·s21). During both the summer and winter months at the pine ecosystem, net CO2 exchange in the daytime was a curvilinear function of PPFD, with no significant seasonal differences in slope or intercept. In contrast, net CO 2 exchange at the cypress ecosystem was minimal during the daytime in the winter. Net CO 2 exchange during the nighttime was an exponential function of air temperature at both sites, with Q 10 values of 2.0 and 1.9 for the pine and cypress ecosystems, respectively. Lower nighttime fluxes of CO2 occurred at the cypress ecosystem across the entire temperature range. Both of these relatively sparse canopies stored CO2 during stable atmospheric conditions. Mean maximum net CO2 exchange during the daytime and mean nighttime net CO 2 exchange for these ecosystems were highly contrasting, and together resulted in a relatively low rate of annual carbon accumulation in the wetland when compared to the aggrading pine ecosystem. However, values reported here are within the ranges of values for other boreal, temperate, and tropical forest ecosystems. Key words: carbon cycle; cypress wetland; eddy covariance; Florida; net carbon exchange; slash pine plantation.

INTRODUCTION Managed slash pine (Pinus elliottii var. elliottii Englm.) and naturally regenerated cypress (Taxodium ascendens Brongn.) ecosystems are two of the major constituents of the ‘‘pine flatwoods’’ of the lower Coastal Plain in the southeastern United States (Meyers and Ewel 1990, Ewel 1998). Pine flatwoods occupy approximately half of Florida, two-thirds of which are uplands dominated by slash pine, and up to a third of which are shallow depressions and drainages dominated primarily by cypress. Although they occur in the same landscape and climatic regime, these two ecosystems differ considerably. The topographic variation of the region results in contrasting hydrologic regimes (Liu 1996). Standing water occurs in the wetlands throughout much of the year, resulting in saturated soil conditions and the accumulation of peat. Sandy subsoils of the uplands also saturate periodically as the water table fluctuates in response to precipitation, but surface flooding is infrequent and short lived. The phenologies of the dominant tree species in the canopies of these two ecosystems are correspondingly distinct, with evergreen pines occurring in the uplands, and deciduous cypress and swamp tupelo Manuscript received 5 October 1998; accepted 20 November 1998.

(Nyssa sylvatica var. biflora (Walter) Sarg.) occurring in the wetlands. Species composition of the understories also differs markedly. The contrasting hydrology and canopy phenology of these two ecosystems imply fundamental differences in the way that carbon is fixed, stored, and cycled, which must be resolved in order to understand the dynamics of landscape-level carbon balances. Midsummer values of leaf-level net photosynthesis for slash pine (Teskey et al. 1994a) and cypress (Brown 1981, correcting for differences in leaf area expression) are similar, suggesting that canopy phenology is the major factor responsible for any difference in annual net carbon dioxide (CO2) gain between these ecosystems. Differences in net ecosystem CO2 exchange are also a function of autotrophic and heterotrophic respiration (Lavigne et al. 1997, Ryan et al. 1997, Law et al. 1998), with presumably lower heterotrophic respiration rates occurring in wetlands where accumulations of .1 m of peat are common (Ewel and Odum 1984). The dynamics of landscape level carbon balances for pine flatwoods are also a function of forest management practices. The collective differences in the natural features of these two ecosystems has led to the development of distinct management regimes. Slash pine stands are managed primarily on an even-aged basis

936

CO2 EXCHANGE IN FLORIDA FORESTS

August 1999

TABLE 1. Structural characteristics for the slash pine and cypress ecosystems in 1996. Pine ecosystem

Parameter Stem density (stems/ha) Tree height (m) Stem diameter at 1.3 m dbh (cm) Stand basal area (m2/ha)

1301 19.2 17.4 31.4

6 6 6 6

81 1.0 0.2 1.4

Cypress ecosystem 2563 26.5 15.0 64.0

6 6 6 6

297 2.0 0.4 6.9

Note: Sample sizes are four plots (625 m2/plot) for the slash pine ecosystem, and three plots (528 m2/plot) for the cypress ecosystem.

for pulpwood production on rotation lengths of 20–25 yr. Clearcutting is followed by intensive site preparation and planting of pine seedlings. Fertilizer is usually applied at least once (during planting), and herbicides are increasingly being used to reduce competition during the early phase of stand development. Managed slash pine stands now dominate the uplands of the pine flatwoods in Florida, comprising 1.9 Mha (Brown 1996). Although also typically clearcut for stemwood, cypress-dominated wetlands are harvested on longer rotations and with no site preparation. The canopy regenerates naturally, primarily by resprouting, and no fertilizers or herbicides are used. Previous estimates from models that scale cuvette and small-chamber measurements up to the ecosystem level indicate that annual net CO2 exchanges of cypress wetlands (Brown 1981) are much smaller in magnitude than those of the pine uplands (Cropper and Gholz 1993). Direct measurements of net CO2 exchange have not been made, but have recently become possible using micrometeorological techniques (e.g., Wofsy et al. 1993, Baldocchi et al. 1996, Beverland et al. 1996, 1997, Moncrieff et al. 1997). As an important first step in understanding the dynamics of landscape-level carbon balances, we made measurements using these techniques over a mature, rotation-aged slash pine plantation and an adjacent, mature, unmanaged cypress wetland. The objectives of this research were to (1) estimate net exchange of CO2 from these two ecosystems over a wide range of meteorological and phenological conditions, (2) elucidate the importance of major environmental and physiological controls over net exchange rates, and (3) estimate annual net exchanges of CO2 from these two ecosystems using relatively simple models, and compare them to biomass and litter accumulation estimates. METHODS

AND

MATERIALS

Study sites Two study sites were established in 1994, 15 km northeast of Gainesville, Alachua County, Florida (298449 N, 82899300 W). Long-term (1955–1995) mean January and July temperatures were 148C and 278C, respectively (NOAA 1996). Mean monthly minimum and maximum temperatures throughout the study pe-

937

riod (July 1994 to July 1997) ranged from 5.98 to 19.58C and 21.58 to 32.98C for these two months, respectively. Long-term mean annual rainfall for Gainesville was 1332 mm, and annual precipitation in 1996 was 1227 mm. The pine ecosystem was dominated by a 24-yr-old (in 1996) plantation of Pinus elliottii var. elliottii established on industry land and managed for pulpwood production. Soils of the pine site are ultic alaquods (sandy, siliceous, thermic) that are poorly drained and low in organic matter and available nutrients. The distributions of discontinuous subsurface spodic (organic) and argillic (clay) horizons range between 30–70 cm and 100–200 cm depth, respectively (Gaston et al. 1990). The water table fluctuated from at the surface to .1.5 m depth during the measurement period. Mixed genotype seedlings were planted at harvest density following stem-only harvest of the previous stand, chopping and broadcast burning of residues, and bedding. The stand had not been thinned or fertilized since establishment. Tree densities, heights, and stem diameters were measured on four 25 3 25 m plots (Table 1). Understory vegetation consisted of native species reestablished naturally after site preparation, primarily Serenoa repens (W. Bartram) Small, Ilex glabra (L.) A. Gray, and Myrica cerifera L. The planting block used for this study was 100 ha in area with a minimum fetch from the tower of ;800 m (see Plate 1), and it was surrounded by other slash pine plantations between 9 and 20 yr in age. Tree inventories and measurements of diameter at 1.3 m (dbh, in centimeters) and height (in meters) were conducted annually at the pine site (Table 1). Tree biomass and growth increments were estimated from allometric relationships based on dbh (Gholz et al. 1991; Table 2). Understory biomass was estimated from census data in each plot using allometric relationships based on various TABLE 2. Distribution of carbon and annual carbon accumulation rates for the slash pine ecosystem in 1996; data are means 6 1 SD. Component

Pinus elliottii Branches 1 twigs Stems Coarse roots Total wood Understory Foliage Total live wood Forest floor Total

Biomass (g C/m2) 537.1 6206.4 1336.4 8079.9

6 6 6 6

40.5 315.9 68.0 423.2

40.4 6 30.5 150.4 6 95.4 2596.0 6 651.8 10 866.7

Increment (g C·m22·yr21) 39.3 303.1 65.7 410.1

6 6 6 6

2.6 11.8 2.5 16.8

0† 0† 335.0 6 126.2 745.1

Note: Carbon in components was estimated using allometric biomass relationships based on dbh for trees (Gholz et al. 1991) and various plant dimensions for understory species (Gholz et al, personal observations), with organic matter assumed to be 50% carbon. Carbon accumulation on the forest floor was estimated from litterfall measurements, assuming a 15%/yr mass loss (Gholz et al. 1985). † Components initially assumed to be in steady state.

KENNETH L. CLARK ET AL.

938

Ecological Applications Vol. 9, No. 3

TABLE 3. Annual fine litterfall (g C·m22·yr21 6 1 SD) for the slash pine and cypress ecosystems. Sample sizes are 10 litterfall traps (1 m2) for each ecosystem. Pine ecosystem Litterfall component Needles Leaves Subtotal of foliage Nonfoliar Total

1995 279.7 8.4 288.1 93.6 381.7

6 6 6 6 6

38.1 11.4 35.1 43.0 68.7

Cypress ecosystem 1996 262.3 8.7 271.0 123.6 394.7

6 6 6 6 6

32.6 14.1 33.1 106.3 126.2

1996 162.6 105.3 268.0 57.8 331.2

6 6 6 6 6

22.6 11.7 28. 20.3 39.2

plant dimensions (H. L. Gholz et al., personal observations). Fine litterfall was collected every two weeks from 10 1 3 1 m traps at random locations in two of the four measurement plots (Table 3). Annual needle litterfall in similar nearby pine stands ranged from 422 to 589 g·m22·yr21, with corresponding maximum and minimum seasonal LAI of 6.5 and 3.7 (Gholz et al. 1991). Forest floor mass was sampled in plots throughout the duration of the study, and coarse woody debris was sampled in each plot at the end of the study. The cypress wetland ecosystem was dominated by an uneven-aged (maximum estimated tree age is ;130 yr; Brown 1981), naturally regenerated stand of Taxodium ascendens in a nearly circular 12.8-ha swamp (see Plate 2). Soils are spodic psammaquents overlain by a thick (1–2 m) organic surface layer of peat. Seasonal fluctuations in water level were ,0.7 m over a typical year, with open water in the center of the pond

ranging from 0.5 to 1.2 m deep during the study. The subcanopy also included numerous individual trees of Nyssa sylvatica var. biflora as well as slash pine (Table 1). Understory shrubs were primarily Leucothoe racemosa (L.) A. Gray, Myrica cerifera L., Vaccinium spp., Ilex spp., Woodwardia virginica (L.) Sm., and Osmunda spp. Sphagnum spp. and other bryophytes occurred at the base of trees and on the water surface. Minimum fetch from the tower at this ecosystem was 200 m, and it was surrounded by variously aged slash pine plantations to the east and west, and similar cypress-dominated wetlands to the north and south. Tree inventories and measurements of dbh and height were conducted once at the cypress site (Table 1). Litterfall was collected every two weeks from 10 1 3 1 m traps at random locations near the meteorological tower (Table 3). Liu et al. (1997) described seasonal LAI for a nearby cypress wetland, and reported max-

PLATE 1. The slash pine canopy and tower (note person mid-way up) at minimum LAI (winter 1996).

PLATE 2. The deciduous cypress canopy and tower in midwinter.

CO2 EXCHANGE IN FLORIDA FORESTS

August 1999

imum and minimum seasonal values of 5.0 and 1.5, respectively. Maximum seasonal LAI based on allometric relationships at this ecosystem was previously estimated at 5.9 (Brown 1981). This ecosystem was used extensively as a control site in previous wetland studies (Ewel and Odum 1984).

939

Most (.90%) net ecosystem fluxes of CO2 were estimated using the eddy covariance technique. The vertical eddy flux of CO2 at a fixed plane above the canopy can be estimated as:

Because we did not measure CO2 concentrations at different heights through the canopy, half-hourly changes in CO2 concentration at the height of the sampling inlet were used to estimate FDS in the volume of air beneath the inlet (24 m3 for the pine ecosystem or 31 m3 for the cypress ecosystem) (e.g., Hollinger et al. 1994). For the analyses presented here, only daytime and nighttime data collected during periods when the canopy was well coupled with the atmosphere were used. We defined well-coupled periods as those where the friction velocity, u*, was greater than 0.2 m/s (e.g., Goulden et al. 1996, 1997).

Feddy 5 r9c w9

Relaxed eddy accumulation measurements

Eddy covariance measurements

(1)

where Feddy is the eddy flux of CO2, r9c 5 rc 2 rc, the instantaneous deviation from the mean density (or concentration) of CO2, and w9 5 w 2 w, the instantaneous deviation of the vertical wind speed from the mean vertical wind speed. To estimate the net ecosystem flux of CO2 between these forests and the atmosphere, the flux associated with the rate of change of CO2 storage in the air column below the eddy covariance system is also included:

FCO2 5 Feddy 1 FDS

(2)

where FCO2 5 net ecosystem flux of CO2, and FDS is the flux associated with the change in storage of CO2. Net eddy fluxes of CO2 were measured using a closed-path eddy covariance system (EdiSol; Moncrieff et al. 1997). The system was composed of (1) a three-dimensional sonic anemometer (A1002R, Gill Instruments Limited, Lymington, UK) mounted at the top of a 24-m meteorological tower at the pine ecosystem or a 31-m tower at the cypress ecosystem, (2) a fastresponse, closed-path infrared gas analyzer (LI-6262, LI-COR Incorporated, Lincoln, Nebraska), (3) a 30 m long, 0.4 cm internal diameter (ID) nylon tube (Phase Separations, UK) and a small air pump, and (4) a laptop PC running EdiSol software. The inlet of the tube was placed between the upper and lower sensors of the sonic anemometer, and air was drawn through the LI6262 at a rate of 6.0 L/min. A 3-m copper tube was attached to the outlet end of the nylon tube in front of the LI-6262 to remove temperature fluctuations. The mean lag time from the tube inlet on the tower to the LI-6262 at the base was 7.5 s. The EdiSol software carries out coordinate rotation of the raw sonic anemometer signals to obtain turbulence statistics perpendicular to the local streamline. The maximum values for the covariances between turbulence and CO2 concentrations were compared to a 200-s running mean to calculate instantaneous fluxes. Average net CO2 exchange was then calculated at half-hour intervals. Finally, fluxes were corrected for the nonideal frequency response of the LI-6262, sensor separation loss, and the frequency attenuation of the gas concentration down the sampling tube using transfer functions (Moncrieff et al. 1997).

Fewer measurements of net ecosystem exchanges of CO2 were made using relaxed eddy accumulation (REA; Beverland et al. 1996, 1997) at both sites during summer and fall 1996, when the EdiSol system was not available. The system we used was identical to one previously evaluated in side-by-side comparisons with the EdiSol system in Scotland, where it was demonstrated that half-hourly net CO2 exchanges were not significantly different (Beverland et al. 1997). The system was composed of (1) a one-dimensional sonic anemometer (No. 27A, Campbell Scientific, Logan, Utah), (2) two two-way fast-response solenoid valves (Clippard Minimatic, Clippard Engineering, Cincinnati, Ohio), (3) a 213 datalogger (Campbell Scientific), (4) a comparator and valve driver circuit, (5) a closedpath infrared gas analyzer (LI-6262), (6) two 30 m long, 0.4 cm ID teflon tubes, and (7) two small air pumps. The comparator was used to compare instantaneous vertical wind speed signals from the sonic anemometer to the mean vertical wind speed approximated by a 200-s running average calculated by the 213 datalogger. The 213 was programmed to open the fastresponse valves to sample air from upward- and downward-moving eddies into the separate gas lines. Air flow was controlled with rotometers, and air samples were accumulated in separate polypropylene bags. Air samples were analyzed for CO2 with the LI-6262 on absolute mode by continuously switching between bags at 3-min intervals using two three-way valves, also controlled by the datalogger. Net CO2 fluxes were then calculated at half-hour intervals by

Feddy 5 bswrair(Cup 2 Cdown)

(3)

where Feddy is the net eddy flux of CO2 (mmol CO2·m22·s21), b is a dimensionless coefficient for which we assumed a value of 0.56 (Beverland et al. 1996, Katul et al. 1996), sw is the standard deviation of the vertical windspeed (m/s), rair is the mean air density (in moles per cubic meter), and (Cup 2 Cdown) is the difference in measured CO2 concentrations (micromoles of CO2 per mole) averaged over the half hour. We used the same method to calculate FDS and Feco as with the eddy covariance system. Only data that were collected during periods when the canopy was well

Ecological Applications Vol. 9, No. 3

KENNETH L. CLARK ET AL.

940

coupled with the atmosphere (e.g., u* . 0.2 m/s) were used for the analyses here.

Meteorological and soil moisture measurements Continuous meteorological measurements were made only at the pine ecosystem, with the exception of photosynthetically active photon flux density (PPFD) measurements which were made at both ecosystems. Incoming shortwave radiation (LI-200, LiCor, Incorporated, Lincoln, Nebraska), PPFD (LI-190, LI-COR, Incorporated), net radiation (No. Q7, Radiation and Energy Balance Systems, Incorporated, Seattle, Washington), air temperature and relative humidity (No. ES-110, Omnidata, Incorporated, Ogden, Utah; mounted in a Stevens enclosure), windspeed and direction (No. 12-002, R. M. Young Company, Traverse City, Michigan), and precipitation (TI-525, Texas Instruments, Incorporated, Dallas, Texas) were measured at 22–23 m on the tower at the pine ecosystem. Soil temperature was measured at 5 cm depth (No. ES-060, Omnidata, Incorporated). Meteorological data were recorded with automated data loggers (Easy Logger No. EL824-GP, Omnidata, Incorporated). Water table depth was measured with a Stevens water depth gauge (No. F-68, Leupold and Stevens, Incorporated, Beaverton, Oregon). Time domain refractometry and gravimetric soil moisture measurements were made at regular intervals throughout the sampling period. The average density of air (ra) was calculated from simultaneous measurements of air temperature, relative humidity, and barometric pressure. Barometric pressure data were obtained from the Gainesville Regional Airport, ;6 km southwest of the site. To estimate mean daytime net CO2 exchanges from the pine and cypress ecosystems, we developed nonlinear regression equations to predict net CO2 exchange using continuous meteorological data. Following Ruimy et al. (1995), we fit a rectangular hyperbola to the relationship between PPFD and FCO2:

FCO2 5

a PPFD Fsat 2R a PPFD 1 Fsat

(4)

where a is the apparent quantum yield (dFCO2/dPPFD at PPFD 5 0), Fsat is the net CO2 exchange at light saturation, and R is the mean net CO2 exchange at PPFD 5 0. To estimate mean nighttime net CO2 exchanges from the two ecosystems, half-hourly nighttime net CO2 exchange rates were regressed on air and soil temperature using an exponential function with the form

FCO2 5 aebT

(5)

where a and b are regression coefficients, and T is the mean half-hourly air or soil temperature. SigmaPlot 4.0 Regression Wizard software (SPSS Incorporated, Chicago, Illinois) was used to estimate parameters in Eqs. 4 and 5. We used the convention that positive net CO2 exchange was from the atmosphere to the ecosystem,

FIG. 1. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem as a function of photosynthetically active photon flux density (PPFD, mmol·m22·s21) during the daytime under well-coupled conditions (u* $ 0.2 m/s) for warm (April to October, ●) and cool (November to March, ▫) seasons.

and negative net CO2 exchange (ecosystem respiration) was from the ecosystem to the atmosphere. RESULTS

Net CO2 exchange at the pine ecosystem Measurements made at the slash pine ecosystem during the daytime (with u* . 0.2 m/s) from May to October, when ambient air temperatures were warm (178– 348C) and LAI was relatively high, indicate that net CO2 exchange data fit a curvilinear function of PPFD with a 5 0.044 6 0.009 (mean 6 1 SE), Fsat 5 26.537 6 1.329, and R 5 4.616 6 1.163 (r 2 5 0.70, n 5 218, P , 0.05; Fig. 1, solid symbols). The apparent ‘‘compensation point’’ for net CO2 exchange occurred at a PPFD of ;125 mmol·m22·s21, and mean net CO2 exchange at 1800 mmol PPFD·m22·s21 was 15.2 mmol CO2·m22·s21. At higher PPFD levels, net CO2 exchange decreased slightly with increasing PPFD. Measurements made during well-mixed daytime conditions from November to March, when air temperatures were relatively cool and LAI relatively low, indicate that net CO2 exchange was a similar function of PPFD with a 5 0.038 6 0.005, Fsat 5 29.280 6 1.809, and R 5 3.044 6 0.552 (r2 5 0.78, n 5 353, P , 0.05; Fig. 1, open symbols). During this period, the apparent light compensation point for net CO2 exchange was similar to that of warm periods (;90 mmol·m22·s21), but a net CO2 exchange rate of 15.2 mmol CO2·m22·s21 occurred at only 1300 mmol PPFD·m22·s21. However, neither the slopes nor intercepts of the regression lines of net CO 2 exchange on PPFD for these two periods were significantly different.

August 1999

CO2 EXCHANGE IN FLORIDA FORESTS

FIG. 2. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem as a function of vapor pressure deficit of the atmosphere (VPD, kPa) during the daytime when PPFD . 900 mmol·m22·s21 and air temperature . 258C.

The plot of net CO2 exchange as a function of vapor pressure deficit of the atmosphere (VPD) at PPFD . 900 mmol·m22·s21 and air temperature .258C indicates only a weak relationship between these variables (r 2 5 0.16), suggesting that a decrease in stomatal conductance large enough to affect net CO2 exchange rates occurred only at the highest VPD values (.2.0 kPa) (Fig. 2). Highest VPD values occurred primarily during the late afternoon in April and May before the onset of summer thunderstorm activity, and when PPFD levels were relatively low (points in Fig. 2 at VPD . 2 kPa all occurred with PPFD , 1500 mmol·m22·s21). However, when the same data were plotted as a function of air temperature, a linear relationship between air temperature and net CO2 exchange was apparent (r2 5 0.23, n 5 114, P , 0.05) (Fig. 3). The lower net CO2 exchange values typically occurred at the highest PPFD levels (points in Fig. 1 at PPFD . 1500 mmol·m22·s21), because high temperatures were coincident with high PPFD levels. These results together suggest that high ambient temperatures during the early afternoon increased respiration rates, resulting in lower rates of net CO2 exchange, rather than stomatal closure due to high VPD levels. Soil moisture levels and water table depths fluctuated considerably throughout the measurement period, and were apparently unrelated to net CO2 exchange rates. When the canopy was well coupled to the atmosphere, nighttime net CO2 exchange at the pine ecosystem was significantly related to air temperature (FCO2 5 1.509 exp(0.058T), r2 5 0.36, n 5 477, P , 0.05; Fig. 4). Above-canopy flux measurements indicate a Q10 of ;2.0 when calculated using air temperature, and a Q10 of ;2.3 when calculated using soil

941

FIG. 3. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem as a function of air temperature (.258C) during the daytime when PPFD . 900 mmol·m22·s21.

temperature at 5 cm depth (data not shown). During stable evening and early-morning hours, however, halfhour net CO2 exchange rates were a strong function of increased turbulence. This was likely due to ventilation of air below the canopy to the atmosphere and the release of CO2 that had been stored under stable nighttime conditions, despite the fact that the canopy had a low LAI relative to other forest ecosystems. For example, patterns of half-hourly net CO2 exchange mea-

FIG. 4. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem as a function of air temperature (8C) during the nighttime under well-coupled conditions (u* $ 0.2 m/s).

942

KENNETH L. CLARK ET AL.

FIG. 5. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem on 28–29 September 1995, during well-coupled conditions (u* $ 0.2 m/s).

sured during a windy night and morning (28–29 September 1995; Fig. 5) contrast strongly with those measured during a cloudy afternoon and night with light winds (6–7 October 1995; Fig. 6). During the transition from stable to well-coupled conditions in the morning, the highest half-hour net exchange rate reached approximately 220 mmol CO2·m22·s21, but such rates were sustained for only half an hour, after which positive net CO2 exchange typically occurred. Using the u* threshold of 0.2 m/s, data collected under these conditions were filtered out and not used in the regression analyses. Inclusion of the FDS term accounted for the relatively small changes in CO2 storage by the canopy during more turbulent conditions (e.g., Fig. 5).

Ecological Applications Vol. 9, No. 3

FIG. 6. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the slash pine ecosystem on 6–7 October 1995, during cloudy, poorly coupled conditions (u* variable, but ,0.2 m/s from 0200 to 0800).

5.062 6 7.889, R 5 0.0315 6 0.380; r 2 5 0.18, n 5 92, P , 0.05) (Fig. 7; open symbols). During well-coupled conditions in the nighttime, net CO2 exchange at the cypress ecosystem was also largely controlled by temperature during both periods (FCO2 5 0.938 exp(0.065T), r 2 5 0.45, n 5 179, P , 0.05; Fig. 8). Mean net CO2 exchange rates were 62% and 72% of those measured at the pine ecosystem at

Net CO2 exchange at the cypress ecosystem Daytime measurements made at the cypress ecosystem when the canopy was well coupled to the atmosphere from June to October, and when ambient air temperatures were warm (17–338C) and LAI maximum, indicated that net CO2 exchange of the wetland also fit a cuvilinear function of PPFD with a 5 0.027 6 0.007, Fsat 5 16.039 6 1.358, and R 5 3.148 6 0.832 (r 2 5 0.64, n 5 144, P , 0.05; Fig. 7, solid symbols). The apparent compensation point for net CO2 exchange occurred at a PPFD of ;150 mmol·m22·s21. In contrast to the pine ecosystem, the mean maximum rate of net CO2 exchange at 1800 mmol·m22·s21 was 8.9 mmol CO2·m22·s21, only 59% of that measured under similar conditions over the pine canopy. Daytime measurements made during the winter months, when the overstory was leafless, indicate little relationship between PPFD and net CO2 exchange rates, with rates not significantly different from zero over a wide range of environmental conditions (a 5 0.002 6 0.002, Fsat 5

FIG. 7. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the cypress wetland ecosystem as a function of photosynthetically active photon flux density (PPFD, mmol·m22·s21) during the daytime under well-coupled conditions (u* $ 0.2 m/s) for growing (June to October, ●) and dormant (November to January, ▫) seasons.

CO2 EXCHANGE IN FLORIDA FORESTS

August 1999

FIG. 8. Net exchange of carbon dioxide (mmol CO2·m22·s21) above the cypress wetland ecosystem as a function of air temperature (8C) during the nighttime under wellcoupled conditions (u* $ 0.2 m/s).

ambient air temperatures of 08C and 208C, respectively. The Q10 calculated as a function of air temperature from these data was 1.9. DISCUSSION Two methodological limitations potentially affected our results. First, we did not run the EdiSol and the REA systems in a side-by-side comparison. However, we could not detect significant differences in half-hourly net CO2 exchange values between systems under similar PPFD and air temperatures at the slash pine site. Also, side-by-side comparisons have been made in Scotland, where no statistically significant differences could be detected for either sensible heat or CO 2 fluxes (Beverland et al. 1996, 1997). Second, the fetch is relatively limited at the cypress site when compared to the pine site and to other sites used in eddy covariance studies. Although this is the case, the marked differences in the patterns of net CO2 exchange rates between the evergreen pine and deciduous cypress, especially during the daytime in the winter (Figs. 1 and 7), strongly suggest that a large portion of the flux signal originated within the wetland.

Net CO2 exchange at the pine ecosystem The relationship between PPFD and net CO2 exchange at the pine ecosystem was a similar function of PPFD during both summer and winter months, and although LAI at the pine ecosystem during the winter months was only 60–70% of that in the warmer months (Gholz et al. 1991), daytime net CO2 exchange rates were also similar for both periods. The nearly linear increase in net CO2 exchange up to a PPFD of ;1000 mmol·m22·s21 indicates a progressive light saturation of

943

net CO2 assimilation as radiation is intercepted at progressively lower depths by foliage in the canopy and understory, similar to patterns observed elsewhere (Ruimy et al. 1995, citations in Table 5). Apparently the moderately lower midday air temperatures during the winter had relatively little effect on net CO2 exchange rates, although Teskey et al. (1994a) reported that leaf-level net CO2 assimilation rates for slash pine increased by 33% as ambient air temperature increased from 158 to 258C. It is likely that the similarity in net CO2 exchange rates between summer and winter periods resulted from compensatory temperature-dependent rates of respiration. During the summer months, high ambient air temperatures associated with PPFD . 1500 mmol·m22·s21 likely increased both autotrophic and heterotrophic respiration rates, leading to a slight decrease in net CO 2 exchange rates with increasing PPFD levels (Figs. 1 and 3). Relatively low respiration rates occurred in the winter, particularly heterotrophic and root respiration in the soil, because winter soil temperatures at 5 cm depth were ;78–108C lower than those measured during the summer. These results support the importance of ambient temperatures in regulating seasonal changes in net CO2 exchange through nonlinear effects on respiration rates (Lavigne et al. 1997, Ryan et al. 1997, Law et al. 1998). The pattern of net CO2 exchange as a function of PPFD at the pine ecosystem is similar to the shape of the light response curve estimated from cuvette data using foliage of slash pine (Teskey et al. 1994a) and understory foliage (H. L. Gholz et al., personal observations). Approximately 70% of leaf-level net CO 2 assimilation was predicted by PPFD in both of those data sets, similar to our results obtained for net CO2 exchange at the ecosystem level. However, the lack of increased net CO2 assimilation at PPFD . 1500 mmol·m22·s21 at the leaf level is primarily due to light saturation of photosynthesis, while ‘‘light saturation’’ at the ecosystem level is also apparently due to increased respiration rates at the higher temperatures. Vapor pressure deficit of the atmosphere had relatively little effect on leaf-level net CO2 assimilation of slash pine and understory shrub foliage at PPFDs . 900 mmol·m22·s21 until values reached ;2.5 kPa (Teskey et al. 1994a; H. L. Gholz et al., personal observations), again similar to the results reported here. The fact that this relationship is largely borne out at the ecosystem level indicates that cuvette data can be effectively ‘‘scaled up’’ to estimate net CO2 exchange for this ecosystem using standard meteorological data. Parameters such as average VPD and stomatal conductance that are difficult to obtain at the level of the whole canopy may not need to be modeled explicitly. If such relationships for pine stands are borne out across the lower Coastal Plain landscape, then regional CO2 exchange models for these ecosystems could be

Ecological Applications Vol. 9, No. 3

KENNETH L. CLARK ET AL.

944

Ewel and Odum 1984). Low rates of net CO2 exchange during the daytime in the winter months were clearly a result of the leafless overstory.

Comparison of annual net CO2 exchange with ground-based measurements of carbon accumulation

FIG. 9. Estimated mean daily net exchange of carbon (g C·m22·d21 6 1 SE), by month, at the slash pine and cypress ecosystems in 1996.

relatively simple yet accurate, and driven with only standard meteorological data.

Net CO2 exchange at the cypress ecosystem Measurements made in summer at the cypress ecosystem indicate much lower net CO2 exchange rates for an equivalent PPFD when compared to the pine ecosystem. Leaf area index of the cypress ecosystem is somewhat lower, and autotrophic respiration rates were likely larger due to greater biomass (Table 1). When compared to the pine ecosystem, relatively lower net CO2 exchange rates at night for an equivalent ambient air temperature were probably due to the relatively low decomposition rate of the more recalcitrant litter fractions and peat in the wetland, despite the fact that autotrophic respiration rates may be higher (Brown 1981,

Using Eqs. 4 and 5 and the continuous meteorological data collected in 1996 and 1997, annual net CO 2 accumulation at the pine ecosystem was estimated to be 740 and 608 g C·m22·yr21, respectively. Devolving these net values, total daytime net CO2 gains were estimated to be 1529 and 1416 g C·m22·yr21, and nighttime releases were 789 and 808 g C·m22·yr21 in 1996 and 1997, respectively. The lower annual net CO2 exchange in 1997 was primarily due to increased daytime cloudiness, rather than to higher nighttime air temperatures and respiration rates. In 1996, the highest estimated mean daily rate of net CO2 exchange at the pine ecosystem occurred in April (3.0 6 0.2 g C·m22·d21, mean 6 1 SE), and the lowest in October (0.7 6 0.3 g C·m22·d21) (Fig. 9). Sensitivity of predicted annual CO2 exchange to values of the various parameters in Eqs. 4 and 5 was investigated by altering values separately by 65%, 610%, and 61 SE of the mean value. Using 61 SE of the mean, predicted annual net CO2 exchange for the pine ecosystem in 1996 was most sensitive to values of parameters a and R in Eq. 4, and least sensitive to values of parameters a and b in Eq. 5 (Table 4). Overall, using 610% of the mean values resulted in maximum deviations of 218.3% to 116.9% for all parameters, within 61 SD of the error in litterfall measurements (632.0% of mean litterfall for 1996, Table 3). Similarly, the sensitivity of predicted annual CO 2 exchange to errors in the measurement of meteorological driving variables was investigated by altering daytime PPFD values by 625, 650, and 6100 mmol·m22·s21, and nighttime air temperatures by 60.5, 61.0, and 62.08C. Values of 625 mmol PPFD·m22·s21

TABLE 4. Sensitivity analysis of (A) model parameters and (B) meteorological variables used to estimate annual net CO2 exchange at the pine ecosystem for 1996 (740 g C·m22·yr21). All data are deviations from predicted annual net CO2 exchange (%). Deviations from mean of model parameter Parameter A) Model parameters Daytime net CO2 exchange (Eq. 4) a Fsat R Nighttime net CO2 exchange (Eq. 5) a b B) Meteorological variables Nighttime air temperature (8C) Daytime PPFD (mmol m22 s21)

21

SE

210%

25%

15%

110%

11

230.9 28.9 130.0

214.3 218.3 13.0

27.0 28.9 11.5

16.7 18.6 21.5

113.0 116.9 23.0

125.1 18.7 230.0

17.6 17.3

110.7 110.8

15.4 15.6

25.3 25.9

210.6 212.2

27.5 27.8

21.08 16.0 250 220.5

20.58 13.1 225 29.9

10.58 23.1 125 19.4

11.08 26.1 150 118.3

22.08 111.6 2100 243.8

SE

12.08 213.0 1100 134.6

August 1999

CO2 EXCHANGE IN FLORIDA FORESTS

resulted in a deviation of ,10% of predicted annual net CO2 exchange, and values of 618C for nighttime air temperatures resulted in deviations of approximately 66% (Table 4), well within the error in litterfall measurements. Approximately 55% (410 g C·m22·yr21) of the annual net CO2 exchange for 1996 could be accounted for as stem, branch, and coarse root (total wood) increment (Table 2). Estimated annual net CO2 exchange for that year was close to the sum of total wood increment and the net accumulation of fine litter (745 g C·m22·yr21, assuming a mass loss during decomposition of 15%/yr for annual fine litterfall; Gholz et al. [1985], and that soil carbon was constant over time). Accumulation of carbon in stem biomass is of critical interest to forest managers, as this represents the economic potential of plantation ecosystems. Annual accumulation rates are highly dependent upon stand age in even-aged stands, as well as on nutrient status. Maximum average stem carbon accumulation rates of fertilized slash pine plantations in the flatwoods are reported to approach 500 g C·m22·yr21 (averaged over 0–15 yr in a fertilized plantation; McCrady and Jokela 1998), so that maximum total annual aboveground carbon accumulation may exceed 1000 g C·m22·yr21 on these sites. Although the pine ecosystem used in this study was near rotation age, carbon accumulation in stems was still substantial (303 g C·m22·yr21; Table 2). This may indicate that environmental conditions during much of 1996 were optimal for growth (e.g., frequent clear-sky conditions and moderate temperatures), in spite of the below-average annual precipitation. Using a similar procedure to that for the pine ecosystem, annual net CO2 exchange for the cypress ecosystem was estimated at 84 and 37 g C·m22·yr21 in 1996 and 1997, respectively. Total daytime net CO2 gains were estimated to be 647 and 614 g C·m22·yr21, and nighttime releases were 562 and 577 g C·m22·yr21 in 1996 and 1997, respectively. The highest estimated mean daily rate of net CO2 exchange at the cypress ecosystem in 1996 occurred in May (1.2 6 0.1 g C·m22·d21, mean 6 1 SE), and the lowest in December (20.9 6 0.1 g C·m22·d21) (Fig. 9). We lack information on biomass or peat accumulation for this ecosystem. However, if we assume that carbon accumulation in wood in this stand is zero, because this ecosystem contains mature trees with a high stem density, and that the decomposition rate of cypress litter is 30–45%/yr (as measured using litter bags at this site in an earlier study; Ewel and Odum 1984), net accumulation of carbon from fine litterfall (182 to 232 g C·m22·yr21; Table 3) is 2.1–2.7 times greater than the estimated annual net CO2 exchange rate in 1996. This imbalance may well be the result of accumulated errors inherent in the extrapolation of both half-hourly flux measurements, particularly in March and April when the canopy is leafing out, and in the vegetation/detrital measurements

945

to obtain comparable annual values for net carbon accumulation at the ecosystem level. Although it is likely that our litterfall estimate adequately characterized the footprint area of the pine ecosystem, we are less certain of how litterfall varies spatially in the cypress ecosystem. Additionally, we have little data on decomposition rates of litter other than of cypress in this wetland (i.e., that of Nyssa and understory shrubs), but it is likely to be more rapid than cypress litter. The FDS term was larger for the cypress ecosystem during the growing season when compared to the pine ecosystem; thus it is likely that a portion of the respired CO2 was reassimilated at this site. Recent estimates of CO2 ‘‘recycling’’ range from 2 to 8% of net CO2 exchange rates, with higher values occurring at lower latitudes (Sternberg 1989, Lloyd et al. 1996, Brooks et al. 1997). In Florida, recycling is most likely to occur during the typically clear mornings in summer when LAI, respiration rates, and incident radiation are all high, especially after nights with low u* when significant amounts of CO2 have been stored in (and under) the canopy. Despite these uncertainties, our mean estimate of 61 g C·m22·yr21 for 1996 and 1997 compares well with the value of 70 g C·m22·yr21 obtained through extrapolation of enclosure measurements by Brown (1981), supporting the relatively low rates of net C accumulation for this ecosystem.

Comparisons with net CO2 exchange at other forest ecosystems When compared to ecosystem-level ‘‘light response curves’’ for net CO2 exchange estimated for other forest ecosystems, apparent light compensation points for both the pine and cypress ecosystems were similar. For example, when all the data for relatively high LAI and warm weather conditions were analyzed together, light compensation points for the pine and cypress ecosystems were ;125 and 150 mmol·m22·s21, respectively, as compared to 100–250 mmol·m22·s21 for a variety of boreal, temperate, and tropical forests (Table 5). When compared to a large number of above-canopy and chamber flux measurements reviewed by Ruimy et al. (1995), a and R estimated from measurements at the pine ecosystem during the growing season were similar (a 5 0.044 vs. 0.044 6 0.009, R 5 4.29 vs. 4.62 6 1.163). However, Fsat calculated for the pine ecosystem is lower than the mean value for their data set (43.34 vs. 26.54 6 1.33). At the leaf level, the maximum rate of net CO2 assimilation, Pmax, of slash pine is lower than a number of other Pinus species (Teskey et al. 1994a, b). Low maximum rates of net CO2 exchange may be linked to the relatively low nitrogen (N) concentrations in foliage at the slash pine ecosystem (10.0 6 1.1 mg N/g for first-yr foliage, 9.3 6 1.0 mg N/g for 2nd-yr foliage, mean 6 1 SD). When compared to the data set analyzed by Ruimy et al. (1995), a, R, and Pmax were all lower at the cypress wetland.

Ecological Applications Vol. 9, No. 3

KENNETH L. CLARK ET AL.

946

TABLE 5. Light compensation points and CO2 parameters for the present study and other selected forest ecosystems.

Forest ecosystem Slash pine plantation Cypress wetland Boreal black spruce Boreal black spruce Boreal jack pine Boreal aspen Boreal woodland Northern hardwood Southern hardwood Beech forest Nothofagus forest Radiata pine plantation Amazonian rain forest Amazonian rain forest

Mean maximum Mean nighttime Light hourly ecosystem respiration§ compensation net CO2 Annual net point† exchange‡ 0 8C 20 8C CO2 exchange\ 125 150 200 ··· 200 ··· ··· 100

14 8 9 8 12 8 7 22

1.5 1 1.5 1.8 1.5 1 1 1.5

4.8 3.4 6 6.9 5 8†† 4.5 4.5

200 ··· 100–250 100 250 ···

28 18 12 18 15 15

··· ··· 2 1 ··· ···

··· 3.2 5 5 6–7‡‡ 6

740, 610¶ 84, 37¶ ··· ;0 ··· 160 ··· 220 525 472 ··· ··· 102 ···

Reference this study this study Jarvis et al. 1997 Goulden et al. 1997 Baldocchi et al. 1997 Black et al. 1996 Fan et al. 1995 Wofsy et al. 1993, Goulden et al. 1996 Greco and Baldocchi 1996 Valentini et al. 1996 Hollinger et al. 1994 Arneth et al. 1998 Grace et al. 1996 Fan et al. 1990

† Units: mmol PPFD·m22·s21. ‡ Units: mmol CO2·m22·s 21, measured at 1500 mmol PPFD·m22·s21. § Units: mmol CO2·m22·s 21. \ Units: g C·m22·yr21. ¶ Estimates for 1996 and 1997. †† Estimated at a soil temperature of 158C. ‡‡ Estimated from a light response curve at 0 mmol PPFD·m22·s21.

Mean maximum net CO2 exchange rates measured at the pine ecosystem were low compared to those reported for temperate deciduous forests during the growing season, but were greater than those reported for boreal sites, and similar to those in tropical forest ecosystems (Table 5). For example, at a temperate hardwood forest in Tennessee and a northern hardwood forest in Massachusetts, net CO2 exchange at 1500 mmol PPFD·m22·s21 in the summer was ;28 and 22 mmol CO2·m22·s21, respectively (Baldocchi and Vogel 1996, Goulden et al. 1996). Mean maximum net CO2 exchange rates at the cypress wetland ecosystem at 1500 mmol PPFD·m22·s21 during the growing season were similar to boreal forests dominated by aspen (Black et al. 1996) and black spruce (Fan et al. 1995, Goulden et al. 1997, Jarvis et al. 1997), but lower than those at a boreal jack pine forest (Baldocchi et al. 1997). Wintertime net CO2 exchange rates at the boreal sites are likely to be considerably lower than those during the winter at the pine site in Florida or tropical sites, but not necessarily different from those of deciduous warm temperate forest ecosystems. Nighttime ecosystem respiration was estimated at ;1.5 and 1 mmol CO2·m22·s21 at an air temperature of 08C, and at 5 and 4 mmol CO2·m22·s21 for warm-weather periods at the pine and cypress ecosystems, respectively. These rates are within the ranges of those reported from a number of other sites (Table 5). About 48–71% of net CO2 emissions at the boreal sites were due to efflux from the soil (Lavigne et al. 1997), compared with ;60% at the Florida pine ecosystem (Ewel et al. 1987). At the tropical sites, ;80% of the net CO2 emissions were due to soil respiration (Fan et al. 1990).

The median Q10 of 2 for the relationship between net nighttime CO2 exchange and air temperature for both the Florida sites is typical for respiration rates of foliage, stems, and fine roots of slash pine (Cropper and Gholz 1991, Ryan et al. 1994). Other sites have reported similar Q10 values for the relationship between nighttime net CO2 exchange and air temperature. For example, Goulden et al. (1997) estimated a Q10 of 2.3 for a boreal black spruce forest, and Valentini et al. (1996) estimated a Q10 of 2.2 for a beech forest in central Italy. The median Q10 of 2.3 for the relationship between nighttime net CO2 exchange and soil temperature at 5 cm depth at the Florida pine ecosystem is close to the Q10 of 2.5 estimated for soil CO2 efflux at this site (Fang et al. 1998), and is within the range reported for net CO2 exchange and soil temperature from other sites. For example, Jarvis et al. (1997) reported a Q10 of 2.5 for this relationship for a boreal black spruce forest, and Goulden et al. (1996) reported a Q10 of 2.2 for a northern hardwood forest in Massachusetts. Interestingly, Wofsy et al. (1993) and Goulden et al. (1996) also showed large negative values for net CO2 exchange from the latter site in the winter, even though both air and soil temperatures were much lower than those measured in Florida. Summer mean daily carbon accumulation rates at the pine ecosystem were generally lower than those estimated for temperate deciduous forests during the growing season (3.0 to 6.0 g C·m22·d21; Goulden et al. 1996), but greater than those estimated for the tropical rainforest sites during both dry and wet seasons (1.1 and 0.6 g C·m22·d21, respectively; Grace et al. 1996). When

August 1999

CO2 EXCHANGE IN FLORIDA FORESTS

compared to the tropical sites, the higher rate of net carbon accumulation at the Florida pine ecosystem is likely due to lower respiration rates, because total biomass, forest floor mass, and soil organic matter contents are much lower, and to the greater biomass increment in the relatively young pine plantation (Table 1). Summer mean daily carbon accumulation at the cypress ecosystem was similar to rates estimated for boreal black spruce and the tropical rainforest sites. Annual net CO2 exchange rates for forest ecosystems vary widely (Table 5). Many reports of tower-based eddy covariance measurements do not include a detailed accounting of the internal sources and sinks of carbon in the ecosystem; thus an interpretation of this variation is problematic. Nevertheless, some patterns are apparent. First, all of the hardwood forests in Table 5 are relatively young and have high annual net CO2 exchange rates. The northern and southern hardwood forests are recovering from agricultural clearing or other disturbances early in this century, and the aspen stand in Canada represents an early successional stage of ecosystem development following fire. Second, the low-latitude temperate forests (southern hardwood, beech) have much greater annual net CO2 exchange rates than either the boreal or the tropical rainforests, with the former likely due to the short growing season, and the latter due to the relatively long time since major disturbance. However, clear interpretation of the carbon balances of forest ecosystems, estimated using either tower-based or ground-based measurements, would benefit from the inclusion of a greater range of forest ecosystems in terms of environmental conditions, stand ages, and management histories. ACKNOWLEDGMENTS We thank Ian Beverland, Shuguang Liu, Chang Ming Fang, Suzy Brock, Jose Luis Hierro, Nate Wafford, Amy Konopacky, Ryan Harris, and Steven Smitherman for assistance in the field. We thank the Jefferson-Smerfit Corporation for allowing access to the slash pine ecosystem. Eddy provided security at the cypress site. This research was funded by Department of Energy, National Institute for Global Environmental Change (NIGEC). This is Florida Agricultural Experiment Station, Journal Series No. R-06647. LITERATURE CITED Arneth, A., F. M. Kelliher, T. M. McSeveny, and J. N. Byers. 1998. Fluxes of carbon and water in a Pinus radiata forest subject to soil water deficit. Australian Journal of Plant Physiology 25:557–570. Baldocchi, D. D., R. Valentini, S. Running, W. Oechel, and R. Dahlman. 1996. Strategies for measuring and modelling carbon dioxide and water vapor fluxes over terrestrial ecosystems. Global Change Biology 2:159–168. Baldocchi, D. D., and C. A. Vogel. 1996. Energy and CO2 flux densities above and below a temperate broad-leaved forest and a boreal pine forest. Tree Physiology 16:5–16. Baldocchi, D. D., C. A. Vogel, and B. Hall. 1997. Seasonal variation of carbon dioxide exchange rates above and below a boreal jack pine forest. Agricultural and Forest Meteorology 83:147–170. Beverland, I. J., D. H. O’Neill, S. L. Scott, and J. B. Moncrieff. 1996. Design, construction, and operation of a flux

947

measurement system using the conditional-sampling technique. Atmospheric Environment 30A:3209–3220. Beverland, I. J., S. L. Scott, D. H. O’Neill, and J. B. Moncrieff. 1997. Simple battery powered device for flux measurements by conditional sampling. Atmospheric Environment 31A:277–281. Black, T. A., G. den Hartog, H. H. Neumann, P. D. Blanken, P. C. Yang, C. Russell, Z. Nesic, X. Lee, S. G. Chen, R. Staebler, and M D. Novak. 1996. Annual cycles of water vapour and carbon dioxide fluxes in and above a boreal aspen forest. Global Change Biology 2:219–229. Brooks, J. R., L. B. Flanagan, G. T. Varney, and J. R. Ehleringer. 1997. Verical gradients in photosynthetic gas exchange and refixation of respired CO2 within boreal forest canopies. Tree Physiology 17:1–12. Brown, M. J. 1996. Forest statistics for Florida, 1995. U.S. Forest Service Southern Research Station SRS 6. Brown, S. 1981. A comparison of the structure, primary productivity, and transpiration of cypress ecosystems in Florida. Ecological Monographs 51:403–427. Cropper, W. P. Jr.,, and H. L. Gholz. 1991. In situ needle and fine root respiration in mature slash pine (Pinus elliottii) trees. Canadian Journal of Forest Research 21:1589–1595. Cropper, W. P., Jr., and H. L. Gholz. 1993. Simulation of the carbon dynamics of a Florida slash pine plantation. Ecological Modeling 66:231–249. Ewel, K. C. 1998. Pondcypress swamps. Pages 405–420 in M. G. Messina and W. H. Conner. Southern forested wetlands: ecology and management. CRC, Boca Raton, Florida, USA. Ewel, K. C., W. P. Cropper, Jr., and H. L. Gholz. 1987. Soil CO2 evolution in Florida slash pine plantations. I. Changes through time. Canadian Journal of Forest Research 17:325– 329. Ewel, K. C., and H. T. Odum. 1984. Cypress swamps. University Presses of Florida, Gainesville, Florida, USA. Fan, S-C., S. C. Wofsy, P. S. Bakwin, and D. J. Jacob. 1990. Atmosphere–biosphere exchange of CO2 and O3 in the Central Amazon forest. Journal of Geophysical Research 95: 16851–16864. Fan, S-M., M. L. Goulden, J. W. Munger, B. C. Daube, P. S. Bakwins, S. C. Wofsy, J. S. Amthor, D. R. Fitzjarrald, K. E. Moore, and T. R. Moore. 1995. Environmental controls on the photosynthesis of a boreal lichen woodland: a growing season of whole-ecosystem exchange measurements by eddy correlation. Oecologia 102:443–452. Fang, C., J. B. Moncrieff, H. L. Gholz, and K. L. Clark. 1998. Soil CO2 efflux and its spatial variation in a Florida slash pine plantation. Plant and Soil 205:135–146. Gaston, L., P Nkedi-Kizza, G. Sawka, and P. S. Rao. 1990. Spatial variability of morphological properties at a Florida flatwoods site. Soil Science Society of America Journal 54: 527–533. Gholz, H. L., C. S. Perry, W. P. Cropper, Jr., and L. C. Hendry. 1985. Litterfall, decomposition, and nitrogen and phosphorus dynamics in a chronosequence of Slash Pine (Pinus elliottii) plantations. Forest Science 31:463–478. Gholz, H. L., S. A. Vogel, W. P. Cropper, Jr., K. McKelvey, K. C. Ewel, R. O. Teskey, and P. J. Curran. 1991. Dynamics of canopy structure and light interception in Pinus elliottii stands, North Florida. Ecological Monographs 61:33–51. Goulden, M. L., B. C. Daube, S.-M. Fan, D. J. Sutton, A. Bazzaz, J. W. Munger, and S. C. Wofsy. 1997. Physiological response of a black spruce forest to weather. Journal of Geophysical Research 102:28987–28996. Goulden, M. L., J. W. Munger, S.-M. Fan, B. C. Daube, and S. C. Wofsy. 1996. Measurements of carbon sequestration by long-term eddy covariance: methods and a critical evaluation of accuracy. Global Change Biology 2:169–182. Grace, J., Y. Malhi, J. Lloyd, J. McIntyre, A. C. Miranda, P.

948

KENNETH L. CLARK ET AL.

Meir, and H. S. Miranda. 1996. The use of eddy covariance to infer the net carbon dioxide uptake of Brazilian rain forest. Global Change Biology 2:209–217. Greco, S., and D. D. Baldocchi. 1996. Seasonal variations of CO2 and water vapour exchange rates over a temperate deciduous forest. Global Change Biology 2:183–197. Hollinger, D. Y., F. M. Kelliher, J. N. Byers, J. E. Hunt, T. M. McSeveny, and P. L. Weir. 1994. Carbon dioxide exchange between an undisturbed old-growth temperate forest and the atmosphere. Ecology 75:134–150. Jarvis, P. G., J. M. Massheder, S. E. Hale, J. B. Moncrieff, M. Rayment, and S. L. Scott. 1997. Seasonal variation in carbon dioxide, water vapor, and energy exchanges of a boreal black spruce forest. Journal of Geophysical Research 102:28953–28966 Katul, G. G., P. L. Finklestein, J. F. Clarke, and T. G. Ellestad. 1996. An investigation of the conditional sampling method used to estimate fluxes of active, reactive, and passive scalars. Journal of Applied Meteorology 35:1835–1845. Lavigne, M. B., M. G. Ryan, D. E. Anderson, D. D. Baldocchi, P. M. Crill, D. R. Fitzjarrald, M. L. Goulden, S. T. Gower, J. M. Massheder, J. H. McCaughey, M. Rayment, and R. G. Striegl. 1997. Comparing nocturnal eddy covariance measurements to estimates of ecosystem respiration made by scaling chamber measurements at six coniferous boreal sites. Journal of Geophysical Research 102: 28977–28985. Law, B. E., M. G. Ryan, and P. M. Anthoni. 1998. Seasonal and annual respiration of a ponderosa pine ecosystem. Global Change Biology, in press. Liu, S. 1996. Evapotranspiration from cypress (Taxodium ascendens) wetlands and slash pine (Pinus elliottii) uplands in North-Central Florida. Dissertation. University of Florida, Gainesville, Florida, USA. Liu, S., H. Riekerk, and H. L. Gholz. 1997. Leaf litterfall, leaf area index, and radiation transmittance in cypress wetlands and slash pine plantations in North-Central Florida. Wetlands Ecology and Management 4:257–271. Lloyd, J., B. Kruijt, D. Y. Hollinger, R. J. Francey, S.-Chin Wong, F. M. Kelliher, A. C. Miranda, G. D. Farquhar, J. H. C. Gash, N. N. Vygodskaya, I. R. Wright, H. S. Miranda, and E.-Detlef Schulze. 1996. Vegetation effects on the isotopic composition of atmospheric CO2 at local and regional scales: theoretical aspects and a comparison between

Ecological Applications Vol. 9, No. 3

rain forest in Amazonia and a boreal forest in Siberia. Australian Journal of Plant Physiology 23:371–399. McCrady, R. L., and E. L. Jokela. 1998. Canopy dynamics, light interception, and radiation use efficiency of selected loblolly pine families. Forest Science 44:64–72. Meyers, R. L., and J. J. Ewel, editors. 1990. Ecosystems of Florida. University of Central Florida Press, Orlando, Florida, USA. Moncrieff, J. B., J. M. Massheder, A. Verhoeh J. Elbers, B. Heusinkveld, S. Scott, H. de Bruin, P. Kabat, and P. J. Jarvis. 1997. A system to measure surface fluxes of energy, momentum and carbon dioxide. Journal of Hydrology 188– 189:589–611. NOAA. 1996. Climatological data, Florida. National Oceanic and Atmospheric Administration, National Climatic Center Environmental Data Service, Asheville, North Carolina, USA. Ruimy, A., P. G. Jarvis, D. D. Baldocchi, and B. Saugier. 1995. CO2 fluxes over plant canopies and solar radiation: a review. Advances in Ecological Research 26:1–68. Ryan, M. G., M. B. Lavigne, and S. T. Gower. 1997. Annual carbon cost of autotrophic respiration in boreal forest ecosystems in relation to species and climate. Journal of Geophysical Research 102:28871–28883. Ryan, M. G., S. Linder, J. M. Vose, and R. M. Hubbard. 1994. Dark respiration in pines. Ecological Bulletins (Copenhagen) 43:50–63. Sternberg, L. 1989. A model to estimate carbon dioxide recycling in forests using 13C/12C ratios and concentrations of carbon dioxide. Agricultural and Forest Meteorology 48: 163–173. Teskey, R. O., H. L. Gholz, and W. P. Cropper, Jr. 1994a. Influence of climate and fertilization on net photosynthesis of mature slash pine. Tree Physiology 14:1215–1227. Teskey, R. O., D. Whitehead, and S. Linder. 1994b. Photosynthesis and carbon gain by pines. Ecological Bulletins (Copenhagen) 43:35–49. Valentini, R., P. DeAngelis, G. Matteucci, R. Monaco, S. Dore, and G. E. Scarascia Mugnozza. 1996. Seasonal net carbon dioxide exchange of a beech forest with the atmosphere. Global Change Biology 2:199–207. Wofsy, S. C., M. L. Goulden, J. W. Munger, S. M. Fan, P. S. Bakwin, B. C. Daube, S. L. Bassow, and F. A. Bazzaz. 1993. Net exchange of CO2 in a mid-latitude forest. Science 260:1314–1317.