Open Access

Differential responses of carbon and water vapor fluxes to climate among evergreen needleleaf forests in the USA

  • Pradeep Wagle1, 13Email author,
  • Xiangming Xiao1, 2,
  • Thomas E. Kolb3,
  • Beverly E. Law4,
  • Sonia Wharton5,
  • Russell K. Monson6,
  • Jiquan Chen7,
  • Peter D. Blanken8,
  • Kimberly A. Novick9,
  • Sabina Dore10,
  • Asko Noormets11 and
  • Prasanna H. Gowda12
Ecological Processes20165:8

DOI: 10.1186/s13717-016-0053-5

Received: 16 February 2016

Accepted: 23 May 2016

Published: 31 May 2016

Abstract

Introduction

Understanding the differences in carbon and water vapor fluxes of spatially distributed evergreen needleleaf forests (ENFs) is crucial for accurately estimating regional or global carbon and water budgets and when predicting the responses of ENFs to current and future climate.

Methods

We compared the fluxes of ten AmeriFlux ENF sites to investigate cross-site variability in net ecosystem exchange of carbon (NEE), gross primary production (GPP), and evapotranspiration (ET). We used wavelet cross-correlation analysis to examine responses of NEE and ET to common climatic drivers over multiple timescales and also determined optimum values of air temperature (T a) and vapor pressure deficit (VPD) for NEE and ET.

Results

We found larger differences in the NEE spectra than in the ET spectra across sites, demonstrating that spatial (site-to-site) variability was larger for NEE than for ET. The NEE and ET were decoupled differently across ENF sites because the wavelet cospectra between ET and climate variables were similar at all sites, while the wavelet cospectra between NEE and climate variables were higher (i.e., closer coupling between NEE and climatic drivers) in semi-arid and Mediterranean sites than in other sites. Ecosystem water use efficiency (EWUE) based on annual GPP/ET ranged from 1.3 ± 0.18 to 4.08 ± 0.62 g C mm−1 ET, while EWUE based on annual net ecosystem production (NEP)/ET ranged from 0.06 ± 0.04 to 1.02 ± 0.16 g C mm−1 ET) among ENFs. Responses of NEE and ET to T a varied across climatic zones. In particular, for ENF sites in semi-arid and Mediterranean climates, the maximum NEE and ET occurred at lower ranges of T a than in sites with warm and humid summers. The optimum T a and VPD values were higher for ET than for NEE, and ET was less sensitive to high values of T a and VPD.

Conclusions

Large spatial variability in carbon and water vapor fluxes among ENFs and large variations in responses of NEE and ET to major climate variables among climatic zones necessitate sub-plant functional type parameterization based on climatic zones to better represent climate sensitivity of ENFs and to reduce uncertainty in model predictions.

Keywords

Drought Ecosystem water use efficiency Net ecosystem CO2 exchange Spatial variability Wavelet cross-correlation analysis

Introduction

Forest ecosystems play a significant role in global carbon and water cycles because they cover over four billion hectares of the Earth’s land surface (Dixon et al. 1994). Evergreen needleleaf forests (ENFs) cover about 57 % of the forested land in the USA (Smith et al. 2009). In the past two decades, eddy covariance (EC) systems have been deployed at several (>20) ENFs across the USA to help understand the seasonality, magnitudes, and processes controlling surface-atmosphere fluxes (Dore et al. 2012; Falk et al. 2008; Hadley and Schedlbauer 2002; Hollinger et al. 1999; Jassal et al. 2009; Monson et al. 2002; Noormets et al. 2007; Noormets et al. 2010; Powell et al. 2008; Schwarz et al. 2004; Stoy et al. 2008). With the exception of the 10–20 years following a stand-replacement disturbance, ENFs are generally net sinks of carbon from the atmosphere (Amiro et al. 2010; Law et al. 2001). These independent site-level studies have reported varying rates of carbon and water vapor exchange and carbon uptake period (CUP) among ENFs due to substantial differences in forest age, land use history, topography, edaphic and climatic conditions, and species composition. Some studies have shown that the magnitudes of fluxes are highly sensitive to hydrologic conditions (Noormets et al. 2010; Stoy et al. 2005; Whelan et al. 2013) and canopy development after a disturbance (Clark et al. 2004). In addition to climate and structural changes, variations in biotic responses (Richardson et al. 2007) also cause variation in forest-atmosphere carbon exchange. Forest carbon cycling and the distribution and composition of forest species are tightly coupled with water availability (Berner and Law 2015). The broad distribution of ENFs across climatic gradients adds to the complexity of estimating fluxes for large regions and widespread vegetation types, since the sensitivity of fluxes for different ENF ecosystems to climate variability is not completely understood. Such understanding is necessary to enhance our ability to predict the potential responses of ENFs to the changing climate.

Direct measurements of carbon uptake and water loss by terrestrial ecosystems using the EC technique make it possible to characterize ecosystem water use efficiency (EWUE) (Arneth et al. 2006; Law et al. 2002) and to quantify the coupling between carbon and water cycles (Yu et al. 2004). At the ecosystem level, EWUE is generally estimated as the ratio of carbon uptake (gross primary production, GPP or net ecosystem production, NEP) to water loss (evapotranspiration, ET). However, the relationship between carbon gain and water loss is complicated at the ecosystem level. The EWUE is different from the concept of water use efficiency (WUE) that is conventionally calculated at leaf or individual plant level because several additional factors, such as heterotrophic respiration, decomposition of organic materials, and evaporation of soil moisture, can potentially influence EWUE (Emmerich 2007). The influence of vapor pressure deficit (VPD) on canopy conductance (Bierhuizen and Slatyer 1965; Monteith and Greenwood 1986) can have confounding effects on the individual responses of carbon uptake and water loss to the climatic conditions. Beer et al. (2009) proposed the “inherent water use efficiency (IWUE)” approach to account for the effect of VPD on WUE at the ecosystem level. In addition, higher temperature leads to higher VPD which limits carbon assimilation via stomatal (Turner et al. 1984) or non-stomatal (Morison and Gifford 1983) effects. Elevated temperatures will also increase photorespiration and heterotrophic respiration and, in turn, reduce carbon assimilation (Long 1991). The variability in EWUE among ecosystems reflects the water-use strategy (i.e., trade-off between water loss and carbon gain) of different species (Donovan and Ehleringer 1991) and provides an important insight into the relationships between carbon uptake and water loss. However, understanding of the variations in carbon and water vapor fluxes and EWUE over climatic gradients in ENFs is still lacking.

An increasing number of EC flux towers now allow synthesis, comparative analysis, and modeling of ecosystem-level fluxes (Falge et al. 2002; Gilmanov et al. 2010; Turner et al. 2003; Wagle et al. 2015a; Xiao et al. 2014). Comparison and synthesis of eddy flux data from multiple sites for a plant functional type (PFT) across a large climatic gradient facilitates investigation of complex biophysical controls on carbon and water vapor fluxes (Schwarz et al. 2004; Thornton et al. 2002). However, few studies have investigated the variability in carbon and water vapor fluxes and their differential responses to major climate variables (e.g., rainfall, temperature) at widely distributed diverse ENFs. It is well known that plants adapt to prevailing climatic conditions to optimize their functions under a varying climate (Berry and Bjorkman 1980; Mooney et al. 1978). We hypothesize that carbon and water vapor fluxes of various ENFs in different climatic zones respond differently to major climatic variables. Thus, investigating the response of carbon uptake and ET to major climatic variables at multiple timescales and sites within widely distributed diverse ENFs is necessary to move beyond site-level measurements and into realistic carbon and water budget estimates over regions or continents.

Eddy fluxes are controlled by physical and biological processes that change over a wide range of temporal scales ranging from seconds to decades (Baldocchi and Wilson 2001; Katul et al. 2001; Schimel et al. 2001). Eddy fluxes are governed by complex, turbulent eddy motion (rich in spectral properties) over seconds, while interactions with weather patterns, bulk turbulent flow characteristics, eco-physiological and biochemical characteristics of plant canopy, and the hydrological conditions induce variability in fluxes at hourly to monthly timescales (Katul et al. 2001). Similarly, synoptic weather patterns superimposed on seasonal variations dominate the variability in fluxes over monthly to annual timescales. Wavelet cross-correlation spectral analysis suitably interprets eddy flux time series data and assesses the contributions of the processes that act on different frequencies (Katul et al. 2001; Ouyang et al. 2014). The spectral analysis quantifies the energetic frequencies/timescales of net ecosystem exchange of carbon (NEE) or ET fluxes and relates them to the timescales of variability in climatic drivers. Descriptions of the use of spectral analysis in investigations of eddy fluxes can be found in previous publications (Baldocchi et al. 2001; Katul et al. 2001; Katul and Parlange 1995; Ouyang et al. 2014; Torrence and Compo 1998). To date, only a few studies have used wavelet cross-correlation analysis to investigate variability in vegetation surface fluxes at a small number of ENF sites (Katul et al. 2001; Novick et al. 2015; Stoy et al. 2005). Thus, this current study investigates the responses of carbon and water vapor fluxes to common climatic drivers by using the wavelet cross-correlation spectral analysis on ten widely distributed ENFs across the USA (Fig. 1).
Fig. 1

Spatial distribution of the AmeriFlux sites used in this study. The base map of forest cover is derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) land-cover map

By analyzing eddy flux and climate data from ten widely distributed ENFs, the following questions are addressed in this study: (1) to what extent do the magnitudes and annual or seasonal sums of carbon and water vapor fluxes differ among sites? (2) how do the relationships between carbon uptake and water loss and seasonal variations of EWUE differ among sites? (3) how do the responses of NEE and ET to climate forcings differ among sites and over multiple timescales (from week to interannual)? and (4) to what extent do the responses of NEE and ET to T a and VPD differ among sites? Furthermore, by examining the time series in flux measurements, we assess conditional statistics for carbon and water vapor fluxes associated with changes in seasonal and interannual climatic conditions.

Methods

Study sites

We accessed the eddy flux data from ten AmeriFlux ENF sites (Fig. 1), with each site having a minimum of 4 years of surface flux data over a period from the late 1990s through 2013. The sites are distributed across the conterminous USA (Fig. 1) and cover a broad range of climates (Table 1). Detailed site information is available in previous studies (see references in Table 1) or on the AmeriFlux website (http://ameriflux.lbl.gov/).
Table 1

Site characteristics of ten evergreen needleleaf forest sites included in this study

Site (symbol)

Latitude longitude elevation (m)

MAT MAP SWC MAR

Study period

Vegetation

Climate

Stand age

References

Austin Cary, FL (US-SP1)

29.7381

20.19

2001–2006 (missing 2002, 2004)

Pine forest (longleaf pine, slash pine)

Warm and humid summers and mild winters

80

(Powell et al. 2008)

−82.2188

833

44

6028

Duke Forest Loblolly Pine, NC (US-DK3)

35.9782

14.87

2001–2008

Loblolly pine

Warm and humid summers and mild winters

31

(Novick et al. 2015)

−79.0942

1051

163

29.45

5185

North Carolina Loblolly Pine, NC (US-NC2)

35.8031

15.09

2005–2013

Loblolly pine

Warm and humid summers and mild winters

22

(Noormets et al. 2010)

−76.6679

1105

12

25.93

4380

Howland Forest Main, ME (US-Ho1)

45.2041

6.51

1996–2004

Red spruce, eastern hemlock, other conifers

Continental temperate

100

(Hollinger et al. 1999)

−68.7402

809

60

4096

Howland Forest West Tower, ME (US-Ho2)

45.2091

6.42

1999–2004

90 % evergreen needleleaf and 10 % broadleaf deciduous

Continental temperate

100

(Hollinger et al. 1999)

−68.7470

786

91

4132

Niwot Ridge, CO (US-NR1)

40.0329

2.41

1999–2007

Subalpine fir, Engelmann spruce, Lodgepole pine

Continental temperate to montane

110

(Monson et al. 2002)

−105.5464

665

3050

13.37

5972

Flagstaff Unmanaged Forest, AZ (US-Fuf)

35.0890

8.59

2006–2010

Ponderosa pine

Semi-arid

100

(Dore et al. 2012)

−111.7620

562

2180

28.83

6661

Metolius Intermediate Pine, OR (US-Me2)

44.4523

7.41

2002–2007 (missing 2003)

Ponderosa pine

Semi-arid

66

(Thomas et al. 2009)

−121.5574

431

1253

22.72

5827

Blodgett Forest, CA (US-Blo)

38.8952

11.16

2000–2006

Ponderosa pine (mixed evergreen coniferous)

Mediterranean

24

(Kurpius et al. 2002)

−120.6327

1306

1315

24.79

5706

Wind River, WA (US-Wrc)

45.8205

9.33

1999–2008 (missing 2005)

Douglas-fir and western hemlock

Mediterranean

~500

(Wharton et al. 2012)

−121.9519

2099

371

29.29

4825

MAT, MAP, SWC, and MAR were determined for each year first, then averaged over the study period

MAT mean annual average temperature (°C), MAP mean cumulative annual precipitation (mm), SWC mean annual average volumetric soil water content (%), MAR mean cumulative annual solar radiation (MJ m−2 year−1)

Flux and climate data

The EC technique was used to measure carbon and water vapor fluxes at all sites. We acquired site-specific gap-filled level 4 eddy flux data (half hourly, daily, and weekly) and climate data, including air temperature (T a), VPD, photosynthetically active radiation (PAR), soil water content (SWC), and precipitation, from the AmeriFlux data archive (http://ameriflux.ornl.gov/) or from site investigators (A. Noormets, US-NC2; K. Novick, US-DK3; S. Dore, US-Fuf; S. Wharton, US-Wrc). Marginal distribution sampling (MDS) (Reichstein et al. 2005) or the artificial neural network (ANN) (Papale and Valentini 2003) was used to fill gaps in the AmeriFlux Level 4 carbon flux data. For the AmeriFlux dataset, we used the gap-filled carbon flux data based on the MDS method. The flux data obtained from site investigators were corrected and gap filled by researchers of each site. Site-specific data gap-filling procedures can be found in site-specific publications (Table 1). For example, site researchers filled the gaps using the MDS method for US-DK3 site. No obvious differences were reported for different gap-filling approaches in comparison studies (Moffat et al. 2007), implying that any differences due to different gap-filling techniques are unlikely to significantly affect the current analysis. The GPP was computed as the balance between ecosystem respiration (ER) and NEE by partitioning NEE data (Lasslop et al. 2010; Reichstein et al. 2005). To compare the responses of NEE and ET to T a and VPD among sites, we categorized half-hourly daytime (global radiation >5 W m−2) NEE and ET values in ten classes of increasing T a and VPD, with class intervals of 3–5 °C for T a and 0.3–0.5 kPa for VPD depending on the available range of T a and VPD. The optimum values of T a and VPD for NEE and ET were determined from the curves relating NEE and ET with T a and VPD, respectively.

To characterize the seasonality of fluxes and net CUP, we averaged 8-day composite flux data for the entire study period from a single site into a single composite year and produced a mean 8-day time series of flux and climate data. The number of negative NEE days (i.e., carbon gain by the ecosystem) during the growing season was determined as CUP (Churkina et al. 2005; Wagle et al. 2015b). Integrated fluxes (calculated by summing daily values) at the annual (NEEyr, GPPyr, and ETyr) and seasonal (NEECUP, GPPCUP, and ETCUP: for the length of CUP) scales, and maximum daily values (NEEmax, GPPmax, and ETmax) of carbon uptake and ET during the growing season were calculated for each site year.

To quantify the coupling between carbon and water cycles at the ecosystem level, we computed EWUE based on GPP and ET (EWUEGPP_yr = GPPyr/ETyr) and NEP and ET (EWUENEP_yr = NEPyr/ETyr) at the annual scale. Similarly, EWUEGPP_CUP and EWUENEP_CUP for the length of CUP were derived from the ratio between GPPCUP and ETCUP and between NEPCUP and ETCUP, respectively. Seasonal patterns (8-day averages) of EWUE were determined from the time series of NEP, GPP, and ET.

Statistical analysis

We examined the cross-site relationship between integrated values of NEE, GPP, and ET at the seasonal and annual timescales. For each site, the relationships between GPP and ET and NEE and ET were also examined using 8-day average values during CUP over the study period.

The coefficient of determination (R 2) was calculated to show the strength of relationships between variables and the coefficient of variation (CV) for annual or seasonal integrals, and the maximum rates of fluxes (NEE, GPP, and ET) was calculated to characterize their spatial variations. We used a linear regression to analyze the relationship of the optimum T a for NEE and ET with mean annual average temperature (MAT). We presented relationships that best fit the data based on highest R 2 values.

We determined the wavelet spectra of NEE or ET time series to identify the proportion of the variability in NEE or ET attributable to processes operating at specific timescales/frequencies. The wavelet cospectra between two time series of fluxes (e.g., NEE, ET) and climatic drivers (e.g., VPD, T a, PAR, and SWC) were also explored to identify the frequencies at which they significantly co-resonate. An orthogonal wavelet transformation with the Haar basis function was performed to characterize the energetic frequencies/timescales of NEE and ET and to relate those to the timescales of variability in major climatic drivers. Flux and meteorological half-hourly time series data were normalized to have zero mean and unit variance and were zero-padded (gaps in data were also zero-padded) to a length equal to the power of 2 (218 in this case), which is greater than the length of the data record (up to 157,825 half-hourly data points) for any site in this study. We only determined and summed the wavelet coefficients within dyadic timescales ranging from 21 (hourly) to 216 (~4 years) because of the unreliability of the wavelet coefficients associated with longer timescales (217 or ~8 years) as in a previous study (Stoy et al. 2005). We performed the wavelet analyses in MATLAB using the Wavelet Toolbox (Mathworks, Natick, MA, USA).

Results

Seasonality and magnitudes of carbon uptake and ET

Seasonal patterns and magnitudes of carbon uptake (NEE and GPP) and ET varied among sites (Fig. 2) because carbon uptake was suppressed during extreme winters, mainly for continental temperate or montane sites where T a was generally less than 0 °C before DOY 100 and after DOY 300, and dry summers, mainly for US-SP1, US-Wrc, and US-Fuf. Table 2 shows large variations in magnitudes of NEE, GPP, and ET across study sites, with a CV of 26, 34, and 17 %, respectively. Mean net CUP varied from 121 days at US-Wrc to 365 days at US-DK3 and US-NC2, with a CV of 33 % (Table 2). The CUP was longer at sites with higher MAT, except at the US-SP1 site. Consequently, CUP showed a positive linear correlation (R 2 = 0.63, P = 0.01) with MAT when the US-SP1 site was excluded from the cross-site analysis (Additional file 1: Figure S1). The length of CUP showed a strong positive linear correlation with spatial variations in mean NEECUP (R 2 = 0.81, P < 0.001) and NEEyr (R 2 = 0.73, P = 0.002) (Additional file 1: Figure S1).
Fig. 2

Seasonal changes of net ecosystem CO2 exchange (NEE), gross primary production (GPP), and evapotranspiration (ET) at ten forest sites. The 8-day average values over the study period from a single site were averaged into a single composite year to produce mean 8-day time series

Table 2

Mean net carbon uptake period (CUP, days) and maximum rates of net ecosystem CO2 exchange (NEEmax, g C m−2 day−1), gross primary production (GPPmax, g C m−2 day−1), and evapotranspiration (ETmax, mm day−1) with the standard deviation and coefficient of variation (CV) at ten forest sites

Site

CUP

NEEmax

GPPmax

ETmax

US-SP1

235

−4.16 ± 0.61

7.97 ± 2.07

4.16 ± 0.47

US-DK3

365

−7.75 ± 0.90

17.48 ± 2.24

5.76 ± 2.57

US-NC2

365

−7.02 ± 1.88

14.98 ± 0.97

6.26 ± 0.68

US-Ho1

217

−5.33 ± 0.72

12.50 ± 0.83

4.2 ± 0.34

US-Ho2

217

−5.12 ± 0.36

11.67 ± 1.30

4.32 ± 0.57

US-NR1

169

−3.78 ± 0.80

7.06 ± 0.77

4.91 ± 0.64

US-Fuf

210

−3.42 ± 0.45

6.71 ± 0.65

3.96 ± 0.50

US-Me2

297

−5.97 ± 0.55

9.75 ± 0.46

4.06 ± 0.43

US-Blo

321

−5.27 ± 1.28

7.65 ± 1.25

5.22 ± 0.92

US-Wrc

121

−5.38 ± 0.49

9.51 ± 1.63

4.31 ± 0.59

CV (%)

33

26

34

17

NEEmax, GPPmax, and ETmax were calculated for each year first, then averaged over the study period. NEE data over the study period were averaged for the same date into a single composite year at each site to determine the CUP

Variability in carbon and water budgets

Large variations in annual and seasonal sums of NEE, GPP, and ET were observed across study sites, with a CV of 83, 45 and 36 %, respectively, on an annual scale and 61, 58, and 52 %, respectively, on the seasonal scale (Table 3). Average NEEyr and NEECUP over all sites were −314 ± 261 (±SD) g C m−2 and −361 ± 222 g C m−2, respectively. In addition to site-to-site variability, ENFs showed large interannual variability in the carbon uptake potential in response to disturbances and climate forcings such as drought. Several sites switched from a net sink to a source of carbon or remained carbon neutral during drought years. For example, the US-SP1 site was a source of carbon in the drought year of 2001 (64 g C m−2 year−1) when the extreme drought caused premature senescence and loss of 50 % annual litterfall in May and June of 2000 and 2002, and the site was a large sink of carbon in 2005 (−200 g C m−2 year−1) when the site received above-average rainfall (>1228 mm). At the US-Fuf site, NEEyr varied from −19 g C m−2 in 2009 when annual precipitation was 296 mm to −174 g C m−2 in 2006 when annual precipitation was 686 mm. At the US-Blo site, NEEyr varied from −75 g C m−2 in 2001 to −931 g C m−2 in 2006, with the weak sink in 2001 likely due to pre-commercial thinning and mastication in 2000 that left debris to decompose on-site. The US-Wrc was a net source or a sink of carbon, largely depending on the springtime and early summer climatic conditions (NEEyr varied from 100 g C m−2 in 2003 to −221 g C m−2 in 1999).
Table 3

Annual and seasonal (carbon uptake period, CUP) sums of net ecosystem CO2 exchange (NEEyr and NEECUP, g C m−2), gross primary production (GPPyr and GPPCUP, g C m−2), and evapotranspiration (ETyr and ETCUP, mm) with the standard deviation and coefficient of variation (CV) at ten forest sites. Ecosystem water use efficiency (EWUE, g C mm−1 ET) was derived from the ratio between GPP and ET sums (EWUEGPP) and between net ecosystem production (NEP = −NEE) and ET sums (EWUENEP) at annual (year) and seasonal (CUP) timescales

Site

NEEyr

GPPyr

ETyr

EWUEGPP_yr

EWUENEP_yr

NEECUP

GPPCUP

ETCUP

EWUEGPP_CUP

EWUENEP_CUP

US-SP1

−70 ± 127

968 ± 344

581 ± 48

1.67 ± 0.57

0.13 ± 0.23

−118 ± 19

661 ± 232

376 ± 44

1.74 ± 0.45

0.32 ± 0.09

US-DK3

−501 ± 176

2337 ± 319

829 ± 75

2.82 ± 0.31

0.60 ± 0.19

Same

Same

Same

Same

Same

US-NC2

−788 ± 304

2839 ± 282

1053 ± 107

2.81 ± 0.45

0.77 ± 0.35

Same

Same

Same

Same

Same

US-Ho1

−256 ± 49

1472 ± 108

367 ± 57

4.08 ± 0.62

0.72 ± 0.21

−313 ± 44

1428 ± 98

343 ± 45

4.21 ± 0.54

0.93 ± 0.22

US-Ho2

−233 ± 39

1347 ± 94

361 ± 63

3.79 ± 0.44

0.67 ± 0.18

−290 ± 37

1296 ± 86

339 ± 52

3.87 ± 0.42

0.88 ± 0.21

US-NR1

−34 ± 23

737 ± 99

572 ± 77

1.30 ± 0.18

0.06 ± 0.04

−159 ± 16

666 ± 84

391 ± 36

1.71 ± 0.21

0.41 ± 0.04

US-Fuf

−112 ± 70

937 ± 94

506 ± 46

1.85 ± 0.14

0.22 ± 0.14

−188 ± 63

726 ± 89

371 ± 40

1.96 ± 0.13

0.50 ± 0.14

US-Me2

−497 ± 91

1335 ± 91

484 ± 40

2.78 ± 0.33

1.02 ± 0.16

−490 ± 74

1296 + 91

445 ± 39

2.93 ± 0.35

1.10 ± 0.13

US-Blo

−579 ± 276

1195 ± 257

681 ± 69

1.76 ± 0.37

0.83 ± 0.36

−578 ± 268

1141 ± 253

662 ± 67

1.72 ± 0.37

0.85 ± 0.36

US-Wrc

−66 ± 88

1370 ± 233

517 ± 71

2.71 ± 0.67

0.13 ± 0.15

−182 ± 40

617 ± 75

230 ± 35

2.74 ± 0.54

0.82 ± 0.26

CV (%)

83

45

36

40

67

61

58

52

36

35

NEE, GPP, and ET sums and EWUE were determined for each year first, then averaged over the study period

Relationship between carbon uptake and ET and variability in EWUE

To investigate the relationship between carbon uptake and water loss at individual sites, we performed regression analyses between time series (8-day average) of NEE or GPP and ET during CUP. We observed different (i.e., weak to strong) relationships between GPP and ET (R 2 values ranging from 0.35 to 0.81), while NEE and ET showed weak relationships at most of the sites (R 2 < 0.20 in eight out of ten sites) (Additional file 2: Table S1). The cross-site regression analysis also shows that ET was more strongly correlated with GPP than NEE. Spatial variation in ETyr explained 31 and 37 % of spatial variability in NEEyr and GPPyr, respectively, while spatial variation in ETCUP explained 54 and 68 % of the spatial variability in NEECUP and GPPCUP, respectively.

We observed large differences in the seasonality and magnitudes of EWUEGPP and EWUENEP (Fig. 3). The EWUEGPP was nearly constant throughout the year at sites with warm and humid summers. At continental or montane sites, EWUEGPP was around zero in winter and fluctuated greatly during the active growing season when GPP increased more than ET. At the semi-arid US-Me2 pine site, EWUEGPP increased steadily as drought progressed in summer and EWUEGPP peaked in early September when drought stress peaked. At the Mediterranean sites, the seasonal pattern of EWUEGPP differed from those of GPP and ET, with a maximum in winter when ET is smaller and a minimum in the peak growing season (i.e., dry summer months) when ET increased more than GPP due to drought effects. Seasonal change in EWUENEP was different from EWUEGPP. In general, EWUENEP started to increase with increasing T a, reached maximum values in spring, declined in warm-dry summers, and again rose in fall and declined in winter. As a result of large spatial variability in annually or seasonally integrated NEP, GPP, and ET values, we found a wide range of EWUEGPP_yr (1.3–4.08 g C mm−1 ET), EWUEGPP_CUP (1.71–4.21 g C mm−1 ET), EWUENEP_yr (0.06–1.02 g C mm−1 ET), and EWUENEP_CUP (0.32–1.1 g C mm−1 ET), with a CV of 40, 36, 67, and 35 %, respectively (Table 3).
Fig. 3

Seasonal changes (8-day averages) of ecosystem water use efficiency (EWUEGPP and EWUENEP) based on gross primary production (GPP) and evapotranspiration (ET) and net ecosystem production (NEP) and ET at ten forest sites, respectively

Spectral characteristics of NEE and ET across sites

Large spatial variability in carbon uptake, ET, and EWUE among ENFs calls for a closer examination of the spectral characteristics of NEE and ET and the cospectra and coherence of NEE and ET with governing climate variables such as VPD, T a, PAR, and SWC from week to interannual timescales using the wavelet analysis. The normalized wavelet spectra show that there were larger differences in the NEE spectra across sites than in the ET spectra at seasonal-to-interannual timescales (Fig. 4). Over longer timescales, variability in NEE and ET was relatively smaller in warm and humid summer sites and larger in semi-arid and Mediterranean sites. The comparison of wavelet spectra of NEE and ET for each site (Fig. 5) shows that they deviated from each other especially at seasonal-to-interannual timescales. The deviation between NEE and ET spectra was relatively larger in temperate continental and warm and humid summer sites and smaller in montane, semi-arid, and Mediterranean sites. The wavelet cospectra of NEE and ET to major governing climate variables showed that climate variables were positively correlated to NEE and ET at all timescales, but climate variables were more positively correlated to ET than NEE at seasonal and interannual timescales (Figs. 6 and 7). In particular, climate variables were less positively correlated to NEE at warm and humid summer sites than at other sites, while the wavelet cospectra between ET and climate variables were of a similar magnitude at all sites, except at the US-SP1 site. Variations in NEE and ET most resonated with VPD at weekly timescales and with T a at seasonal and interannual timescales.
Fig. 4

Normalized wavelet spectra of net ecosystem CO2 exchange (NEE, μmol m−2 s−1) and evapotranspiration (ET, mm 30-min−1) at ten forest sites. Half-hourly data were used in the analysis

Fig. 5

A comparison of the normalized wavelet spectra of net ecosystem CO2 exchange (NEE, μmol m−2 s−1) and evapotranspiration (ET, mm 30-min−1) at each forest site. Half-hourly data were used in the analysis

Fig. 6

Normalized wavelet cospectra between net ecosystem CO2 exchange (NEE) and major climate variables (vapor pressure deficit (VPD), photosynthetically active radiation (PAR), air temperature (T a), and soil water content (SWC)) at ten forest sites. Half-hourly data were used in the analysis. SWC data were not available in some sites

Fig. 7

Normalized wavelet cospectra between evapotranspiration (ET) and major climate variables (vapor pressure deficit (VPD), photosynthetically active radiation (PAR), air temperature (T a), and soil water content (SWC)) at ten forest sites. Half-hourly data were used in the analysis. SWC data were not available in some sites

Responses of NEE and ET to T a and VPD

The responses of NEE and ET to T a and VPD differed among ENFs (Fig. 8), but were similar in the same climatic zone with the exception of US-Blo in semi-arid/Mediterranean. The maximum NEE occurred near T a = 30 °C at warm and humid summer sites, ~22 °C at continental temperate sites, and ~15 °C at a montane, semi-arid, and Mediterranean (US-Wrc) sites, except at the US-Blo (Mediterranean) site where it occurred at ~25 °C. The ET increased alongside T a up to 33–35 °C at warm and humid summer sites, while it plateaued beyond 20 °C at continental temperate or montane, semi-arid, and Mediterranean sites, except at the US-Blo (Mediterranean) site where ET continued to rise with increasing T a up to 30 °C. The optimum T a values for NEE and ET showed a strong positive correlation (R 2 = 0.86, P < 0.001). The MAT was positively correlated with the optimum T a for NEE (R 2 = 0.68, P < 0.01) and ET (R 2 = 0.72, P < 0.01) across sites. The maximum NEE occurred near VPD = 1.5 kPa at warm and humid summer, continental temperate or montane sites, and at a semi-arid site (US-Fuf). It occurred at less than 1 kPa at a semi-arid site (US-Me2) and at a Mediterranean site (US-Wrc), but at ~2 kPa at the other Mediterranean site (US-Blo). The maximum ET occurred near VPD = 2.5 kPa at all sites except US-Fuf where it occurred at slightly less than 2 kPa. Results show that the optimum T a and VPD values were higher for ET than for NEE, and NEE declined more than ET at high T a and VPD values.
Fig. 8

Responses of net ecosystem CO2 exchange (NEE) and evapotranspiration (ET) to air temperature (T a) and vapor pressure deficit (VPD) at ten forest sites. Daytime (solar radiation >5 W m−2) half-hourly NEE and ET data over the study period were aggregated in ten classes of increasing T a and VPD

Discussion

The average NEEyr over all sites was −314 ± 261 (±SD) g C m−2 year−1, indicating that ENFs in the USA are generally net carbon sinks, except for early seral forests, but there are large variances among sites. The NEEyr was <−450 g C m−2 year−1 in four out of ten sites, which is outside of the range of observed NEEyr in two thirds of the ecosystems in the continental USA surveyed by Novick et al. (2015). A mean NEEyr of −183 ± 270 g C m−2 year−1 was reported for 506 site-years of data representing a range of global biomes (Baldocchi 2008). The average NEEyr became more negative (i.e., larger C sinks) with increasing MAT, similar to previous studies (Lindroth et al. 2008; Valentini et al. 2000), as the length of CUP increased with increasing MAT. The length of CUP was an important determinant of spatial variations in mean NEEyr, which is consistent with the findings of previous studies in boreal and deciduous forests (Baldocchi et al. 2005; Suni et al. 2003).

Despite being in the same PFT type, we observed a large spatial variability in carbon uptake and ET among ENF sites. The spatial variability was larger for carbon uptake than for ET (i.e., larger CV for NEE and GPP than for ET in Tables 2 and 3), which was consistent with findings for grasslands across the USA (Wagle et al. 2015a). Results of the wavelet analysis further support the larger spatial variability of NEE than ET, as we found larger differences in the NEE spectra than in the ET spectra across sites at longer (seasonal-to-interannual) timescales (Fig. 4). Larger spatial variability of NEE than of ET can be partly attributed to the fact that NEE is a more complex process than ET because NEE involves soil respiration, which is greatly influenced by the soil microbial community (Monson et al. 2006) and substrate availability and quality (Raich and Schlesinger 1992), which do not necessarily influence ET. Variation in NEE across multiple spatial and temporal scales is driven by dynamic variation of two biological processes at the ecosystem level: GPP or ER (which includes both autotrophic and heterotrophic respiration) in response to environmental effects (Stoy et al. 2009). In addition, carbon uptake (NEE or GPP) is constrained more than ET by VPD through restriction of stomatal regulation (Kuglitsch et al. 2008; Wagle and Kakani 2014).

The wavelet analysis showed that NEE and ET deviated from each other at seasonal-to-interannual timescales at all ENFs (Fig. 5) because climate variables were more positively correlated to ET (Fig. 7) than NEE (Fig. 6). However, the deviation between NEE and ET was greater in warm and humid summer and temperate continental sites than in other sites (Fig. 5), indicating that NEE and ET are decoupled differently across study sites. These differences in coupling between NEE and ET across sites can be attributed to differences in the wavelet cospectra of NEE and ET with climate variables. Unlike the wavelet cospectra between ET and climate variables, the wavelet cospectra between NEE and climate variables differed across climatic zones and were generally higher in semi-arid and Mediterranean sites, indicating closer coupling between NEE and climate drivers in these sites. As a result, we observed a relatively larger variability in NEE (Fig. 4) and a smaller deviation between NEE and ET spectra in semi-arid and Mediterranean sites than in warm and humid summer sites (Fig. 5).

The wavelet analysis also indicated that T a and VPD dominated the variability in NEE and ET over most timescales (Figs. 6 and 7). The responses of NEE and ET to T a and VPD were very different across climatic zones (Fig. 8). In particular, NEE declined more than ET at high T a and VPD values, and the maximum NEE occurred at lower ranges of T a for ENF sites in semi-arid and Mediterranean climate, which would promote greater coupling between NEE and these climate variables. The wavelet analysis further confirms that drivers that vary seasonally or annually were more important for NEE in semi-arid and Mediterranean sites. Over longer timescales, NEE is influenced by seasonal changes in climatic conditions, leaf dynamics, phenology, rainfall, snow events, and growing season duration (Baldocchi et al. 2001; Stoy et al. 2005). Larger variability in NEE in semi-arid and Mediterranean sites at longer timescales might be related to larger year-to-year variability in soil water availability and drought duration and the length of CUP in these sites. A positive linear relation between annual or seasonal sums of NEE and CUP length in this study suggests that CUP variability has a profound influence on interannual NEE fluxes, which is consistent with an earlier study (Randerson et al. 1999). Unlike the different magnitudes of the wavelet cospectra between NEE and climate variables, the wavelet cospectra between ET and climate variables were similar in magnitude at all sites because the optimum T a and VPD values were higher for ET than for NEE, and ET was less sensitive to high values of T a and VPD. The optimum T a values for NEE and ET increased with increasing MAT, suggesting the thermal acclimation of ecosystem properties (Mooney et al. 1978; Niu et al. 2012).

This study also highlights that the carbon source-sink status of ENFs was conditional upon seasonal changes in climatic conditions and that ENF stands of any age or climatic zone can be either a carbon source or carbon neutral under certain climatic conditions. Such alterations in carbon dynamics have been observed in several grasslands and forests worldwide (Amiro et al. 2006; Barr et al. 2007; Ciais et al. 2005; Wagle et al. 2015a). Although old-growth stands are traditionally viewed as net carbon sources or carbon neutral (DeBell and Franklin 1987) due to increased respiration and decreased photosynthesis, previous studies have suggested that old-growth forests can still be large carbon sinks under favorable climatic conditions (Carey et al. 2001; Luyssaert et al. 2008; Wharton et al. 2012). The carbon source or sink strength of any stand age depends on local climatic conditions (Chen et al. 2004; Wofsy et al. 1993). However, many studies reported that individual-tree-level carbon and water vapor fluxes varied with tree age and/or height (Bond 2000; Irvine et al. 2004; Law et al. 2001; Novick et al. 2009). For example, different seasonal EWUE trends of three Douglas-fir stands of 7, 19, and 58 years old were observed in similar ecosystems (within 50 km of each other) because of the greater impact of stand age on GPP than ET (Jassal et al. 2009). Our results also demonstrated that the maximum NEE and ET occurred at a higher range of T a or VPD for younger stands (24 years old) at the US-Blo site compared to older stands in semi-arid and Mediterranean sites (Fig. 8). In addition to climate controls and differences in stand age (varying from ~20 to ~500 years in this study) of ENFs, disturbance history and management practices might have also contributed to some portions of the spatial variability in carbon and water vapor fluxes, but evaluation of those contributions is outside the scope of this study.

Understanding spatial patterns of EWUE is important for elucidating the relationship between carbon uptake and water loss. Different seasonal patterns (Fig. 3) and magnitudes (Table 3) of EWUEGPP and EWUENEP and the difference in the relationships between NEE or GPP and ET (Additional file 2: Table S1) across sites indicate differences in trade-off between carbon uptake and water loss by ENFs. Our analyses illustrate that this observed large spatial variability in EWUE can be attributed to a larger spatial variations in carbon fluxes than in ET, a greater sensitivity of carbon fluxes than ET to climate, and differences in the coupling of carbon uptake and ET among sites. Some variations in EWUE among sites can also be attributed to differences in climatic conditions and dominant species (Yu et al. 2008). In addition, ET is composed of evaporation (E) and transpiration (T), and the fractions vary among sites and seasons, which affect EWUE estimations. These results suggest that improved representation of spatial and temporal differences in EWUE among ENFs is necessary in modeling.

Different methods to calculate EWUE complicate comparisons among studies. However, EWUE values in our study were within the reported range from past studies. Annual EWUEGPP of Douglas-fir stands of different ages on the east coast of Vancouver Island, BC, Canada, varied from 0.5 to 5.3 g C mm−1 ET (Jassal et al. 2009). The average EWUEGPP (derived from the slope of the relationship between monthly GPP and ET) value for evergreen conifers in FLUXNET was 0.65 g C mm−1 ET (Law et al. 2002). The EWUEGPP of ponderosa pine in central Oregon was 2.9 g C mm−1 ET (Law et al. 2002) and EWUEGPP of conifer plantation forest in eastern China was 2.5 g C mm−1 ET (Yu et al. 2008).

The same physiological and structural parameters are assigned to all ENF sites based on generalized PFTs for simulation of ecosystem processes. For example, production efficiency models (i.e., light use efficiency models) use a single light use efficiency value and terrestrial ecosystem models use a single optimum T a value per PFT based on the assumption that PFT-specific physiological parameters do not change with space or time (Raich et al. 1991; Running et al. 1999). Our study highlights large differences in the magnitudes of carbon and water vapor fluxes and their responses to climate among ENFs. Our results are consistent with the findings of previous studies (Law 2014; Van der Molen et al. 2011) that generalized PFTs cannot capture the differences in sensitivity of different species to climate. These results suggest that modeling carbon and water vapor fluxes of ENFs and projections of response of ENFs to climate based on the same PFT parameters can potentially cause extensive errors. Thus, sub-PFT parameterization based on climatic zones is required to better represent climate sensitivity of ENFs and to reduce uncertainty in model predictions.

Conclusions

This study highlights large spatial variability of carbon and water vapor fluxes and EWUE of widely distributed ENFs in the USA. We found large variations in optimum T a (15–30 °C for NEE and 20–35 °C for ET) and VPD (1–2 kPa for NEE and ~2.5 kPa for ET) for NEE and ET among ENFs. Furthermore, NEE showed greater sensitivity to high T a and VPD than did ET. The NEE and ET responded differently to climate variables as the wavelet cospectra between ET and climate variables were similar in magnitude at all sites, while the wavelet cospectra between NEE and climate variables were generally higher in semi-arid and Mediterranean sites. These results highlight that modeling carbon and water budgets of ENFs based on the same biome-specific parameters can potentially cause extensive errors in model predictions. Therefore, canopy photosynthesis and transpiration models used in predicting responses to future climate and in estimating regional or global carbon and water budgets should be considered using more refined parameterization of ENFs based on climatic zones. Long-term observations and a more complete sample of plant groups within the ENFs of the world are required for further validation of our results and for global application.

Abbreviations

ANN, artificial neural networking; CUP, carbon uptake period; CV, coefficient of variation; E, evaporation; ENF, evergreen needleaf forests; ET, evapotranspiration; EWUE, ecosystem water use efficiency; EWUEGPP_CUP, EWUE based on GPP and ET for the length of CUP; EWUEGPP_yr, EWUE based on GPP and ET at annual scale; EWUENEP_CUP, EWUE based on NEP and ET for the length of CUP; EWUENEP_yr, EWUE based on NEP and ET at annual scale; GPP, gross primary production; MAT, mean annual temperature; MDS, marginal distribution sampling; NEE, net ecosystem CO2 exchange; NEECUP, GPPCUP, and ETCUP, integrated NEE, GPP, and ET for the length of CUP; NEEyr, GPPyr, and ETyr, integrated NEE, GPP, and ET at the annual scale; NEP, net ecosystem production; PAR, photosynthetically active radiation; PFT, plant functional type; R 2, coefficient of determination; SWC, soil water content; T, transpiration; T a, air temperature; VPD, vapor pressure deficit; WUE, water use efficiency

Declarations

Acknowledgements

This study was supported in part by grants from the Agriculture and Food Research Initiative of the USDA National Institute of Food and Agriculture (NIFA; Grant No. 2013-69002 to P. Wagle, X. Xiao, and P. Gowda, and Grant No. 2013-67003-20652 to B. Law) and the National Science Foundation EPSCoR (IIA-1301789 to X. Xiao). Data were obtained from the AmeriFlux data archive (http://ameriflux.ornl.gov/). Data from the US-Me2 site were supported by US Department of Energy (Grant No. 65076) to B. Law; US-Fuf site was supported by the North American Carbon Program/USDA CREES NRI (2004-35111-15057, 2008-35101-19076) and Science Foundation Arizona (CAA 0-203-08) to T. Kolb; US-NR1 site was supported by grants from US Department of Energy [the National Institute for Climate Change Research (NICCR) and Terrestrial Carbon Processes Program (TCP)] and the National Science Foundation Environmental Biology (Grant 0918565); and US-Wrc was supported by an agreement among the University of Washington, the Pacific Northwest Research Station, and the Gifford Pinchot National Forest. Work at US-NC2 has been supported by DOE BER-TES awards number 7090112 and 11-DE-SC-0006700, USDA NIFA CAP 560 Award 2011-68002-30185, and USDA Forest Service Eastern Forest Environmental Threat Assessment Center Grant 08-JV-11330147-038. We are grateful to two anonymous reviewers and the editor for constructive suggestions which helped us improve the manuscript.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors’ Affiliations

(1)
Department of Microbiology and Plant Biology, Center for Spatial Analysis, University of Oklahoma
(2)
Institute of Biodiversity Science, Fudan University
(3)
School of Forestry, Northern Arizona University
(4)
Department of Forest Ecosystems and Society, Oregon State University
(5)
Atmospheric, Earth and Energy Division, Lawrence Livermore National Laboratory
(6)
School of Natural Resources and the Environment and Laboratory of Tree Ring Research, University of Arizona
(7)
Department of Geography, Michigan State University
(8)
Department of Geography, University of Colorado
(9)
School of Public and Environmental Affairs, Indiana University—Bloomington
(10)
Department of Environmental Science, Policy, and Management, University of California Berkeley
(11)
Department of Forestry and Environmental Resources, North Carolina State University
(12)
USDA-ARS Grazinglands Research Laboratory
(13)
USDA-ARS Grazinglands Research Laboratory

References

  1. Amiro B, Barr A, Black T, Iwashita H, Kljun N, McCaughey J, Morgenstern K, Murayama S, Nesic Z, Orchansky A (2006) Carbon, energy and water fluxes at mature and disturbed forest sites, Saskatchewan, Canada. Agr Forest Meteorol 136(3):237–251View ArticleGoogle Scholar
  2. Amiro B, Barr A, Barr J, Black T, Bracho R, Brown M, Chen J, Clark K, Davis K, Desai A (2010) Ecosystem carbon dioxide fluxes after disturbance in forests of North America. J Geophys Res-Biogeosci (2005–2012):115(G4)Google Scholar
  3. Arneth A, Veenendaal E, Best C, Timmermans W, Kolle O, Montagnani L, Shibistova O (2006) Water use strategies and ecosystem-atmosphere exchange of CO2 in two highly seasonal environments. Biogeosciences Discuss 3(2):345–382View ArticleGoogle Scholar
  4. Baldocchi D (2008) TURNER REVIEW No. 15. ‘Breathing’ of the terrestrial biosphere: lessons learned from a global network of carbon dioxide flux measurement systems. Aust J Bot 56(1):1–26View ArticleGoogle Scholar
  5. Baldocchi DD, Wilson KB (2001) Modeling CO2 and water vapor exchange of a temperate broadleaved forest across hourly to decadal time scales. Ecol Model 142(1):155–184View ArticleGoogle Scholar
  6. Baldocchi D, Falge E, Wilson K (2001) A spectral analysis of biosphere–atmosphere trace gas flux densities and meteorological variables across hour to multi-year time scales. Agr Forest Meteorol 107(1):1–27View ArticleGoogle Scholar
  7. Baldocchi DD, Black T, Curtis P, Falge E, Fuentes J, Granier A, Gu L, Knohl A, Pilegaard K, Schmid H (2005) Predicting the onset of net carbon uptake by deciduous forests with soil temperature and climate data: a synthesis of FLUXNET data. Int J Biometeorol 49(6):377–387View ArticleGoogle Scholar
  8. Barr AG, Black T, Hogg E, Griffis T, Morgenstern K, Kljun N, Theede A, Nesic Z (2007) Climatic controls on the carbon and water balances of a boreal aspen forest, 1994–2003. Glob Change Biol 13(3):561–576View ArticleGoogle Scholar
  9. Beer C, Ciais P, Reichstein M, Baldocchi D, Law B, Papale D, Soussana JF, Ammann C, Buchmann N, and Frank D (2009) Temporal and among-site variability of inherent water use efficiency at the ecosystem level. Glob Biogeochem Cy 23(2).Google Scholar
  10. Berner LT, Law BE (2015) Water limitations on forest carbon cycling and conifer traits along a steep climatic gradient in the Cascade Mountains, Oregon. Biogeosciences 12(22):6617–6635View ArticleGoogle Scholar
  11. Berry J, Bjorkman O (1980) Photosynthetic response and adaptation to temperature in higher plants. Annu Rev Plant Phys 31(1):491–543View ArticleGoogle Scholar
  12. Bierhuizen J, Slatyer R (1965) Effect of atmospheric concentration of water vapour and CO2 in determining transpiration-photosynthesis relationships of cotton leaves. Agr Meteorol 2(4):259–270View ArticleGoogle Scholar
  13. Bond BJ (2000) Age-related changes in photosynthesis of woody plants. Trends Plant Sci 5(8):349–353View ArticleGoogle Scholar
  14. Carey EV, Sala A, Keane R, Callaway RM (2001) Are old forests underestimated as global carbon sinks? Glob Change Biol 7(4):339–344View ArticleGoogle Scholar
  15. Chen J, Ustin SL, Suchanek TH, Bond BJ, Brosofske KD, Falk M (2004) Net ecosystem exchanges of carbon, water, and energy in young and old-growth Douglas-fir forests. Ecosystems 7(5):534–544View ArticleGoogle Scholar
  16. Churkina G, Schimel D, Braswell BH, Xiao X (2005) Spatial analysis of growing season length control over net ecosystem exchange. Glob Change Biol 11(10):1777–1787View ArticleGoogle Scholar
  17. Ciais P, Reichstein M, Viovy N, Granier A, Ogée J, Allard V, Aubinet M, Buchmann N, Bernhofer C, Carrara A (2005) Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437(7058):529–533View ArticleGoogle Scholar
  18. Clark KL, Gholz HL, Castro MS (2004) Carbon dynamics along a chronosequence of slash pine plantations in north Florida. Ecol Appl 14(4):1154–1171View ArticleGoogle Scholar
  19. DeBell DS, Franklin JF (1987) Old-growth Douglas-fir and western hemlock: a 36-year record of growth and mortality. W J Appl For 2(4):111–114Google Scholar
  20. Dixon RK, Solomon A, Brown S, Houghton R, Trexier M, Wisniewski J (1994) Carbon pools and flux of global forest ecosystems. Science 263(5144):185–190View ArticleGoogle Scholar
  21. Donovan LA, Ehleringer JR (1991) Ecophysiological differences among juvenile and reproductive plants of several woody species. Oecologia 86(4):594–597View ArticleGoogle Scholar
  22. Dore S, Montes‐Helu M, Hart SC, Hungate BA, Koch GW, Moon JB, Finkral AJ, Kolb TE (2012) Recovery of ponderosa pine ecosystem carbon and water fluxes from thinning and stand‐replacing fire. Glob Change Biol 18(10):3171–3185View ArticleGoogle Scholar
  23. Emmerich WE (2007) Ecosystem water use efficiency in a semiarid shrubland and grassland community. Rangeland Ecol Manag 60(5):464–470.View ArticleGoogle Scholar
  24. Falge E, Tenhunen J, Baldocchi D, Aubinet M, Bakwin P, Berbigier P, Bernhofer C, Bonnefond J-M, Burba G, Clement R (2002) Phase and amplitude of ecosystem carbon release and uptake potentials as derived from FLUXNET measurements. Agr Forest Meteorol 113(1):75–95View ArticleGoogle Scholar
  25. Falk M, Wharton S, Schroeder M, Ustin S (2008) Flux partitioning in an old-growth forest: seasonal and interannual dynamics. Tree Physiol 28(4):509–520View ArticleGoogle Scholar
  26. Gilmanov TG, Aires L, Barcza Z, Baron V, Belelli L, Beringer J, Billesbach D, Bonal D, Bradford J, Ceschia E (2010) Productivity, respiration, and light-response parameters of world grassland and agroecosystems derived from flux-tower measurements. Rangeland Ecol Manag 63(1):16–39View ArticleGoogle Scholar
  27. Hadley JL, Schedlbauer JL (2002) Carbon exchange of an old-growth eastern hemlock (Tsuga canadensis) forest in central New England. Tree Physiol 22(15-16):1079–1092View ArticleGoogle Scholar
  28. Hollinger D, Goltz S, Davidson E, Lee J, Tu K, Valentine H (1999) Seasonal patterns and environmental control of carbon dioxide and water vapour exchange in an ecotonal boreal forest. Glob Change Biol 5(8):891–902View ArticleGoogle Scholar
  29. Irvine J, Law B, Kurpius M, Anthoni P, Moore D, Schwarz P (2004) Age-related changes in ecosystem structure and function and effects on water and carbon exchange in ponderosa pine. Tree Physiol 24(7):753–763View ArticleGoogle Scholar
  30. Jassal RS, Black TA, Spittlehouse DL, Brümmer C, Nesic Z (2009) Evapotranspiration and water use efficiency in different-aged Pacific Northwest Douglas-fir stands. Agr Forest Meteorol 149(6):1168–1178View ArticleGoogle Scholar
  31. Katul GG, Parlange MB (1995) Analysis of land surface heat fluxes using the orthonormal wavelet approach. Water Resour Res 31(11):2743–2749View ArticleGoogle Scholar
  32. Katul G, Lai C-T, Schäfer K, Vidakovic B, Albertson J, Ellsworth D, Oren R (2001) Multiscale analysis of vegetation surface fluxes: from seconds to years. Adv Water Resour 24(9):1119–1132View ArticleGoogle Scholar
  33. Kuglitsch F, Reichstein M, Beer C, Carrara A, Ceulemans R, Granier A, Janssens I, Koestner B, Lindroth A, Loustau D (2008) Characterisation of ecosystem water-use efficiency of European forests from eddy covariance measurements. Biogeosciences Discuss 5(6):4481–4519View ArticleGoogle Scholar
  34. Kurpius MR, McKay M, Goldstein AH (2002) Annual ozone deposition to a Sierra Nevada ponderosa pine plantation. Atmos Environ 36(28):4503–4515View ArticleGoogle Scholar
  35. Lasslop G, Reichstein M, Papale D, Richardson AD, Arneth A, Barr A, Stoy P, Wohlfahrt G (2010) Separation of net ecosystem exchange into assimilation and respiration using a light response curve approach: critical issues and global evaluation. Glob Change Biol 16(1):187–208View ArticleGoogle Scholar
  36. Law BE (2014) Regional analysis of drought and heat impacts on forests: current and future science directions. Glob Change Biol 20(12):3595–3599View ArticleGoogle Scholar
  37. Law B, Goldstein A, Panek J, Bauer M, Fracheboud J, Anthoni P, Hultman N (2001) Carbon dioxide and water vapor exchange by young and old ponderosa pine ecosystems during a dry summer. Tree Physiol 21(5):299–308View ArticleGoogle Scholar
  38. Law B, Falge E, Gu LV, Baldocchi D, Bakwin P, Berbigier P, Davis K, Dolman A, Falk M, Fuentes J (2002) Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation. Agr Forest Meteorol 113(1):97–120View ArticleGoogle Scholar
  39. Lindroth A, Lagergren F, Aurela M, Bjarnadottir B, Christensen T, Dellwik E, Grelle A, Ibrom A, Johansson T, Lankreijer H (2008) Leaf area index is the principal scaling parameter for both gross photosynthesis and ecosystem respiration of Northern deciduous and coniferous forests. Tellus B 60(2):129–142View ArticleGoogle Scholar
  40. Long S (1991) Modification of the response of photosynthetic productivity to rising temperature by atmospheric CO2 concentrations: has its importance been underestimated? Plant Cell Environ 14(8):729–739View ArticleGoogle Scholar
  41. Luyssaert S, Schulze E-D, Börner A, Knohl A, Hessenmöller D, Law BE, Ciais P, Grace J (2008) Old-growth forests as global carbon sinks. Nature 455(7210):213–215View ArticleGoogle Scholar
  42. Moffat AM, Papale D, Reichstein M, Hollinger DY, Richardson AD, Barr AG, Beckstein C, Braswell BH, Churkina G, Desai AR (2007) Comprehensive comparison of gap-filling techniques for eddy covariance net carbon fluxes. Agr Forest Meteorol 147(3):209–232View ArticleGoogle Scholar
  43. Monson RK, Turnipseed AA, Sparks JP, Harley PC, Scott-Denton LE, Sparks K, Huxman TE (2002) Carbon sequestration in a high-elevation, subalpine forest. Glob Change Biol 8(5):459–478View ArticleGoogle Scholar
  44. Monson RK, Lipson DL, Burns SP, Turnipseed AA, Delany AC, Williams MW, Schmidt SK (2006) Winter forest soil respiration controlled by climate and microbial community composition. Nature 439(7077):711–714View ArticleGoogle Scholar
  45. Monteith J, Greenwood D (1986) How do crops manipulate water supply and demand? [and discussion]. Philos T Roy Soc A 316(1537):245–259View ArticleGoogle Scholar
  46. Mooney HA, Björkman O, Collatz GJ (1978) Photosynthetic acclimation to temperature in the desert shrub, Larrea divaricata I. Carbon dioxide exchange characteristics of intact leaves. Plant Physiol 61(3):406–410View ArticleGoogle Scholar
  47. Morison JI, Gifford RM (1983) Stomatal sensitivity to carbon dioxide and humidity a comparison of two C3 and two C4 grass species. Plant Physiol 71(4):789–796View ArticleGoogle Scholar
  48. Niu S, Luo Y, Fei S, Yuan W, Schimel D, Law BE, Ammann C, Altaf Arain M, Arneth A, Aubinet M (2012) Thermal optimality of net ecosystem exchange of carbon dioxide and underlying mechanisms. New Phytol 194(3):775–783View ArticleGoogle Scholar
  49. Noormets A, Chen J, Crow TR (2007) Age-dependent changes in ecosystem carbon fluxes in managed forests in northern Wisconsin, USA. Ecosystems 10(2):187–203View ArticleGoogle Scholar
  50. Noormets A, Gavazzi MJ, McNulty SG, Domec JC, Sun G, King JS, Chen J (2010) Response of carbon fluxes to drought in a coastal plain loblolly pine forest. Glob Change Biol 16(1):272–287View ArticleGoogle Scholar
  51. Novick K, Oren R, Stoy P, Juang J-Y, Siqueira M, Katul G (2009) The relationship between reference canopy conductance and simplified hydraulic architecture. Adv Water Resour 32(6):809–819View ArticleGoogle Scholar
  52. Novick KA, Oishi AC, Ward EJ, Siqueira MBS, Juang J-Y, Stoy PC (2015) On the difference in the net ecosystem exchange of CO2 between deciduous and evergreen forests in the southeastern United States. Glob Change Biol 21(2):827–842View ArticleGoogle Scholar
  53. Ouyang Z, Chen J, Becker R, Chu H, Xie J, Shao C, John R (2014) Disentangling the confounding effects of PAR and air temperature on net ecosystem exchange at multiple time scales. Ecol Complex 19:46–58View ArticleGoogle Scholar
  54. Papale D, Valentini R (2003) A new assessment of European forests carbon exchanges by eddy fluxes and artificial neural network spatialization. Glob Change Biol 9(4):525–535View ArticleGoogle Scholar
  55. Powell TL, Gholz HL, Clark KL, Starr G, Cropper WP, Martin TA (2008) Carbon exchange of a mature, naturally regenerated pine forest in north Florida. Glob Change Biol 14(11):2523–2538Google Scholar
  56. Raich J, Schlesinger WH (1992) The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus B 44(2):81–99View ArticleGoogle Scholar
  57. Raich J, Rastetter E, Melillo J, Kicklighter D, Steudler P, Peterson B, Grace A, Moore B III, Vorosmarty C (1991) Potential net primary productivity in South America: application of a global model. Ecol Appl 1(4):399–429View ArticleGoogle Scholar
  58. Randerson J, Field C, Fung I, Tans P (1999) Increases in early season ecosystem uptake explain recent changes in the seasonal cycle of atmospheric CO2 at high northern latitudes. Geophys Res Lett 26(17):2765–2768View ArticleGoogle Scholar
  59. Reichstein M, Falge E, Baldocchi D, Papale D, Aubinet M, Berbigier P, Bernhofer C, Buchmann N, Gilmanov T, Granier A (2005) On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm. Glob Change Biol 11(9):1424–1439View ArticleGoogle Scholar
  60. Richardson AD, Hollinger DY, Aber JD, Ollinger SV, Braswell BH (2007) Environmental variation is directly responsible for short‐but not long‐term variation in forest‐atmosphere carbon exchange. Glob Change Biol 13(4):788–803View ArticleGoogle Scholar
  61. Running SW, Nemani R, Glassy JM, Thornton PE (1999) MODIS daily photosynthesis (PSN) and annual net primary production (NPP) product (MOD17) Algorithm Theoretical Basis Document. University of Montana, SCF At-Launch Algorithm ATBD Documents (available online at: http://ntsg.umt.edu/sites/ntsg.umt.edu/files/modis/ATBD/ATBD_MOD17_v21.pdf)
  62. Schimel DS, House JI, Hibbard KA, Bousquet P, Ciais P, Peylin P, Braswell BH, Apps MJ, Baker D, Bondeau A (2001) Recent patterns and mechanisms of carbon exchange by terrestrial ecosystems. Nature 414(6860):169–172View ArticleGoogle Scholar
  63. Schwarz P, Law B, Williams M, Irvine J, Kurpius M, Moore D (2004) Climatic versus biotic constraints on carbon and water fluxes in seasonally drought‐affected ponderosa pine ecosystems. Glob Biogeochem Cy 18(4):GB4007View ArticleGoogle Scholar
  64. Smith WB, Miles PD, Perry CH, Pugh SA (2009) Forest resources of the United States, 2007: a technical document supporting the forest service 2010 RPA Assessment. General Technical Report-USDA Forest Service (WO-78)Google Scholar
  65. Stoy PC, Katul GG, Siqueira MB, Juang J-Y, McCarthy HR, Kim H-S, Oishi AC, Oren R (2005) Variability in net ecosystem exchange from hourly to inter-annual time scales at adjacent pine and hardwood forests: a wavelet analysis. Tree Physiol 25(7):887–902View ArticleGoogle Scholar
  66. Stoy PC, Katul GG, Siqueira M, Juang JY, Novick KA, McCarthy HR, Oishi AC, Oren R (2008) Role of vegetation in determining carbon sequestration along ecological succession in the southeastern United States. Glob Change Biol 14(6):1409–1427View ArticleGoogle Scholar
  67. Stoy P, Richardson A, Baldocchi D, Katul G, Stanovick J, Mahecha M, Reichstein M, Detto M, Law B, Wohlfahrt G (2009) Biosphere-atmosphere exchange of CO2 in relation to climate: a cross-biome analysis across multiple time scales. Biogeosciences 6(10):2297–2312View ArticleGoogle Scholar
  68. Suni T, Berninger F, Markkanen T, Keronen P, Rannik Ü, Vesala T (2003) Interannual variability and timing of growing‐season CO2 exchange in a boreal forest. J Geophys Res: Atmos 108(D9)Google Scholar
  69. Thomas CK, Law BE, Irvine J, Martin JG, Pettijohn JC, Davis KJ (2009) Seasonal hydrology explains interannual and seasonal variation in carbon and water exchange in a semiarid mature ponderosa pine forest in central Oregon. J Geophys Res: Biogeosci (2005–2012): 114(G4)Google Scholar
  70. Thornton P, Law B, Gholz HL, Clark KL, Falge E, Ellsworth D, Goldstein A, Monson R, Hollinger D, Falk M (2002) Modeling and measuring the effects of disturbance history and climate on carbon and water budgets in evergreen needleleaf forests. Agr Forest Meteorol 113(1):185–222View ArticleGoogle Scholar
  71. Torrence C, Compo GP (1998) A practical guide to wavelet analysis. B Am Meteorol Soc 79(1):61–78View ArticleGoogle Scholar
  72. Turner NC, Schulze E-D, Gollan T (1984) The responses of stomata and leaf gas exchange to vapour pressure deficits and soil water content. Oecologia 63(3):338–342View ArticleGoogle Scholar
  73. Turner DP, Ritts WD, Cohen WB, Gower ST, Zhao M, Running SW, Wofsy SC, Urbanski S, Dunn AL, Munger JW (2003) Scaling Gross Primary Production (GPP) over boreal and deciduous forest landscapes in support of MODIS GPP product validation. Remote Sens Environ 88(3):256–270View ArticleGoogle Scholar
  74. Valentini R, Matteucci G, Dolman A, Schulze E-D, Rebmann C, Moors E, Granier A, Gross P, Jensen N, Pilegaard K (2000) Respiration as the main determinant of carbon balance in European forests. Nature 404(6780):861–865View ArticleGoogle Scholar
  75. Van der Molen M, Dolman A, Ciais P, Eglin T, Gobron N, Law B, Meir P, Peters W, Phillips O, Reichstein M (2011) Drought and ecosystem carbon cycling. Agr Forest Meteorol 151(7):765–773View ArticleGoogle Scholar
  76. Wagle P, Kakani VG (2014) Growing season variability in evapotranspiration, ecosystem water use efficiency, and energy partitioning in switchgrass. Ecohydrology 7(1):64–72View ArticleGoogle Scholar
  77. Wagle P, Xiao X, Scott RL, Kolb TE, Cook DR, Brunsell N, Baldocchi DD, Basara J, Matamala R, Zhou Y (2015a) Biophysical controls on carbon and water vapor fluxes across a grassland climatic gradient in the United States. Agr Forest Meteorol 214:293–305Google Scholar
  78. Wagle P, Xiao X, Suyker AE (2015b) Estimation and analysis of gross primary production of soybean under various management practices and drought conditions. ISPRS J Photogramm 99:70–83Google Scholar
  79. Wharton S, Falk M, Bible K, Schroeder M, Paw UK (2012) Old-growth CO2 flux measurements reveal high sensitivity to climate anomalies across seasonal, annual and decadal time scales. Agr Forest Meteorol 161:1–14View ArticleGoogle Scholar
  80. Whelan A, Mitchell R, Staudhammer C, Starr G (2013) Cyclic occurrence of fire and its role in carbon dynamics along an edaphic moisture gradient in longleaf pine ecosystems. PLoS One 8(1):e54045View ArticleGoogle Scholar
  81. Wofsy S, Goulden M, Munger J, Fan S, Bakwin P, Daube B, Bassow S, Bazzaz F (1993) Net exchange of CO2 in a mid-latitude forest. Science 260(5112):1314–1317View ArticleGoogle Scholar
  82. Xiao J, Ollinger SV, Frolking S, Hurtt GC, Hollinger DY, Davis KJ, Pan Y, Zhang X, Deng F, Chen J (2014) Data-driven diagnostics of terrestrial carbon dynamics over North America. Agr Forest Meteorol 197:142–157View ArticleGoogle Scholar
  83. Yu G-R, Wang Q-F, Zhuang J (2004) Modeling the water use efficiency of soybean and maize plants under environmental stresses: application of a synthetic model of photosynthesis-transpiration based on stomatal behavior. J Plant Physiol 161(3):303–318View ArticleGoogle Scholar
  84. Yu G, Song X, Wang Q, Liu Y, Guan D, Yan J, Sun X, Zhang L, Wen X (2008) Water‐use efficiency of forest ecosystems in eastern China and its relations to climatic variables. New Phytol 177(4):927–937View ArticleGoogle Scholar

Copyright

© Wagle et al. 2016