Real-Time Partitioning of Diurnal Stem CO₂ Efflux into Local Stem Respiration and Xylem Transport Processes

Abstract

The apparent respiratory quotient (ARQ) of tree stems, defined as the ratio of net stem CO₂ efflux (E_{S_CO2}) to net stem O₂ influx (E_{S_O2}), offers insights into the balance between local respiratory CO₂ production and CO₂ transported via the xylem. Traditional static chamber methods for measuring ARQ can introduce artifacts and obscure natural diurnal variations. Here, we employed an open flow-through stem chamber with ambient air coupled with cavity ring-down spectrometry, which uses the molecular properties of CO₂ and O₂ molecules to continuously measure E_{S_CO2}, E_{S_O2}, and ARQ, at the base of a California cherry tree (Prunus ilicifolia) during the 2024 growing season. Measurements across three stem chambers over 3–11-day periods revealed strong correlations between E_{S_CO2} and E_{S_O2} and mean ARQ values ranging from 1.3 to 2.9, far exceeding previous reports. Two distinct diurnal ARQ patterns were observed: daytime suppression with nighttime recovery, and a morning peak followed by gradual decline. Partitioning E_{S_CO2} into local respiration and xylem-transported CO₂ indicated that the latter can dominate when ARQ exceeds 2.0. Furthermore, transported CO₂ exhibited a higher temperature sensitivity than local respiration, with both processes showing declining temperature sensitivity above 20 °C. These findings underscore the need to differentiate stem CO₂ flux components to improve our understanding of whole-tree carbon cycling.

1. Introduction

A large but highly uncertain fraction of carbon assimilated by photosynthesis is returned to the atmosphere as CO₂ during aerobic respiration in plant tissues [1], with estimates suggesting up to 70% in tropical forests [2]. Stem respiration contributes approximately a quarter of the global above-ground auto-trophic respiration, with an estimated annual emission of around 11.20 ± 5.88 Pg C, comparable to total anthropogenic emissions [3]. In trees, respiratory CO₂ produced in stem tissues during maintenance and growth processes can be emitted directly to the atmosphere via radial diffusion, measured as net stem CO₂ efflux (E_{S_CO2}, µmol m⁻² s⁻¹) [4,5,6,7,8,9,10,11]. However, field observations of diurnal E_{S_CO2} dynamics remain scarce due to the lack of open-flow chamber systems, with most studies relying on static chambers that recirculate air in a closed loop and estimate fluxes from accumulation rates [12]. These static systems often create artificial gas gradients, suffer from condensation and leakage, and can underestimate fluxes, obscuring natural diel variation [13,14]. This stands in contrast to common practice in leaf respiration studies, where open-flow systems continuously refresh headspace air, enabling real-time measurements. The absence of such capabilities for stems has limited our ability to assess how stem respiration responds to environmental and physiological drivers.

Beyond radial emission, respiratory CO₂ from stem tissues may also be transported upward in the transpiration stream due to its solubility, forming water-soluble carbon species (H₂CO₃, HCO₃⁻, CO₃²⁻) in dynamic equilibrium. Stable isotope studies have shown that transported CO₂ can be re-assimilated by leaves, potentially mitigating photosynthetic limitation under high temperature and transpiration [15,16,17,18,19,20,21,22,23,24,25,26,27]. For instance, one study found that up to 17% of ¹³CO₂ infused into tree stems was re-assimilated, with the remainder emitted via E_{S_CO2} [16]. Although the magnitude of this recycling effect is still debated, internal transport of CO₂ from roots, stems, and soils to the canopy may influence stem CO₂ efflux and contribute to the whole-tree carbon balance [17,18]. When stem E_{S_CO2} was partitioned in Pinus sylvestris, the authors found that transport-related fluxes accounted for up to 40% of the total stem CO₂ efflux during the peak growing season [28]. They employed a combination of stem chamber measurements, sap flow monitoring, and modeling to distinguish between locally respired CO₂, bark photosynthesis, and CO₂ transported via the xylem. Models like TreSpire attempt to simulate these interactions, but field validation remains limited [29,30].

A promising method to separate local respiration from transported CO₂ is by quantifying the apparent respiratory quotient (ARQ), defined as the ratio of net CO₂ efflux to net O₂ influx (ARQ = E_{S_CO2}/E_{S_O2}). This ratio provides insights into respiratory substrate use, CO₂ re-assimilation, and transport processes [27]. In principle, complete oxidation of non-structural carbohydrates yields an RQ of 1.0, meaning one molecule of CO₂ is released per molecule of O₂ consumed [31]. Thus, ARQ values greater than 1.0 imply that CO₂ efflux exceeds local respiratory production—likely due to additional transported CO₂—while values below 1.0 suggest that CO₂ is either retained within tissues, re-assimilated, or transported away (Figure 1).

Most previous ARQ studies using static chambers have reported values below 1.0, across diverse ecosystems including Mediterranean, tropical, and temperate forests (see Table 1) [28,32,33,34,35,36,37,38,39,40,41,42,43,44]. These findings support the idea that local stem-respired CO₂ may be derive from substrates with low cellular respiratory quotients (RQs) such as organic acids (RQ 0.7–0.9) and lipids (RQ ∼0.7), transported away in the transpiration stream, or refixed before emission. Proposed mechanisms include carboxylation by phosphoenolpyruvate carboxylase (PEPC), which fixes bicarbonate into C₄ acids like oxaloacetate or malate [45,46]. Although the extent of refixation in woody stems remains uncertain, these processes may contribute to low ARQ values in non-photosynthetic tissues. In green stems, additional light-dependent re-assimilation via RuBisCO has also been proposed [47,48,49].

Compounding these biological explanations, technical limitations of static chambers likely exacerbate underestimations of ARQ. Sealed environments quickly develop CO₂ concentration gradients, reduce diffusion rates, and often produce ARQ values 20–40% lower than true cellular RQ [50]. For example, ARQ values from static chambers decreased markedly with elevated CO₂ levels, from ∼1.0 at 300 ppm to <0.1 at 2000 ppm [43,44]. Similar findings in soil respiration studies confirm that dynamic open-flow chambers, which maintain ambient-like conditions, often yield higher and more accurate CO₂ efflux rates than static counterparts [51].

To address these challenges, we developed an open flow-through stem chamber system using ambient air and high-precision cavity ring-down spectroscopy (CRDS) for CO₂ and O₂ detection. This system enables continuous, real-time measurements of E_{S_CO2}, E_{S_O2}, and ARQ under field conditions with minimal disturbance to natural gas gradients. In this study, we deployed this system on a California cherry tree (Prunus ilicifolia), quantifying the diurnal dynamics and temperature sensitivities of stem CO₂ efflux, O₂ influx, and ARQ. By assuming a cellular RQ of 1.0, we partitioned E_{S_CO2} into contributions from local respiration and transported CO₂, providing a methodological framework for dissecting the drivers of stem CO₂ fluxes and their implications for carbon cycling in trees.

Table 1. Summary of stem respiratory quotient (RQ) measurements reported in the literature using static enclosed stem chambers under field conditions, compared to the current study on diurnal RQ observations obtained using an open flow-through stem chamber system with ambient air as the reference.

Reference	Method Used	Field/Lab, Intact/Cut	Range of ARQ Values	Species	Ecosystem	Diurnal Patterns?
Smart, 2004 [27]	Sealed static chamber	Cut branches vs. [CO2]	Branch: 0.088–0.99	Vitis rupestris × V. riparia cv. 3309 Couderc	Cultivated vineyards	No
Hilman et al., 2019 [17]	Sealed static chamber	Intact stems in the field	Stem: 0.23–0.90	Pinus halepensis, Quercus calliprinos	Mediterranean	No
Helm et al., 2023 [18]	Sealed static chamber	Intact stems in the field	Stem: 0.7	European beech (Fagus sylvatica)	Temperate	No
Helm et al., 2023 [52]	Sealed static chamber	Intact stems in the field	Stem: 0.7- 1.0	Populus tremula hybrids	Temperate	No
Hilman et al., 2022 [21]	Sealed static chamber	Stem cores	Stem: Cores: 0.19–0.70	Quercus calliprinos	Mediterranean	No
Hilman et al., 2019 [22]	Sealed static chamber	Intact stems	Stem: 0.39–0.78	Cedrela odorata, Swietenia macrophylla, Quercus ilex, Pinus halepensis.	Tropical, temperate, Mediterranean	No
Hilman and Angert, 2016 [23]	Sealed static chamber	Intact stems	Stem: 0.23–0.90	Quercus calliprinos, Pinus halepensis, Tetragastris panamensis (Engl.) Kuntze	Mediterranean	No
Patterson et al., 2018 [25]	Sealed static chamber	Intact stems in the field	Stem: 0.7–1.3	16 co-occurring temperate tree species, conifers and broadleaf	Temperate	No
Angert and Sherer, 2011 [38]	Sealed static chamber	Intact stems in the field	Stem: 0.61–0.84	Malus domestica, Pinus pinea L., Pinus halepensis Mill.	Mediterranean	No
Angert et al., 2012 [24]	Sealed static chamber	Intact stems in the field	Stem: 0.48–0.84	Hymenaea courbaril, Bertholletia excelsa, Cedrela odorata, Swietenia macrophylla	Tropical	No
Current study	Open flow chamber	Intact stems in the field	Stem: 0.9–2.9	Prunus ilicifolia	Temperate	Yes

Table 1. Summary of stem respiratory quotient (RQ) measurements reported in the literature using static enclosed stem chambers under field conditions, compared to the current study on diurnal RQ observations obtained using an open flow-through stem chamber system with ambient air as the reference.

Reference	Method Used	Field/Lab, Intact/Cut	Range of ARQ Values	Species	Ecosystem	Diurnal Patterns?
Smart, 2004 [27]	Sealed static chamber	Cut branches vs. [CO2]	Branch: 0.088–0.99	Vitis rupestris × V. riparia cv. 3309 Couderc	Cultivated vineyards	No
Hilman et al., 2019 [17]	Sealed static chamber	Intact stems in the field	Stem: 0.23–0.90	Pinus halepensis, Quercus calliprinos	Mediterranean	No
Helm et al., 2023 [18]	Sealed static chamber	Intact stems in the field	Stem: 0.7	European beech (Fagus sylvatica)	Temperate	No
Helm et al., 2023 [52]	Sealed static chamber	Intact stems in the field	Stem: 0.7- 1.0	Populus tremula hybrids	Temperate	No
Hilman et al., 2022 [21]	Sealed static chamber	Stem cores	Stem: Cores: 0.19–0.70	Quercus calliprinos	Mediterranean	No
Hilman et al., 2019 [22]	Sealed static chamber	Intact stems	Stem: 0.39–0.78	Cedrela odorata, Swietenia macrophylla, Quercus ilex, Pinus halepensis.	Tropical, temperate, Mediterranean	No
Hilman and Angert, 2016 [23]	Sealed static chamber	Intact stems	Stem: 0.23–0.90	Quercus calliprinos, Pinus halepensis, Tetragastris panamensis (Engl.) Kuntze	Mediterranean	No
Patterson et al., 2018 [25]	Sealed static chamber	Intact stems in the field	Stem: 0.7–1.3	16 co-occurring temperate tree species, conifers and broadleaf	Temperate	No
Angert and Sherer, 2011 [38]	Sealed static chamber	Intact stems in the field	Stem: 0.61–0.84	Malus domestica, Pinus pinea L., Pinus halepensis Mill.	Mediterranean	No
Angert et al., 2012 [24]	Sealed static chamber	Intact stems in the field	Stem: 0.48–0.84	Hymenaea courbaril, Bertholletia excelsa, Cedrela odorata, Swietenia macrophylla	Tropical	No
Current study	Open flow chamber	Intact stems in the field	Stem: 0.9–2.9	Prunus ilicifolia	Temperate	Yes

2. Materials and Methods

This study was conducted on a California cherry tree (Prunus ilicifolia) located in Berkeley, CA, USA, during the 2024 growing season. The tree was growing approximately 1 m from the analytical laboratory building and accessed via ¼″ Bev-a-line tubing extended out of the laboratory window to reach the base of the tree. During the first experiment, three stem chambers were installed at heights between 0.5 and 1.5 m on the stem with gas exchange measured in parallel. Two additional replicates using a chamber installed at a single location on the stem (0.5–1.5 m height) were also conducted. Continuous stem gas exchange measurements were conducted in three different periods over the summer, lasting 3–9 days, to capture diurnal and weekly patterns (29 June–1 July 2024, 13–19 August 2024, 22–30 August 2024). Air temperature was measured continuously at 1.5 m height near the tree using an Onset Mx100 (HOBO) (https://www.onsetcomp.com/, Cooperation: Onset, Bourne, MA, USA).

Following a light cleaning of the stem surface to remove debris, a 3D-printed opaque stem chamber (∼300 mL internal volume) was secured to the stem using two nylon straps [45]. After installation, a small sheet of reflective insulation was placed over the chamber and secured with two large zip ties to prevent artificial heating by sunlight. An ambient air reservoir (4 L) filled with small packs of silica gel to remove humidity was used to dampen rapid variations in CO₂ and O₂ concentrations and supplied ambient air via ¼″ Bev-a-line tubing to both the stem chamber inlets and ambient air inlets on a continuous air flow controller array. The outlets of the three stem chambers were also connected to the air flow controller array. The continuous flow system comprised six mass-flow controllers (FMA-A2103, 0–100 mL min⁻¹, Omega Engineering, Norwalk, CT, USA), a small vacuum pump, and a 6-port continuous-flow selector valve (EUTB-VLSF6MWE2, Valco-Vici, Houston, TX, USA). This system maintained a constant air flow of 100 mL min⁻¹ through three ambient air sample lines and three stem chambers, independent of the specific air sample being measured by the CRDS analyzers (Figure 2). The output of the continuous-flow valve was connected to two cavity ring-down spectrometers (CRDSs) and cycled between ambient air (10 min) and stem air (10 min). When analyzing three stem chambers sequentially, the valve cycle proceeded as follows: ambient air → stem air 1 → ambient air → stem air 2 → ambient air → stem air 3, completing one full measurement cycle every hour. This setup ensured continuous buffered ambient air flow through both the ambient reference and stem air sample gas lines.

Sample air on the output side of the gas selector valve was diverted to two CRDS systems including an O₂ and isotopic ¹⁸O₂ CRDS (G2207-i, Picarro Inc., Santa Clara, CA, USA) and isotopic ¹²CO₂ and ¹³CO₂ (G2101-i) at a total flow of 100 mL min⁻¹. The O₂ CRDS was placed in the high-precision O₂ concentration mode (instead of ¹⁸O isotope mode). Given the high sensitivity of ARQ to errors in either E_{S_CO2} or E_{S_O2} fluxes [16], calibration of both sensors was performed over the range of O₂ and CO₂ concentrations encountered in the dynamic stem chambers using the dilution method using two high-precision mass-flow controllers with low (0–10 mL min⁻¹) and high (0–1000 mL min⁻¹) flow ranges. For O₂ calibration, a high-purity air cylinder (21% or 210,000 ppm O₂, UHP 5.0, Linde Gas) flowing at 1000 mL min⁻¹ was diluted with 0.0–10.0 mL min⁻¹ of high-purity nitrogen (N2) to generate the following O₂ concentrations (210,000, 209,790, 209,581, 209,372, 209,163, 208,955, 208,748, 208,540, 208,333, 208,126, and 207,920 ppm O₂). ¹²CO₂ measurements were calibrated by dynamic dilution of pure CO₂ gas standard (0–10 mL min⁻¹) with ultra-high-purity CO₂-free air flowing at 1000 mL min⁻¹ to generate CO₂ concentrations of 0, 999, 1996, 2991, 3984, 4975, and 5964 ppm. Linear regression slopes of actual CO₂ and O₂ concentrations generated by dynamic dilution were plotted versus CRDS reported concentrations for CO₂ (Cal_CO₂ = 0.89447) and O₂ (Cal_O₂ = 1.0442).

For each measurement period of ambient and stem chamber air, the last three minutes of O₂ and CO₂ concentrations were averaged and the difference between ambient and stem concentrations was calculated, with calibration corrections, to estimate net stem CO₂ efflux (E_{S_CO2}, µmol m⁻² s⁻¹), O₂ influx (E_{S_O2}, µmol m⁻² s⁻¹), and the apparent respiratory quotient (ARQ) using the following equations, where F is the ambient air flow rate (100 mL min⁻¹ or 0.00167 L s⁻¹), ΔCO₂ and ΔO₂ are the concentration differences between the stem chamber and ambient air for each measurement period (1 ppm = 1 µL/L), 1 µmole/22.4 µL is a conversion factor based on the ideal gas law used to convert the volume (µL) to moles, and A_stem is the enclosed stem surface area (15.3 cm × 6.5 cm or 0.0099 m²). Note that no flow correction for humidity was performed due to the use of dried ambient air.

E_{S_CO2} = F × ΔCO₂ × Cal_CO₂ × (1 µmole/22.4 µL)/A_stem

(1)

E_{S_O2} = F × ΔO₂ × Cal_O₂ × (1 µmole/22.4 µL)/A_stem

(2)

ARQ = E_{S_CO2}/E_{S_O2}

(3)

Assuming a cellular RQ of 1.0 for complete oxidation of non-structural carbohydrates during aerobic mitochondrial respiration, E_{S_CO2} was partitioned into contributions from local respiratory production and transported CO₂ using the relationships in Equations (4) and (5).

E_{S_CO2} = E_{S_CO2_local} + E_{S_CO2_transport}

(4)

E_{S_CO2_local} = −E_{S_O2}

(5)

Diurnal patterns and temperature sensitivities of stem gas exchange variables were evaluated using time-series plots. Linear statistical analyses were performed to assess correlations between E_{S_CO2} and E_{S_O2}, allowing the determination of average stem ARQ values during each measurement period.

3. Results

3.1. Calibration of CO₂ and O₂ CRDS Sensors

The calibration of the CO₂ and O₂ cavity ring-down spectroscopy (CRDS) sensors demonstrated high linearity and precision. Figure 3a shows the relationship between actual and measured CO₂ concentrations. For CO₂, the linear regression analysis yielded a calibration slope (CO₂_actual/CO₂_measured) of 0.89447 ± 0.00178, with an R² value of 0.9999 across the range (0–5000 ppm). Similarly, the O₂ calibration (Figure 3b) also exhibited strong linearity, with a calibration slope (O₂_actual/O₂_measured) of 1.0442 ± 0.00689, and with an R² value of 0.9998, across the range (201,021–205,041 ppm O₂ or 20.1–20.5%). These results confirm that the CRDS sensors provide accurate and reliable measurements of CO₂ and O₂ concentration differences between ambient air and stem air, supporting their use in continuous gas exchange monitoring in stem respiration studies.

3.2. Diurnal Patterns of Raw Stem CO₂ and O₂ Concentrations

Figure 4 illustrates raw CO₂ and O₂ gas concentrations (ppm) of ambient air and stem air exiting three open path flow-through stem chambers installed on the cherry tree outside the analytical laboratory during a continuous 3-day measurement period from 28 June to 1 July 2024. As expected from active aerobic stem respiration in local stem tissues, ambient O₂ concentrations (indicated by the arrow on the top right) were higher in the ambient air (inlet) compared with all three stem chambers (outlet), with a maximum concentration difference of approximately 2430 ppm observed during the measurement period. In contrast, ambient CO₂ concentrations (indicated by the arrow on the bottom left) were lower than the air exiting all three stem chambers, with a maximum concentration difference of 4405 ppm. This observation suggests that more CO₂ is emitted than O₂ consumed within the stem chambers.

Figure 5 illustrates the diurnal patterns of stem CO₂ efflux (E_{S_CO2}), O₂ influx (E_{S_O2}), and ARQ calculated from the raw data across three stem chambers during the continuous 3-day period. Stem E_{S_CO2} effluxes and E_{S_O2} influxes exhibited clear diurnal cycles, with minimum values occurring in the early morning at around 7:00 when air temperature was the lowest. This was followed by an increase in these fluxes as temperature warmed in the early morning. Stem chamber 1 fluxes showed dramatic increases in both E_{S_CO2} and E_{S_O2} fluxes, with a peak between 9:00 and 10:00. E_{S_CO2} and E_{S_O2} fluxes generally stabilized or were slightly suppressed by noon, followed by a slight increase again as temperatures began to cool in the early evening, reaching a maximum around 19:00. ARQ values also showed distinct diurnal patterns, with a range between 0.9 and 1.5, and showed a general increase in all three stem chambers over the 2.5-day experiment. All three stem chambers tended to display elevated ARQ values at night, often reaching a maximum at 7:00 A.M, when air temperatures reached a minimum. ARQ values declined slightly during the daytime (Figure 5), potentially related to a differential effect of temperature, diurnal cycles of photosynthesis and transpiration, and metabolism of respiratory substrates (via phloem loading and transport) on CO₂ and O₂ fluxes.

Regression analysis of E_{S_CO2} versus E_{S_O2} flux data over the 3-day experiment was performed, revealing strong linear relationships for all three stem chambers (Figure 6). Average ARQ values obtained from the linear regression were 1.3, 1.6, and 1.7, and therefore remained well above 1.0 on average. When E_{S_CO2} was partitioned into E_{S_CO2_local} and E_{S_CO2_transport} (Figure 7), clear differences could be found for local versus transported respiratory CO₂ and their influences on net CO₂ efflux. While the contribution of local sources of stem respiration to stem E_{S_CO2} remained higher than transported sources of CO₂, the magnitude of E_{S_CO2_transport} in all three stem chambers showed an increase over the 3-day experiment, possible due to the trend of increasing temperatures. When the temperature sensitivity of E_{S_CO2_local} was evaluated via regression with air temperature (Figure 7d), an exponential pattern was not observed, as expected from a Q₁₀ model (see Section 4). Instead, the temperature sensitivity of E_{S_CO2} continuously decreased with temperature, such that E_{S_CO2} reached a plateau. These patterns underscore the strong coupling between CO₂ efflux and O₂ influx in stems and their dependence on diurnal physiological processes such as CO₂ transport in the transpiration stream, plant hydraulics like transpiration and cellular turgor pressure, and environmental conditions like temperature.

The measurements were repeated later in the growing season two times over longer measurement periods, but with only one stem chamber installed on the stem, allowing higher temporal resolution of the stem gas exchange fluxes (three sequential stem gas exchange observations per hour rather than one). While minimum E_{S_CO2} and E_{S_O2} fluxes were still observed in the early morning, associated with the lowest air temperatures, both CO₂ and O₂ fluxes (including local and transported components of E_{S_CO2}) increased with temperature, showing a maximum in the late afternoon (Figure 8). However, in contrast to the first experiment earlier in the growing season, ARQ showed a peak around noon, reaching values over 3.2, followed by a gradual decline throughout the afternoon. Average ARQ values determined from linear regression analysis showed a mean ARQ of 2.89 ± 0.05, again demonstrating values far exceeding unity (Figure 9a). Thus, stem net CO₂ efflux was dominated by transported CO₂ contributions. While E_{S_CO2_local} and E_{S_CO2_transport} showed decreased temperature sensitivity at temperatures above 20 °C, E_{S_CO2_transport} showed higher temperature sensitivity than E_{S_CO2_local} (Figure 9b). When this experiment was repeated towards the end of the growing season, but at a different location on the stem, net CO₂ efflux was again dominated by transported CO₂ contributions (Figure 10). ARQ generally showed a midday suppression and a maximum value of up to 3.2 during the night. Regression analysis showed an average ARQ of 2.45, and analysis versus air temperature generally showed E_{S_CO2_transport} fluxes higher than E_{S_CO2_local} (Figure 11). As in the other replicates, the apparent temperature sensitivity of the local and transported E_{S_CO2} stem effluxes decreased with temperatures above 20 °C, below which the apparent temperature sensitivity of E_{S_CO2_tranport} was higher than E_{S_CO2_local}.

4. Discussion

4.1. Advantages of Real-Time, Continuous Measurements of CO₂ Efflux

In this article, we present the first system capable of continuous observations of stem E_{S_CO2} and E_{S_O2} fluxes under ambient air conditions, allowing for real-time observations of ARQ, and potentially avoiding issues related to sealed static stem chambers. The apparent respiratory quotient (ARQ), the ratio of net CO₂ efflux to net O₂ influx, provides additional insights into respiratory and transport processes in stems and soils and has the potential to constrain CO₂ transport processes independently of local respiratory production. However, observations from static chamber studies consistently reported ARQ values often much lower than 1.0 (Table 1). This deviates from hypotheses predicting ARQ values greater than 1.0 due to CO₂ transport in the transpiration stream (Figure 1). Static chamber methods are prone to artifacts, and their inability to resolve real-time changes has limited the interpretation of ARQ observations from intact stems in the field. While processes such as re-fixation by PEPC and preferential utilization of different respiratory substrates have been discussed as potential causes of low ARQ values below theoretical expectations for aerobic respiration of carbohydrates (i.e., 1.0), the underestimation of ARQ by the static method should also be considered. In addition, the transport of CO₂ in the transpiration stream has also been discussed as a potential cause of an ARQ below 1.0. The ARQ could decline due to CO₂ transport if more CO₂ is carried away in the transpiration stream than is entering from below. This could then lead to the net CO₂ efflux measured at the stem surface being lower than the true respiratory production. This could occur under conditions of high transpiration rates or when the CO₂ source strength from lower stem segments, roots, and/or soils is weak, leading to a reduction in measured ARQ. In such cases, the discrepancy would indicate that a portion of the respired CO₂ is being internally transported rather than escaping to the atmosphere via diffusion. As a result, the ARQ derived from gas exchange in a stem segment may decline. Conversely, if there is a large respiratory CO₂ source from the soil, roots, and lower stem segments, more CO₂ may enter the segment than is transported away, leading to an increased ARQ above 1.0. In this case, the stem acts as a conduit for significant CO₂ movement from below, amplifying the measured efflux at the segment’s surface. This scenario might be more likely in trees with high soil, root, and stem respiration rates. This indeed appeared to be the case in the current study on a single California cherry tree, with maximum daytime E_{S_CO2} reaching high values between 6 and 25 µmol m⁻² s⁻¹. This is higher than many reports of stem CO₂ efflux, such as from juniper woodlands (up 5 µmol m⁻² s⁻¹) [46], Norway spruce (up to 2.3 µmol m⁻² s⁻¹) [32], loblolly pine (up to 3.3 at 25 °C) [28], and tropical trees (up to 2.1 µmol m⁻² s⁻¹) [33], and more comparable to stem E_{S_CO2} studies of Eastern cottonwood (up to 25.2 µmol m⁻² s⁻¹) [9], Arizona walnut (up to 22.0 µmol m⁻² s⁻¹) [9], Tulip tree (15.9 µmol m⁻² s⁻¹) [9], Douglas fir (up to 8.4 µmol m⁻² s⁻¹) [34] and Japanese cypress (up to 10 µmol m⁻² s⁻¹) [35]. In addition to variations in cellular respiration and growth rates, E_{S_CO2} is also potentially dependent on CO₂ conductance of bark to allow CO₂ diffusion to the atmosphere. Thus, presumably, we can expect strong differences in E_{S_CO2} fluxes between species due to differences in wood anatomy, including wood density, lenticel density, bark thickness, etc.

4.2. Significance of Observed Respiratory Quotient

In this study, continuous observations affirmed that there was an ARQ routinely above 1.0 on multiple time scales (sub-hour to week), as determined using two independently calibrated CRDS systems for CO₂ and O₂ concentrations, a 3D-printed stem chamber with ambient air continuously flowing through, and a mass-flow controller array and continuous flow valve system that maintains air flow through both ambient air and stem air sample gas lines at all times while cycling through each of the samples for online CO₂ and O₂ concentration determination by CRDS. Based on a tree growing outside the laboratory, we show for the first time that stem ARQ values greatly exceeding unity are common, potentially indicating that transported CO₂ contributions to E_{S_CO2} can dominate local respiratory contributions. However, it should be acknowledged that very high values of ARQ observed here at up to 3.2 may involve other processes such as aerobic and anaerobic fermentation in roots and stems, which could generate CO₂ with little O₂ consumption [36]. It should be acknowledged that a key limitation of our approach is the assumption that the cellular respiratory quotient (RQ) is 1.0, which corresponds to the complete oxidation of sugars. In reality, tree stem tissues may respire a variety of substrates that exhibit different RQ values: lipids typically yield an RQ around 0.7, proteins yield RQ values around 0.8, and organic acids often have RQ values in the range of 1.3 to 1.4. This variability implies that using a fixed RQ of 1.0 for partitioning E_{S_CO2} into local and transported components could lead to errors. If substrates with lower RQ values (such as lipids with an RQ around 0.7 or proteins around 0.8) are predominantly respired, then the actual local CO₂ production will be lower than what is assumed by an RQ of 1.0. In this scenario, our method would subtract too large a local component, leading to an underestimation of the transported CO₂. Conversely, if organic acids (with RQ values typically in the range of 1.3–1.4) dominate, local respiration produces more CO₂ than assumed. This extra CO₂, not accounted for in the assumed local production, would be erroneously attributed to transport, thus overestimating the contribution of transported CO₂. However, the primary substrates used for respiration in stems are carbohydrates, such as sucrose and starch, and phloem sap is mainly composed of sucrose, supporting the assumption of an RQ near 1.0 for carbohydrate oxidation. Although alternative substrates like proteins or lipids can be utilized under specific stress or developmental conditions, they typically represent a minor fraction of the respiratory substrate pool. This means that, under normal conditions, the assumption of an RQ of 1.0 may be reasonable for partitioning stem CO₂ fluxes between local respiration and transported components. Future studies could address this issue of variable substrate utilization (and therefore cellular RQ) by employing techniques such as stable isotope labeling, metabolomic profiling, and mass spectrometry analyses of phloem exudates. These methods would help quantify which substrates are being delivered via the phloem and used in respiration, thereby enabling a more accurate assessment of substrate-specific RQ values and improving the partitioning between local respiratory production and CO₂ transported via the transpiration stream.

4.3. Physiological Processes and Modeling of CO₂ Efflux

Utilizing the methods presented here, future research could explore the opportunities for collecting continuous vertical profiles of stem E_{S_CO2}, E_{S_O2}, and ARQ and couple these observations to biochemical (Table 2), hydraulic (soil moisture, sap velocity, stem water potential), stem diameter change (using point dendrometry), and temperature observations (air and stem). By integrating these measurements across the height of the stem, observations would enable E_{S_CO2} partitioning as a function of height as well as an overall partitioning of whole stem CO₂ efflux into above- versus below-ground contributions. These vertical gradients and diurnal stem-atmosphere gas exchange patterns could lead to more robust field data useful for model inter-comparisons, such as those incorporating detailed physiological processes, including hydraulic impacts, CO₂ transport, and environmental sensitivities. In addition, gas exchange measurements at the leaf level would provide valuable context, particularly for quantifying canopy re-assimilation and stem respiratory substrate supply. Future studies could also consider complementing real-time ARQ measurements with stable isotope techniques, such as ¹³CO₂ or ¹³C-labeled substrate tracing, to directly quantify the sources, transport, and fate of respiratory CO₂ within the plant. Isotopic labeling would provide a powerful means of validating the partitioning of CO₂ between local emission, xylem transport, and potential re-assimilation by canopy tissues, thereby strengthening the mechanistic interpretation of ARQ dynamics.

Interestingly, in a series of greenhouse experiments, Salomón and colleagues highlighted the phenomenon of daytime depression in temperature-normalized stem CO₂ efflux in young poplar trees, linking it to suppressed mitochondrial respiration under hydraulic stress and CO₂ transport in the transpiration stream [37]. Their study demonstrated the difficulty in separating these processes due to their simultaneous influence by transpiration dynamics. Experimental manipulation and mechanistic modeling indicated that reductions in turgor pressure rather than CO₂ transport dominated the observed depression, underscoring the need for novel approaches like dynamic stem ARQ measurements to help quantify these components. Consistent with these observations, a recent study investigating the effects of reduced water availability on stem CO₂ efflux in oak (Quercus petraea) and hornbeam (Carpinus betulus) coppices. They found that severe summer drought led to a significant decrease in stem CO₂ efflux, with reductions ranging from 43% to 81% during July and August [53]. This decline was closely associated with a concurrent reduction in stem growth, suggesting that drought-induced limitations on growth processes were a primary factor in the observed decrease in stem respiration [53]. While the study did not directly measure turgor pressure, the authors discussed how the suppression of stem CO₂ efflux under drought conditions could be attributed to hydraulic constraints affecting cell expansion and metabolic activity, potentially linked to reduced turgor pressure. These findings highlight the complex interplay between water availability, stem growth, and respiration, emphasizing the sensitivity of stem CO₂ efflux to hydraulic stress.

Daytime suppression of cellular growth and aerobic respiration associated with a drop in turgor pressure and high transpiration rates is expected to result in a measurable decrease in the temperature sensitivity of local tissue respiration determined from E_{S_O2} measurements. In contrast, if transport of respiratory CO₂ in the transpiration stream is the main cause of the daytime temperature-normalized E_{S_CO2} suppression, then stem E_{S_O2} fluxes would be expected to increase with temperature exponentially, as predicted from Q₁₀ modeling approaches [38]. However, our results suggest that the temperature sensitivity of local stem respiratory processes is suppressed above 20 °C (e.g., Figure 7d, Figure 9b and Figure 11b), potentially due to reductions in turgor pressure constraining growth respiration [37]. This is consistent with the lower apparent temperature sensitivity of E_{S_CO2_local} versus E_{S_CO2_transport} (e.g., Figure 9b and Figure 11b) as well as studies demonstrating that the respiratory Q₁₀ of plant tissues is not constant, but often decreases with temperature [39]. Nonetheless, the results suggest that elevated E_{S_CO2_transport} fluxes at high temperatures are driven by both higher respiration and transpiration rates.

In addition to the more biophysical model TreSpire, stem CO₂ efflux measurements that distinguish between localized and transported respiration would be invaluable for benchmarking and calibrating terrestrial biosphere models (TBMs). Some of these models, such as the Functionally Assembled Terrestrial Ecosystem Simulator (FATES) [40,41,42], estimate respiration rates of independent plant organs including stems, coarse roots, fine-roots, and leaves, typically using relationships that scale a base respiration rate (defined by total nitrogen content of the tissues) by a temperature response function. While some TEBs may represent mechanisms such as dissolved CO₂ transport and re-fixation, these processes are typically not represented in TEBs used in earth system modeling applications. As such, TEBs such as FATES could benefit from representing these aspects of the carbon balance in plants.

4.4. Consequences for Molecular Signaling Pathways

Our method to assess CO₂ efflux partitioning (stem respiration vs. xylem sap transport) will be extremely useful to monitor root-to-shoot signaling via xylem sap composition. In effect, the xylem sap bicarbonate concentration and thus the propensity to liberate CO₂ are directly related to sap pH [43,44], which in turn plays a role in water relations. It has been shown that under water deficit or a high soil CO₂ concentration, there is a change in sap pH [47,48,49]. This stimulates starch degradation via the trehalose 6-phosphate signaling pathway (for a recent study, see Gao et al. 2022 [54]), thereby facilitating acidic invertase activity and soluble sugar accumulation to adjust the apoplastic osmotic potential and facilitate embolism repair [55]. Also, the sap pH affects abscisic acid (ABA) signaling via tissular compartmentalization [50]. In the field, it has also been shown that there is a co-variation between xylem sap pH and ABA content, leading to stomatal closure during the dry season in evergreen and semi-deciduous woody species [51]. It is also worth noting that a low pH and bicarbonate concentration have an effect on phenolic peroxidases and Zn-Cu superoxide dismutase (SOD), two enzymes that have been found repeatedly in the xylem sap proteome [56]. In fact, phenolic peroxidases have higher activity at a low pH [57], while bicarbonate enhances the peroxidase activity of SOD [58].

In other words, there is likely an interaction between CO₂ efflux and drought signaling pathways involving reactive oxygen species (ROS). CO₂ efflux from non-photosynthetic plant tissues—such as stems, roots, and xylem parenchyma—is likely intricately linked to salt, temperature, and drought stress responses through ROS signaling pathways [59]. Under these abiotic stresses, ROS accumulate in organelles like mitochondria and peroxisomes, modulating respiration by altering mitochondrial electron transport and activating alternative oxidase pathways. This directly affects the rate and pattern of CO₂ release from these tissues [60]. As a result, CO₂ efflux may reflect both local respiratory activity and broader oxidative stress signaling under challenging environmental conditions. Collectively, monitoring ARQ and thus estimates of xylem sap transport of CO₂ using our method might be used to follow root-to-shoot molecular signaling processes. Given that stem CO₂ efflux and O₂ influx measurements provide real-time, integrative indicators of respiratory activity in woody tissues, mechanistic interpretation in future studies will be enhanced through targeted biochemical assays. Among the most directly comparable methods are mitochondrial O₂ consumption assays using Clark-type electrodes, isothermal calorimetry to quantify total respiratory heat output, and enzymatic activity measurements of the alternative oxidase (AOX) and cytochrome oxidase (COX) pathways. These methods, along with metabolic profiling of sugars, organic acids, and amino acids, may enable a mechanistic interpretation of gas exchange patterns in relation to respiratory substrate use and pathway activity under varying environmental conditions. These comparisons, summarized in Table 2, are expected to enable a systems-level understanding of how respiration responds to environmental conditions such as drought, high temperature, and salinity.

Table 2. Biochemical measurements that can be quantitatively compared with in vivo stem CO₂ efflux and O₂ influx. The table lists recommended assays for assessing mitochondrial respiration, substrate availability, and redox status, along with the required instruments, measured outputs, and how each measurement links mechanistically and quantitatively to observed gas fluxes. These biochemical metrics provide a framework for integrating physiological gas exchange data with underlying metabolic processes under varying environmental conditions.

Measurement	Instrument Needed	What It Reflects	Units/Output	Quantitative Link to Gas Fluxes
Mitochondrial O2 Consumption Rate [61]	Clark-type O2 electrode or Seahorse XF Analyzer	Total respiratory activity (cytochrome + AOX pathways)	µmol O2 mg−1 protein h−1 or µmol O2 g−1 FW h−1	Directly comparable to stem O2 influx (Es_O2)
CO2 Production in Isolated Tissues or Mitochondria [62]	Infrared gas analyzer (IRGA), GC, or CRDS	Net decarboxylation rate from respiration	µmol CO2 g−1 FW h−1	Directly comparable to stem CO2 efflux (Es_CO2)
AOX and COX Enzyme Activities [63]	Spectrophotometer or O2 electrode with inhibitors (e.g., SHAM, KCN)	Partitioning of O2 consumption across ETC pathways	nmol O2 min−1 mg−1 protein	Explains variation in RQ and O2 uptake efficiency
TCA Cycle Enzyme Activities [64]	Enzymatic activity assay kits with spectrophotometry	Respiratory flux capacity	nmol min−1 mg−1 protein	High activity aligns with elevated CO2 production
ROS Production Rates (H2O2, O2−) [65]	Fluorometric ROS assays (e.g., Amplex Red, MitoSOX) or microplate reader	Redox state and mitochondrial efficiency	nmol g−1 FW h−1	High ROS can suppress respiration and may influence RQ
Antioxidant Enzyme Activities (SOD, CAT, APX) [66]	Spectrophotometer or plate reader	ROS detoxification capacity	nmol min−1 mg−1 protein	Reflects ability to maintain respiration under oxidative stress
Metabolic Profiling: Sugars, Organic Acids, and Amino Acids [67]	GC-MS, LC-MS, or HPLC	Primary carbon substrates	µmol g−1 FW	Correlate with respiratory fluxes and CO2 efflux with composition shifts potentially impacting RQ
ATP/ADP and NADH/NAD⁺ Ratios [68]	Bioluminescence assay kits, HPLC, or LC-MS	Energetic/redox status	Molar ratios	Low ratios often coincide with reduced gas exchange
Calorimetry of Stem Segments [69]	Isothermal microcalorimeter (e.g., TAM (Thermal Activity Monitor))	Total respiratory heat production from all pathways	µW g−1 FW or J h−1	Quantitatively related to O2 consumption (∼470 kJ/mol O2) and CO2 production

Table 2. Biochemical measurements that can be quantitatively compared with in vivo stem CO₂ efflux and O₂ influx. The table lists recommended assays for assessing mitochondrial respiration, substrate availability, and redox status, along with the required instruments, measured outputs, and how each measurement links mechanistically and quantitatively to observed gas fluxes. These biochemical metrics provide a framework for integrating physiological gas exchange data with underlying metabolic processes under varying environmental conditions.

Measurement	Instrument Needed	What It Reflects	Units/Output	Quantitative Link to Gas Fluxes
Mitochondrial O2 Consumption Rate [61]	Clark-type O2 electrode or Seahorse XF Analyzer	Total respiratory activity (cytochrome + AOX pathways)	µmol O2 mg−1 protein h−1 or µmol O2 g−1 FW h−1	Directly comparable to stem O2 influx (Es_O2)
CO2 Production in Isolated Tissues or Mitochondria [62]	Infrared gas analyzer (IRGA), GC, or CRDS	Net decarboxylation rate from respiration	µmol CO2 g−1 FW h−1	Directly comparable to stem CO2 efflux (Es_CO2)
AOX and COX Enzyme Activities [63]	Spectrophotometer or O2 electrode with inhibitors (e.g., SHAM, KCN)	Partitioning of O2 consumption across ETC pathways	nmol O2 min−1 mg−1 protein	Explains variation in RQ and O2 uptake efficiency
TCA Cycle Enzyme Activities [64]	Enzymatic activity assay kits with spectrophotometry	Respiratory flux capacity	nmol min−1 mg−1 protein	High activity aligns with elevated CO2 production
ROS Production Rates (H2O2, O2−) [65]	Fluorometric ROS assays (e.g., Amplex Red, MitoSOX) or microplate reader	Redox state and mitochondrial efficiency	nmol g−1 FW h−1	High ROS can suppress respiration and may influence RQ
Antioxidant Enzyme Activities (SOD, CAT, APX) [66]	Spectrophotometer or plate reader	ROS detoxification capacity	nmol min−1 mg−1 protein	Reflects ability to maintain respiration under oxidative stress
Metabolic Profiling: Sugars, Organic Acids, and Amino Acids [67]	GC-MS, LC-MS, or HPLC	Primary carbon substrates	µmol g−1 FW	Correlate with respiratory fluxes and CO2 efflux with composition shifts potentially impacting RQ
ATP/ADP and NADH/NAD⁺ Ratios [68]	Bioluminescence assay kits, HPLC, or LC-MS	Energetic/redox status	Molar ratios	Low ratios often coincide with reduced gas exchange
Calorimetry of Stem Segments [69]	Isothermal microcalorimeter (e.g., TAM (Thermal Activity Monitor))	Total respiratory heat production from all pathways	µW g−1 FW or J h−1	Quantitatively related to O2 consumption (∼470 kJ/mol O2) and CO2 production

5. Conclusions

This study provides key insights into the dynamic partitioning of stem-respired CO₂ between local emission and upward transport via the transpiration stream, with important implications for understanding whole-tree carbon cycling and improving terrestrial carbon flux models. The observation of ARQ values consistently greater than 1.0 demonstrates that transported CO₂ can be a significant component and even dominate stem efflux, challenging the common assumption that stem CO₂ emissions primarily reflect local respiration and re-assimilation. Diurnal patterns revealed that high temperatures can suppress ARQ, likely due to an increase in local respiration relative to transported CO₂, whereas midday ARQ peaks suggest enhanced CO₂ transport during periods of high transpiration. Notably, the apparent temperature sensitivity of transported CO₂ was greater than that of local mitochondrial respiration, indicating that transpiration-driven transport is more responsive to temperature fluctuations. These findings highlight the role of hydraulic processes in modulating respiratory activity and stem CO₂ transport dynamics, and they suggest that transported CO₂, if re-assimilated by canopy leaves, may help buffer photosynthesis during periods of reduced stomatal conductance, such as warm afternoons with a high vapor pressure deficit. While the results are based on detailed measurements from a single tree near the laboratory, this study introduces a novel, real-time method for quantifying the apparent respiratory quotient (ARQ). When coupled with biochemical, hydraulic, and environmental sensors, this approach can be extended to larger spatial scales and diverse ecosystems to improve our mechanistic understanding of the coupling between plant carbon cycling and water use. Integrating ARQ measurements into carbon and water flux models offers a new framework for capturing the physiological complexity of plant–atmosphere exchange. Further research across forest types and climates will be essential to evaluate the generality of these findings and refine predictions of the global carbon budget.