Excision increases branch water potential but generally reduces leaf gas exchange

Abstract

Measurement of leaf gas exchange is often complicated by the difficulty of sampling leaves from tall or otherwise inaccessible plants. Branch excision is widely used to enable such measurements, but the magnitude, variability and mechanisms of its effects on water potential and gas exchange are not well characterized across species and environments. We quantified the effects of excision on branch water potential, stomatal conductance, photosynthesis and stomatal behavior (parameter g₁) in four angiosperm and one gymnosperm tree species differing in hydraulic traits and water-use strategies. To assess generality, we also conducted a meta-analysis of 26 species spanning a broad range of hydraulic traits, water-use strategies and biomes. Excision consistently increased branch water potential in all species. In angiosperms, excision reduced stomatal conductance and photosynthesis, with decreases in g₁ for some species, whereas in the gymnosperm, excision increased stomatal conductance, photosynthesis and g₁. The meta-analysis showed that excision generally decreased stomatal conductance and photosynthesis, with effect sizes varying by species, water-use strategy and hydraulic traits. These results demonstrate that the effects of excision are strongly species- and trait-dependent. Careful consideration and study-specific corrections are needed when interpreting gas exchange data from excised branches.

Introduction

Photosynthesis fuels the biosphere and is the largest single flux in the global carbon cycle (IPCC 2023). It plays a central role in plant–atmosphere carbon exchange, regulating atmospheric CO₂ concentrations and global climate. Accurate measurement of leaf gas exchange is therefore essential for advancing ecophysiological theory, parameterizing Earth system models and informing projections of plant responses to climate change (Medlyn et al. 2002, Wright et al. 2004, Kattge and Knorr 2007, Blonder et al. 2023). Portable gas exchange analyzers have greatly facilitated these efforts, enabling rapid, high-resolution measurements under both laboratory and field conditions (Garen et al. 2022, Garen and Michaletz 2024, 2025). Nevertheless, use of these instruments still faces methodological challenges, particularly in sampling leaves that are difficult to access in situ.

A common solution is branch excision, followed by recutting under water and transportation for subsequent measurement (e.g. Michaletz et al. 2016). This approach enables access to otherwise inaccessible branches and increases experimental throughput, but it also disrupts the plant hydraulic continuum (Ewers and Fisher 1989, Joesting et al. 2009, Domingues et al. 2010). Although branch excision is widely used across taxa and biomes (Perry 1978, Ewers and Fisher 1989, Monson et al. 2007, Messier et al. 2010, Dusenge et al. 2015, Akalusi et al., 2021), understanding of the magnitude, variability and mechanisms of its effects on water potential and gas exchange across species and environments is limited. Previous comparative studies have been restricted to limited species and environments, most notably wet tropical forests where water availability is seldom limiting (Lakso 1982, Meng and Arp 1993, Santiago and Mulkey 2003, Miyazawa et al. 2011, Verryckt et al. 2020, Kar et al. 2021, Missik et al. 2021). As a result, fundamental questions remain regarding the mechanisms by which excision affects gas exchange and the extent to which these effects are mediated by species’ hydraulic traits (Simovic and Michaletz 2025).

Stomatal conductance is a key regulator of gas exchange and varies dynamically with plant water status (Cowan and Farquhar 1977, Medlyn et al. 2011, Blonder et al. 2023). According to the cohesion-tension theory (Dixon and Joly 1895), transpiration generates surface tension at leaf surfaces which is transmitted by cohesion, adhesion and tension through a continuous water column, effectively pulling water from the soil into the roots, and through the xylem to the leaves (Tyree and Zimmermann 2002). As water availability declines, increasing sap tension reduces xylem water potential, inducing stomatal closure and restricting water loss and carbon assimilation (Tyree and Zimmermann 2002, Kirkham 2005). Manipulation that alters the hydraulic continuum between leaf and soil can therefore alter water potential, stomatal conductance and gas exchange (Lakso 1982, Santiago and Mulkey 2003, Garen et al. 2022). Relieving tension via excision is expected to increase sap water potential and increase stomatal conductance and gas exchange. However, incomplete rehydration, xylem occlusion or anatomical damage may constrain this response, leading instead to reduced stomatal conductance and assimilation.

The unified stomatal optimization (USO) model (Medlyn et al. 2011) provides a quantitative framework for analyzing how stomatal conductance responds to hydraulic changes. The model assumes that stomatal behavior maximizes carbon gain per unit water loss and is expressed as

(1)

where g_s (mol m⁻² s⁻¹) is stomatal conductance, g₀ (mol m⁻² s⁻¹) is an empirical intercept representing baseline g_s, the factor 1.6 is the ratio of the diffusivities of water vapor to CO₂ in air, A (μmol m⁻² s⁻¹) is net assimilation rate, C_a (μmol mol⁻¹) is atmospheric CO₂ concentration, g₁ (kPa^0.5) quantifies the marginal water cost of carbon gain and D (kPa) is leaf-to-air vapor pressure deficit. When D is held constant, Eq. (1) can be linearized by defining , yielding

(2)

In this linearized form, the term is the USO model index (Wu et al. 2020), and g₁ can be estimated as . Lower g₁ values indicate higher intrinsic water-use efficiency (iWUE) and more conservative stomatal regulation, whereas higher values indicate lower iWUE and less conservative regulation (Medlyn et al. 2011, Lin et al. 2015, Wu et al. 2020). Importantly, within the USO framework g₁ reflects the marginal water cost of carbon gain under optimal stomatal behavior. Branch excision may therefore influence g₁ if hydraulic disruption changes the effective cost of water loss relative to carbon assimilation, for example by constraining water supply independently of photosynthetic demand or by altering hydraulic constraints on transpiration. In contrast, if excision reduces stomatal conductance and photosynthesis proportionally, g₁ would not change. Thus, g₁ provides a standardized index for assessing whether excision primarily scales gas exchange rates or shifts stomatal regulation among species differing in water-use strategy and hydraulic traits.

The magnitude and direction of excision effects are likely to be influenced by plant hydraulic traits. Species span a continuum of water-use strategies from isohydry, characterized by conservative stomatal regulation that maintains relatively stable leaf water potentials, to anisohydry, in which stomatal conductance and leaf water potential vary more freely (Klein 2014, Martínez-Vilalta and Garcia-Forner 2017, Bryant et al. 2022). Additional traits, such as xylem type (tracheids or vessels) and exudate type (resin or latex), may further determine a branch’s capacity to recover or maintain hydraulic function following excision (Santiago and Mulkey 2003, Miyazawa et al. 2011, Verryckt et al. 2020, Kar et al. 2021, Missik et al. 2021, Davidson et al. 2022). For example, xylem type influences hydraulic resistance and the mechanical integrity of conduits at the cut surface. Conduit walls in tracheid-based xylem typically have greater thickness-to-span ratios and may therefore be less susceptible to shearing and conduit wall deformation during excision, preserving hydraulic conductivity relative to vessel-based xylem (Simovic and Michaletz 2025). Likewise, exudates may physically impede water uptake at the cut surface (Santiago and Mulkey 2003), potentially constraining hydraulic conductivity and gas exchange following excision.

However, previous studies have largely focused on one or a small number of species, often within a single functional group or environment, limiting our ability to assess how excision effects vary across species differing in hydraulic traits and water-use strategies. In this study, we address these gaps by (i) quantifying effects of excision on measurements of branch water potential and leaf gas exchange across five tree species differing in hydraulic traits and water-use strategies, and (ii) assessing the generality of excision effects using a meta-analysis of 26 species spanning a broad range of taxa, hydraulic traits and biomes. Hydraulic theory predicts that excision should release sap tension and increase branch water potential, which may increase stomatal conductance and photosynthesis (Sperry et al. 2002, Brodribb and Holbrook 2003). However, we hypothesized that the effects of excision on gas exchange would be strongly mediated by species’ hydraulic traits (including xylem type, water-use strategy and exudate type), such that the responses will vary across species. In particular, we expected that traits such as the presence of exudates or xylem conduit geometry (e.g. low wall thickness-to-span ratios in large conduits) may limit full recovery following excision by occluding conduits or promoting shearing and deformation at the cut surface, thereby constraining hydraulic conductivity, reducing stomatal conductance and photosynthesis, and altering stomatal behavior (g₁).

Materials and methods

Plant material and study sites

We studied five focal tree species representing variation in xylem and hydraulic traits relevant to excision effects on hydraulic and photosynthetic functioning (Table 1). Measurements were conducted on sunny days from 1 June 2023 to 15 August 2023 at three sites in Vancouver, British Columbia. All sites are located at ca 100 m elevation and within 13 km of each other and thus share a common temperate maritime macroclimate. Mean annual temperature is 10.4 °C and mean annual precipitation is 1189 mm based on 1981–2010 climate normals (Environment and Natural Resources Canada; https://climate.weather.gc.ca/). Soils at the study sites are generally representative of upland Vancouver soils developed on glacial deposits, predominantly Podzolic (e.g. Humo-Ferric Podzols) with coarse-textured, gravelly sandy loam to loamy sand and moderate to good drainage.

Table 1.

Study species and traits associated with hydraulic functioning and stomatal conductance following excision.

Species	Clade	Xylem type	Water-use strategy	Exudate type
Acer campestre	Angiosperm	Diffuse-porous	Anisohydric	Latex
Betula papyrifera	Angiosperm	Diffuse-porous	Isohydric	None
Carpinus betulus	Angiosperm	Diffuse-porous	Anisohydric	None
Pseudotsuga menziesii	Gymnosperm	Tracheids	Isohydric	Resin
Quercus garryana	Angiosperm	Ring-porous	Anisohydric	None

Betula papyrifera Marshall and Quercus garryana Douglas ex Hook. were located at The University of British Columbia Botanical Garden in Vancouver, BC, Canada (49.253778°, −123.247775°). Acer campestre L. (49.258197°, −123.250518°) and Pseudotsuga menziesii (Mirb.) Franco (49.258444°, −123.246540°) were located on the university campus. Carpinus betulus L. was located at a nearby site in Vancouver (49.2696365°, −123.0727520°). Species were categorized by clade, xylem type, water-use strategy and exudate type based on data reported in the literature (Kattge 2012, Meinzer et al. 2016, Martínez-Vilalta and Garcia-Forner 2017). All study trees were unirrigated.

Sampled branches were located in sun-exposed, lower canopy regions to control for light exposure and sap tension. Branches were excised at lengths of 1–2 m, ensuring they exceeded twice the maximum vessel length of each species. Maximum vessel length was measured using the air filtration technique (Cochard et al. 2010, Michaletz et al. 2012, Missik et al. 2021), yielding mean maximum vessel lengths of 36.34 cm, 31.79 cm, 17.86 cm and 44.46 cm for B. papyrifera, A. campestre, C. betulus and Q. garryana, respectively. Tracheid length for Pseudotsuga menziesii was not measured, as tracheids are much shorter than vessels (maximum lengths of ca 1–3 cm, Koshy 1993) such that excised branches were always much longer than the longest tracheids. For all branches, the youngest undamaged, fully expanded leaf was selected for measurement.

Water potential and gas exchange measurements

We measured diurnal curves of branch water potential and leaf gas exchange for paired intact and excised branches. For each curve, we sampled eight branches at 1-h intervals from 08:30 to 16:00 h (Figure 1). For each species, we measured diurnal curves on five individual trees, with intact and excised curves obtained from the same tree on each day, resulting in 10 curves per species. Each measurement sequence began with intact branches. After collecting data for intact branches, we excised the same branches and eliminated introduced embolisms by re-cutting under water to remove a segment longer than the maximum vessel length. Excised branches were then allowed a 20-min acclimation period before measurements. The average daily air temperature during sampling was 19.9 °C, as recorded with a 32-gauge type K chromel-alumel thermocouple and an HH806AW digital thermometer (Omega Sensing Solutions ULC, St-Eustache QC, Canada). Total annual precipitation in 2023 was 875 mm, as reported by an Environment Canada Vancouver International Airport weather station (49.1947°, −123.1839°) located ~8 km from the study sites.

Figure 1.

A schematic diagram illustrating the experimental methods applied to each branch. (A) A leaf on an intact branch was clamped in the chamber of a LI-COR LI-6800 and allowed to equilibrate for 20 min. During that time, (B) a different leaf was removed from the intact branch, and xylem water potential was measured with a pressure chamber. After 20 min, (C) gas exchange variables were recorded and (D) the branch was excised while the leaf was still in the chamber. (E) The excised branch was then re-cut under water and allowed to equilibrate for 20 min before recording (F) gas exchange variables and (G) water potential. These methods were used to measure mean (H) photosynthesis, (I) water potential and (J) stomatal conductance for eight branches per day. Data are shown for Quercus garryana (data for all species are provided in Figure S1 available as Supplementary Data at Tree Physiology Online). Points represent means of five individual trees, lines are ordinary least squares regressions and shaded areas are 95% confidence bands.

Water potential measurements were taken with a SoilMoisture Plant Water Status Console (Model 3115, SoilMoisture Equipment Corp., Goleta, CA, USA). Distal branch tips were excised and placed into the pressure chamber, with cut ends protruding through a compression fitting in the chamber. The chamber was pressurized slowly until sap first appeared at the cut end, and the balancing pressure (sap tension) was recorded. Measurements were repeated for each intact branch, and for each excised branch following the 20-min acclimation period.

Gas exchange measurements were taken with one of three LI-COR LI-6800 Portable Photosynthesis Systems (LI-COR Biosciences, Lincoln, NE, USA). Measurements occurred on sunny days on sun-exposed leaves, ensuring acclimation to high ambient irradiance prior to insertion in the LI-COR chamber. Gas exchange data were recorded 20 min after securing the leaf in the chamber for intact branches, and 20 min after re-cutting under water for excised branches. Chamber conditions were set to a reference CO₂ concentration of 400 p.p.m., a temperature block (Txchg) of 25 °C, and reference humidity of 50%, resulting in a constant D (1.59 kPa) during all measurements. The photosynthetically active radiation (PAR) was set to saturating levels for each species, defined as 80% of the asymptotic carbon assimilation rate estimated by fitting a photosynthesis-light response model (Marshall and Biscoe 1980) to light response curves measured on eight branches per individual. This yielded the following mean saturating PAR values: B. papyrifera, 800 μmol m⁻² s⁻¹; C. betulus, 500 μmol m⁻² s⁻¹; A. campestre, 2000 μmol m⁻² s⁻¹; Pseudotsuga menziesii, 1000 μmol m⁻² s⁻¹; Q. garryana, 2200 μmol m⁻² s⁻¹. All branches remained secured in the LI-COR chamber throughout the experiment, ensuring the same leaf area was measured during both intact and excised periods. All broadleaves completely filled the chamber area, while for Pseudotsuga menziesii, gas exchange data were area-corrected using leaf area measurements obtained for each branch with a Cannon CanoScan LiDE 400 and the LeafArea package in R (Katabuchi 2019). Any gas exchange measurements not meeting the stability criteria defined by LI-COR (Table S2 available as Supplementary Data at Tree Physiology Online) were excluded from analyses; in total, 174 of 400 measurements (43.5%) were excluded for not meeting stability criteria.

Data compilation

To quantify the general effects of excision on stomatal conductance and photosynthesis rates across diverse plant taxa, we combined data from our five focal species with published data from the literature. In total, our compiled dataset comprises 26 species from two biome types (tropical rainforest and temperate seasonal forest) across three continents (North America, South America and Asia), spanning broad latitudinal climate gradients. Stomatal conductance and photosynthetic rate were included as response variables, and species were categorized by clade, xylem type, water-use strategy and exudate type (Table S1 available as Supplementary Data at Tree Physiology Online).

Studies were included if they reported quantitative measurements for both excised and intact branches under ambient or saturating light conditions and provided extractable data (e.g. in tables, figures or supplementary files). To ensure methodological comparability among studies, we excluded those that did not recut branches under water following excision, as well as those that did not report measurements under defined light conditions. Several otherwise relevant studies were therefore not included because they did not meet these criteria or did not report data in a suitable form.

Statistical analyses

To characterize the effects of excision on water potential and gas exchange variables for our five focal species, we fitted separate mixed-effects models for each response variable (water potential, stomatal conductance and photosynthesis rate). Tree and branch identities were included as nested random effects to account for repeated measurements on individual trees and paired (intact and excised) measurements on branches throughout the day. Fixed effects included air temperature, measurement height and an interaction term between treatment (intact and excised) and species.

To evaluate differences in iWUE between intact and excised branches, we fitted Eq. (2) to our data and obtained the slope parameter m (Wu et al. 2020), from which we estimated g₁ as . We used ANOVA to compare differences in g₁ of intact and excised branches, and species-wise comparisons of g₁ were performed using paired t-tests.

In order to characterize the general effects of excision on stomatal conductance and photosynthesis rates across studies in the meta-analysis, we calculated the mean, coefficient of variation (CV) and sample size for gas exchange measurements of intact and excised branches in our compiled dataset (Table S1 available as Supplementary Data at Tree Physiology Online). To quantify the impact of excision, we used a log-transformed response ratio, R (dimensionless), which compares the means of excised and intact branches as (Hedges et al. 1999)

(3)

where and are the mean values for excised and intact branches, respectively. To ensure robustness, we also calculated an alternative log-transformed response ratio, R₂ (dimensionless), which adjusts for sample size and error according to (Nakagawa et al. 2023):

(4)

where n is the sample size (paired in all cases). Comparing R and R₂ enabled us to confirm consistency across approaches, which was particularly important because standard deviations were not available for all studies included in the meta-analysis. A paired t-test between R and R₂ showed no significant difference (P = 0.354), so Eq. (3) was used for all reported meta-analysis results to allow inclusion of studies that did not report standard deviation.

All statistical analyses and plots were completed using the R computing software version 4.4.3 (R Core Team 2025) with the following packages: ggplot2 (Wickham 2016), lme4 (Bates et al. 2015), lmerTest (Kuznetsova et al. 2017) and emmeans (Lenth et al. 2026). The data and code for reproducing all analyses and plots are available at https://github.com/MichaletzLab/Excision-Photosynthesis-Methods.

Results

Branch excision increased xylem water potentials across all five focal species. Mean water potentials increased significantly from ca −1.4 MPa in intact branches to −0.56 MPa in excised branches (P < 0.0001; Figure 2). Water potentials of intact branches declined with marginal significance throughout the day (P = 0.056), whereas excised branches exhibited no significant diurnal variation (P = 0.26; Figure 2). For each species, branch excision caused significant increases in mean water potential (all P < 7.53 × 10⁻⁶; Figure 3A). Among the species, Q. garryana exhibited the largest increase in water potential, rising from −1.72 to −0.48 MPa, while Pseudotsuga menziesii showed the smallest increase, from −1.49 to −0.97 MPa.

Figure 2.

Diurnal measurements of xylem water potential for intact and excised branches across five focal tree species (Acer campestre, Carpinus betulus, Betula papyrifera, Pseudotsuga menziesii and Quercus garryana). Points represent mean values from five individuals per species, error bars represent standard error and shaded areas are 95% confidence intervals.

Figure 3.

Relationships between water potential, stomatal conductance and photosynthetic rates measured on intact and excised branches for five focal tree species. Points represent individual branches, and shading indicates 95% confidence bands. The dashed 1:1 line shows where intact and excised measurements are equal; points above the 1:1 line indicate higher values for excised branches, while points below indicate higher values for intact branches.

Branch excision generally reduced leaf gas exchange across the focal species. Excision caused significant decreases in mean stomatal conductance for A. campestre (P = 0.0002), C. betulus (P = 0.0004) and Q. garryana (P < 0.0001; Figure 3B) and mean photosynthesis rates (all P < 0.0022; Figure 3C). Across species, excision reduced stomatal conductance and photosynthesis by up to several-fold in species with high intact gas exchange (Figure 3B and C). In contrast, B. papyrifera had non-significant decreases in mean stomatal conductance (P = 0.35). Pseudotsuga menziesii exhibited significant increases in mean stomatal conductance (P = 0.016) and mean photosynthesis rate (P = 0.021) following excision (Figure 3B and C). Across all five species, excision constrained gas exchange variables to low rates. This effect was particularly pronounced in species with high intact values, as indicated by regression slopes <1 for all species; this reduction was significant for stomatal conductance in A. campestre, B. papyrifera and Q. garryana (all P < 0.0001; Figure 3B), and for photosynthesis in all species (all P < 0.004; Figure 3C). At lower intact gas exchange values (below ca 0.075 mol m⁻² s⁻¹ for stomatal conductance and 7.5 μmol m⁻² s⁻¹ for photosynthesis), excised values were sometimes higher or lower than intact values (Figure 3B and C). However, at higher intact gas exchange values, excised values were consistently lower, highlighting the limitation on gas exchange imposed by excision.

Mixed-effects models provided further confirmation of these results and highlighted species-specific differences in response to excision (Figure 4). Water potential increased significantly after excision in all species (P << 0.001; Figure 4A). For stomatal conductance, excision caused significant reductions in A. campestre (P = 0.0001) and Q. garryana (P < 0.0001), while C. betulus experienced a marginally significant decrease (P = 0.070) and B. papyrifera experienced a non-significant reduction (P = 0.082). Pseudotsuga menziesii exhibited a non-significant increase (P = 0.099; Figure 4B). Photosynthesis rates showed similar responses, with significant reductions observed for A. campestre (P < 0.0001), B. papyrifera (P = 0.009) and Q. garryana (P = 4.80 × 10⁻¹⁶), whereas C. betulus had a marginally significant decrease (P = 0.064). Pseudotsuga menziesii exhibited a significant increase in photosynthesis (P = 0.019; Figure 4C).

Figure 4.

Mixed-effects model results illustrating the effect of excision on (A) water potential, (B) stomatal conductance and (C) photosynthetic rates for five focal tree species. Negative effect sizes indicate that the response variable was higher for intact branches than for excised branches, while positive effect sizes indicate that the response variable was higher for excised branches than for intact branches. Error bars represent the 95% confidence intervals, the dashed vertical line represents the no-effect threshold (effect size = 0) and asterisks indicate significant differences at α = 0.05. Mixed-effects models were fitted separately for each response variable and included nested random effects for tree and branch identities to account for repeated measurements and paired branch comparisons; fixed effects included air temperature, measurement height, an interaction term between treatment (intact and excised) and species, and water potential, stomatal conductance and photosynthesis (where applicable).

Stomatal conductance data revealed that branch excision generally had minimal effects on iWUE across the five focal species (Figure 5). Regressions of intact and excised stomatal conductance on the USO model index, , were significantly positive for all five focal species (all P < 0.002; Figure 5A and B) except for intact B. papyrifera (P = 0.37). Excision significantly decreased the USO parameter g₁ for A. campestre (P = 0.025) and C. betulus (P = 0.0093). Excision did not have a significant effect on g₁ for B. papyrifera (P = 0.57) and Q. garryana (P = 0.73). However, excision significantly increased g₁ for Pseudotsuga menziesii (P = 0.031), indicating a decrease in iWUE for this species.

Figure 5.

Effects of branch excision on the parameter g₁ of the USO model (Medlyn et al. 2011). Relationships between stomatal conductance and the USO model index, , are shown for intact (A) and excised (B) branches of five focal species. A (μmol m⁻² s⁻¹) is the assimilation rate, C_a (μmol mol⁻¹) is the atmospheric CO₂ concentration and D (kPa) is the vapor pressure deficit between the leaf and surrounding air. Lines represent ordinary least squares regression fits of Eq. (2) for each species; all regressions were significant (all P < 0.002) except for intact B. papyrifera. Parameter g₁ values from the USO model plotted by species and treatment (C). Points represent estimated values and lines indicate 95% confidence intervals. Higher g₁ values correspond to lower intrinsic water-use efficiency. Excision resulted in significant decreases in g₁ for A. campestre and C. betulus, and a significant increase in g₁ for Pseudotsuga menziesii.

Across the compiled dataset comprising 26 diverse plant species, branch excision reduced stomatal conductance and photosynthetic rates in angiosperms and increased both variables in the single gymnosperm species (Figure 6). Among the angiosperms, these reductions were most pronounced in species with ring-porous or diffuse-porous xylem, anisohydric water-use strategies, and species producing latex or lacking exudates, where 95% CIs excluded a log response ratio of 0 (Figure 6C–G). Species with semi-ring-porous, isohydric water-use strategies, and resin-producing species exhibited no significant changes in gas exchange following excision, with 95% CIs that included a log response ratio of 0 (Figure 6C–G).

Figure 6.

Effects of branch excision on stomatal conductance and photosynthesis rates across 26 plant species. Measurement locations are plotted on a global map (A) and in temperature-precipitation climate space (B; Whittaker 1975) with points sizes proportional to number of species reported at each location. Mean log-transformed response ratios (Eq. (3)) are shown for individual species (C), and grouped by clade (D), xylem type (E), water-use strategy (F) and exudate type (G). Error bars indicate 95% confidence intervals where data were sufficient. Some species could not be classified into all categories. Sample sizes for each category are as follows: gymnosperm n = 22, angiosperm n = 167, semi-ring-porous n = 22, ring-porous n = 30, nonporous n = 22, diffuse-porous n = 108, isohydric n = 46, anisohydric n = 78, resin n = 26, none n = 96, latex n = 34. 1, tropical rainforest; 2, temperate rainforest; 3, tropical seasonal forest; 4, temperate seasonal forest; 5, boreal forest; 6, savannah; 7, grassland; 8, tundra; 9, desert.

Discussion

This study investigated how branch excision influences branch water potential and leaf gas exchange. We first quantified excision effects on water potential, stomatal conductance, photosynthesis and the USO parameter g₁ across five tree species differing in hydraulic traits and water-use strategies. We then assessed the generality of excision effects on gas exchange using a meta-analysis of 26 plant species spanning diverse taxa, traits and biomes. In our experiments, excision consistently increased branch water potential but yielded contrasting gas exchange responses: stomatal conductance and assimilation declined in angiosperms but increased in the gymnosperm, with mixed results for g₁ among angiosperms and an increase in the gymnosperm. Consistent with our experimental results, the meta-analysis showed that excision generally reduced stomatal conductance and photosynthesis across species, although effect sizes varied widely. Gymnosperm responses in the meta-analysis were limited to a single species and should therefore be interpreted cautiously. Together, these findings indicate that while excision increases branch water potential, it generally reduces gas exchange, with the magnitude and direction of responses mediated by hydraulic traits and water-use strategies.

Although excision increased branch water potential, this almost never improved gas exchange in the species we examined. In our experiments, stomatal conductance and photosynthesis declined in four of five species despite increased xylem water potentials (Figures 3 and 4). Likewise, our meta-analysis revealed significant declines in stomatal conductance and photosynthesis in 21 of 26 species (Figure 6). These results are consistent with prior work showing rapid reductions in assimilation following excision in a variety of tree species (Missik et al. 2021). This suggests that mechanisms beyond sap water potential, such as incomplete rehydration, reduced cross-sectional area due to shearing of cut conduits, or cavitation even when cutting under water (Santiago and Mulkey 2003), often limit gas exchange in excised branches. These hydraulic disturbances can alter leaf water status and disrupt the transport of hormones and nutrients (Santiago and Mulkey 2003), inducing stomatal closure and limiting photosynthesis (Miyazawa et al. 2011). Thus, the potential benefit of tension release appears to be outweighed by hydraulic damage or chemical transport impacts, imposing an upper limit on achievable gas exchange, especially in species with high baseline conductance. In contrast, species with inherently lower conductance or photosynthetic rates may be less affected.

Branch excision is a common solution to the challenge of sampling leaves in tall or inaccessible canopies, improving feasibility of experiments and enabling increased throughput. Gas exchange analysers have become essential tools in plant physiology, but measuring intact leaves at height is often impractical or impossible, so excision is widely used. Despite its prevalence, relatively few studies report quantitative, extractable comparisons of gas exchange between intact and excised branches using standardized methods. Many studies evaluate excision only qualitatively as a validation step, or report preliminary comparisons without presenting data suitable for inclusion in our analyses (Ginn 1989, Zhang et al. 1993, Ellsworth and Liu 1994, Dang et al. 1997, Valladares et al. 1997, Bernacchi et al. 2003, Huxman et al. 2003, Warren et al. 2003, Niinemets et al. 2005, Ethier et al. 2006, Warren 2006, Ewers et al. 2007, Lichtenthaler et al. 2007, Monson et al. 2007, Urban et al. 2008, Han 2011). Detailed quantitative comparisons remain limited to a small number of studies that are often focused on a few species or particular environments, notably tropical forests (Lakso 1982, Lange et al. 1986, Meng and Arp 1993, Santiago and Mulkey 2003, Miyazawa et al. 2011, Verryckt et al. 2020, Kar et al. 2021, Missik et al. 2021). Consistent with our results (Figures 4–6), most of these studies report lower photosynthetic rates and stomatal conductance in excised branches, though some exceptions exist under extremely high-rainfall conditions (e.g. Verryckt et al. 2020).

Several hypotheses have been proposed to explain reduced gas exchange after excision. First, loss of hydraulic conductivity due to embolism or mechanical damage during excision may reduce hydraulic conductivity even when branches are cut under water (Santiago and Mulkey 2003). Mechanical shearing, conduit collapse or occlusion at the cut surface may further limit water uptake (Michaletz et al. 2012, Davidson et al. 2022, Simovic and Michaletz 2025). Although releasing water tension increases water potential, our results and previous studies (Missik et al. 2021) show that the reduction in water transport capacity after excision typically outweighs any benefit of tension release, leading to reduced leaf water status and stomatal closure. Second, an alternative hypothesis is that excessively high (less negative) water potentials in excised branches may actually inhibit stomatal opening, as high turgor pressure in surrounding epidermal or subsidiary cells could mechanically compress guard cells, leading to closure (Glinka 1971, DeMichele and Sharpe 1973, Edwards et al. 1976, Roelfsema and Hedrich 2002, Henry et al. 2019, Westbrook and McAdam 2021). Although such decoupling of leaf water status and stomatal behavior has been only infrequently discussed in the literature (Heath 1959), it may contribute to the observed declines in gas exchange (Pichaco et al. 2024). Our findings are consistent with these mechanisms: despite release of sap tension, most excised branches exhibited reduced stomatal conductance and photosynthetic rates. The magnitude of these effects was greatest in species with high baseline gas exchange, suggesting excision imposes an upper limit on achievable rates (Kar et al. 2021).

The divergent responses among species in our experiments underscore the importance of hydraulic traits in mediating excision effects. The only gymnosperm in our experiments (Pseudotsuga menziesii) maintained or increased gas exchange following excision, in contrast to all angiosperm species, which experienced declines. This likely reflects fundamental differences in xylem anatomy (tracheids vs vessels) and stomatal regulation (isohydric vs anisohydric) (Klein 2014). Pseudotsuga menziesii possesses tracheid-based xylem characterized by smaller conduit diameters and higher wall thickness-to-span ratios, which confer greater mechanical resistance to shearing and conduit wall deformation (Simovic and Michaletz 2025) at the cut surface, thereby better preserving hydraulic conductivity following excision. In contrast, angiosperms with vessel-based xylem may be more susceptible to conduit deformation or disruption during excision, leading to reduced hydraulic conductivity and downstream declines in stomatal conductance and photosynthesis. Differences in stomatal regulation may further modulate these responses, as Pseudotsuga menziesii also exhibits relatively conservative, isohydric behavior (Klein 2014), which may reduce sensitivity to transient hydraulic perturbations.

Our meta-analysis broadly supported these experimental results (Figure 6). Excision-induced declines in gas exchange were greatest in angiosperms, in species with ring- or diffuse-porous xylem, and in those with anisohydric water-use strategies, whereas declines were minimal in species with semi-porous xylem and in those with isohydric strategies. As only one gymnosperm species was included in the experiments and the meta-analysis, inferences regarding gymnosperm responses should be interpreted as preliminary. We found mixed support for the hypothesis of Santiago and Mulkey (2003) that excision effects should be especially pronounced in resinous and latex-producing species (Figure 6G): latex-producing species exhibited significant declines in stomatal conductance and photosynthesis following excision, whereas resinous species did not exhibit significant decreases in gas exchange.

The USO parameter g₁ provides insight into how stomatal regulation influences gas exchange responses to excision. In two of our angiosperm species, excision decreased g₁, indicating an increase in iWUE (Medlyn et al. 2011, Lin et al. 2015, Wu et al. 2020). In these species, stomatal conductance declined proportionally more than photosynthesis, resulting in greater carbon gain per unit water loss. Importantly, this response does not imply improved water supply following excision; rather, the increase in branch water potential reflects relaxation of xylem sap tension in a hydraulically disconnected system, while excision may simultaneously reduce hydraulic conductivity via, for example, occlusion by exudates and/or conduit wall deformation associated with shearing at the cut surface. Under such conditions, a decrease in g₁ is consistent with more conservative stomatal regulation despite increased water potential. In the two angiosperm species where excision did not affect g₁, stomatal conductance and photosynthesis declined together, maintaining their ratio and indicating that excision reduced overall gas exchange without altering the balance between carbon gain and water loss. In contrast, Pseudotsuga menziesii exhibited a significant increase in g₁, reflecting reduced intrinsic iWUE. This species was unique in exhibiting increases in both stomatal conductance and photosynthesis following excision, with stomatal conductance increasing proportionally more than assimilation. Together, these results indicate that excision does not induce novel stomatal behavior, but instead accentuates species-specific regulatory strategies. Although excision increased branch water potential across species, responses of stomatal conductance depended on intrinsic hydraulic traits and water-use strategy, demonstrating that water potential alone is insufficient to predict gas exchange responses in excised branches.

Our findings have important implications for studies that rely on excised branches to measure gas exchange. Branch excision is not a neutral manipulation; gas exchange measurements from excised branches generally underestimate those of intact branches, and the magnitude of this bias is strongly species-dependent (Kar et al. 2021, Missik et al. 2021). Even with careful methods such as cutting under water, using segments longer than the maximum vessel length, and allowing a 20-min acclimation, excision impacts were not eliminated. We recommend that researchers quantify excision effects in their system by directly comparing measurements from intact and excised branches, rather than assuming excision has no impact based on prior studies. Where consistent biases are detected, appropriate corrections should be applied. Our meta-analysis provides baseline, trait-based expectations for excision effects (Figure 6): the single gymnosperm species showed small increases in stomatal conductance and photosynthesis, whereas angiosperms generally exhibited substantial declines. Such knowledge can inform experimental design and interpretation of gas exchange data.

Several methodological practices can help minimize excision-induced artifacts. Ensuring sample branches are at least twice the maximum vessel length and making all cuts under water will minimize embolism. For some species, allowing a brief acclimation period after excision can promote partial recovery of water status and gas exchange (Santiago and Mulkey 2003, Miyazawa et al. 2011, Missik et al. 2021). However, acclimation times and sensitivity vary among species, and many (particularly those with large vessels or anisohydric strategies) may exhibit rapid and persistent declines in gas exchange after excision (Missik et al. 2021). In these cases, measurements should be made as soon as possible after excision, as declines in conductance and photosynthesis often occur within minutes (Santiago and Mulkey 2003). Maintaining high humidity around the excised branch may help slow the onset of stomatal closure. For species that are especially sensitive to excision, alternative approaches such as in situ measurement may be preferable. Where excision is unavoidable, resulting measurements should be interpreted as lower-bound estimates of intact gas exchange rates, as excision generally reduces stomatal conductance and photosynthesis relative to intact branches.

Conclusion

We combined new experimental data with a meta-analysis of 26 species to characterize the effects of branch excision on branch water potential and leaf gas exchange. Excision generally reduces gas exchange, but the magnitude of reduction depends on hydraulic traits and water-use strategies. This underscores the close relationships between plant hydraulics and stomatal regulation: species that tightly control water loss or are less sensitive to hydraulic disruption can better buffer against the effects of excision. Our results highlight the need to account for excision effects in analyses and models, as measurements from excised branches are not always equivalent to those from intact plants. We advocate for reporting excision effects and for future studies to quantify and correct for them when possible. While excision remains a necessary and valuable approach, its impacts on plant water relations and gas exchange can be substantial. Recognizing and correcting for these impacts will improve the accuracy and reliability of gas exchange data and predictions in plant ecophysiology.

Our findings have important implications across multiple levels of biological organization. At scales below the branch, they underscore how processes such as embolism formation, maintenance of hydraulic continuity, and stomatal behavior at the cellular and leaf levels can constrain or facilitate whole-branch gas exchange following excision. At larger scales, accounting for excision effects will increase the accuracy of canopy- and ecosystem-level estimates of transpiration and carbon exchange, thereby improving the predictability of plant, land surface and Earth system models that are parameterized using leaf-level gas exchange data.

Author contributions

M.C. and S.T.M. designed the research. M.C. collected and analyzed the data with guidance from S.T.M. M.C. wrote the first draft of the paper, and both authors contributed to revisions.

Conflict of interest

None declared.

Funding

This work was funded by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery grant.

Data availability

All data and code is available at the public GitHub repository: https://github.com/MichaletzLab/Excision-Photosynthesis-Methods.git