Sap flow sensor integration for precision irrigation management in Christmas trees - a scoping review

1 Introduction

Climate change is intensifying global water scarcity by increasing the frequency, duration, and severity of drought events which is posing a critical challenge to agricultural and perennial production systems (Dang et al., 2021; Cook et al., 2018). Across diverse biomes, drought-induced tree mortality has been widely reported, highlighting the sensitivity of woody perennials to water limitation and thermal stress. In both managed and semi-managed systems, water availability has emerged as a primary limiting factor governing plant growth, productivity, and survival, particularly in regions characterized by high evaporative demand and constrained water supplies (Pereira et al., 2004). These trends have elevated the need for efficient, demand-based irrigation strategies that conserve water resources while sustaining crop performance under increasing climatic variability. In addition to evaporation, transpiration is also a dominant pathway of water loss in terrestrial plants and a major component of the global hydrological cycle (An, 2022). Plants characteristics, irrigation management, weather conditions, and environmental parameters all affect evaporation and transpiration, which governs ecosystem-scale water and energy fluxes. Plants absorb water through their roots, transport it upward through the xylem, and release it to the atmosphere via leaf stomata, with more than 90% of absorbed water ultimately lost through transpiration (Han et al., 2024). At both field and landscape level, transpiration influences soil moisture dynamics, runoff generation, and local climate feedback, highlighting its importance for irrigation scheduling and water management.

Plant water use is primarily regulated by stomatal behavior, which are commonly described as an isohydric–anisohydric continuum (Klein, 2014). Isohydric species tightly regulate stomatal conductance to maintain relatively stable leaf water potential during drought, reducing transpiration to avoid hydraulic failure. In contrast, anisohydric species allow greater fluctuations in leaf water potential, sustaining transpiration and carbon assimilation under declining soil moisture at the cost of increased vulnerability to xylem cavitation (Klein et al., 2013). These hydraulic strategies have direct implications for irrigation management because they determine how rapidly transpiration responds to soil drying, atmospheric demand, and irrigation events. For conifer species, which dominate many perennial production systems, species-specific stomatal regulation strongly influences transpiration dynamics, drought tolerance, and water-use efficiency (Guo et al., 2022), underscoring the need for irrigation approaches that account for plant physiological behavior rather than relying solely on soil or atmospheric indicators (Jones, 2004).

Accurate quantification of whole-plant water use is essential for improving irrigation management in perennial systems (Han et al., 2019; Song et al., 2023). Traditional irrigation practices often rely on fixed schedules or soil moisture thresholds that do not adequately capture dynamic plant water demand, leading to either water stress or excessive irrigation. In contrast, plant-based measurements provide integrative information on how trees respond physiologically to environmental conditions and management interventions. Among these approaches, sap flow measurements offer a direct and continuous assessment of transpiration by quantifying the movement of water through the xylem in response to both environmental drivers and internal physiological controls. Sap flow and sap flux density have been widely used to estimate whole-plant water use (Myers et al., 1998; Pataki, 1998; Ryan et al., 2000; James et al., 2002; Ford et al., 2004b), hydraulic conductance (Pataki, 1998; Ryan et al., 2000), canopy transpiration (Granier et al., 1996b; Teskey and Sheriff, 1996; Oren et al., 1999), and stomatal conductance (Granier and Loustau, 1994; Oren et al., 1998), making sap flow sensing a powerful tool for plant-based irrigation scheduling. For instance, an automatic irrigation based on sap flow sensing enabled more precise physiological regulation in Cherry tree, with sap flow closely tracking seasonal changes and climatic drivers such as temperature and solar radiation (Jin et al., 2024). In addition, the study showed that sap flow measurement was used to accurately quantify maize water use, with the novel low-cost sensors measuring daily transpiration (Miner et al., 2017). This validated sap flow approach demonstrated reliable monitoring of plant water use at both individual and field scales, supporting irrigation efficiency, and plant water stress assessment (Capurro et al., 2024).

Christmas trees as commercially cultivated conifers, represent a high-value perennial production system, with a global market value of $5.91 billion in 2023 (Global Market Research Reports, 2024). In the United States, commercial Christmas tree production occurs in all 50 states across approximately 16, 000 tree farms. According to the 2022 Census of Agriculture, more than 14.5 million Christmas trees were harvested in 2022. Between 2000 and 2015, U.S. live Christmas tree imports averaged 2 million trees annually (US $36.1 million), increasing to 2.8 million trees valued at US $68 million by 2022, reflecting growing market demand and production challenges associated with climate variability (Scott, 2024).

Christmas tree production in the United States is dominated by conifer species such as Abies fraseri (Fraser fir), Pseudotsuga menziesii (Douglas-fir), Abies procera (Noble fir), and Abies nordmanniana (Nordmann fir), which are valued for their aesthetic quality, fragrance, and post-harvest needle retention (Darr et al., 2022). However, prolonged drought conditions across major production regions contributed to a decline of approximately 550, 000 harvested trees in 2022 compared with the 2017 agricultural census (Scott, 2024). Historically, irrigation management in Christmas tree plantations has relied largely on empirical observations and grower experience. However, increasing climate-induced drought frequency and variability have exposed the limitations of these approaches, driving a shift toward precision, demand-based irrigation strategies (Lindberg and Cregg, 2022).

Sap flow sensors function as biological monitors that integrate dynamics of the Soil-Plant-Atmosphere Continuum (SPAC), enabling the transition from calendar-based irrigation to plant-driven water management. By linking irrigation decisions directly to transpiration demand, sap flow–based approaches offer the potential to improve water-use efficiency, enhance seedling establishment, and reduce risks associated with over-irrigation, such as nutrient leaching and root-rot diseases (Dong et al., 2023). Although sap flow techniques have been widely applied in orchards, forests, and other woody perennial systems (Noun et al., 2022), several characteristics distinguish Christmas tree production and justify a dedicated review. Unlike fruit orchards where yield is the primary economic outcome or forests where timber volume is often the main objective, Christmas tree production depends heavily on visual quality attributes such as tree height, canopy density, symmetry, color, and needle retention (National Centers for Environmental Information, 2024). Consequently, even moderate water stress can reduce marketability despite limited effects on biomass accumulation. In addition, most Christmas tree species are conifers, whose xylem structure consists primarily of tracheids rather than the vessel elements found in many hardwood species. This fundamental difference in hydraulic architecture influences water transport dynamics, sapwood conductivity, and heat dissipation characteristics, requiring careful interpretation, calibration, and validation of sap flow measurements. By focusing on conifer physiology, Christmas tree production objectives, and practical irrigation applications, this review identifies critical knowledge gaps and provides guidance for future research and technology adoption within the Christmas tree industry. Specifically, it (i) examines the physiological basis of sap flow and its relevance for irrigation management in conifers, (ii) evaluates commonly used sap flow measurement techniques and their suitability across tree developmental stages, and (iii) assesses how sap flow sensing can be integrated with weather-based and soil-based approaches to improve water-use efficiency under climate-driven drought conditions. By consolidating advances across plant ecophysiology, sensor technology, and irrigation science, this narrative review provides a framework for advancing precision irrigation management in perennial cropping systems.

2 Physiological basis of sap flow in conifers and implications for irrigation management

Sap flow represents the movement of water through xylem and provides a direct physiological measure of whole-plant water use (Gimenez et al., 2004). In conifers, sap flow is driven primarily by transpiration, which generates a negative pressure gradient along the soil-plant-atmosphere continuum and pulls water upward from the root system to the canopy (Chuang et al., 2006; Meinzer et al., 2006). Since transpiration integrates soil moisture availability, atmospheric demand, and internal plant regulation, sap flow measurements offer a uniquely informative metric for irrigation management in perennial cropping systems, particularly under variable climatic conditions (Kumar et al., 2022). Accurate interpretation of sap flow signals for irrigation scheduling requires an understanding of the hydraulic architecture and physiological controls that govern water transport in conifers, which is described in the following sections.

2.1 Hydraulic architecture of conifers

The hydraulic architecture of conifers is a complex system shaped by both environmental variability and plant structure. Environmentally, it is governed by soil water availability and atmospheric vapor pressure deficit; structurally, it depends on xylem quantity (sapwood width and path length), variation in xylem properties throughout the plant, root system size, leaf area, capacitance (the volume of water released from storage per unit change in water potential), and physiological factors such as the triggers governing stomatal opening and closure (Hacke et al., 2015).

Conifers possess distinctive xylem anatomy and drought-response mechanisms that strongly influence transpiration dynamics and sap flow measurement. In angiosperms, xylem is composed of fibers and water-conducting vessel elements. Conifer xylem is more homogenous, consisting of tracheids that provide mechanical support and conduct water. These tracheids are typically 1–4 mm in length and 20-80 µm in diameter. In a tracheid, water movement occurs both axially and laterally between adjacent cells through bordered pits in the cell walls (Figure 1). The bordered pit membrane is composed of a porous margo surrounding a dense central torus, which balances hydraulic efficiency with safety (Domec et al., 2006; Sviderskaya et al., 2020; Song et al., 2022). Under well-watered conditions, the pit membrane remains centrally positioned, allowing water to pass through the margo pores with minimal resistance. As soil moisture declines and xylem tension increases, the torus is displaced against the pit aperture, effectively sealing the conduit and preventing the spread of embolism from cavitating functional tracheid (Domec et al., 2006; Song et al., 2022). The torus–margo mechanism boosts hydraulic safety and raises resistance to sap flow, resulting in reduced peak sap velocities and more consistent radial flow patterns in conifer stems (Hu et al., 2022; Song et al., 2022).

These anatomical features have direct implications for sap flow measurements. It increases the sensitivity of measurements, probe positioning, sapwood depth estimation, and species-specific calibration coefficients, particularly when thermal-based techniques are applied (Hu et al., 2022).

2.2 Transpiration, stomatal regulation, and drought response

Transpiration in conifers occurs primarily through stomata located on needles and accounts for most of the water absorbed by the root system (Zhang et al., 2024). Based on the amount of soil moisture, conifers have different responses to external variables. Under adequate soil moisture, stomatal conductance responds dynamically to atmospheric drivers such as vapor pressure deficit and solar radiation, allowing transpiration to support photosynthesis, nutrient transport, and evaporative cooling (Conard et al., 1997; Meinzer et al., 2013). However, under drought conditions, stomatal closure becomes the primary defense mechanism, reducing water loss through transpiration and limiting the risk of hydraulic failure (Conard et al., 1997; Meinzer et al., 2013). Therefore, this response entails a fundamental physiological trade-off. Reduced stomatal conductance also restricts CO₂ uptake, suppressing photosynthesis and growth (Zhang et al., 2024). The survival of seedlings on hot, dry sites depends on maintaining relatively high transpiration rates and stomatal conductance (Kolb and Robberecht, 1996). These processes enable effective evaporative and conductive cooling of stems and needles, helping to prevent tissues from reaching lethal temperatures under extreme heat and drought conditions.

2.3 Isohydric vs. anisohydric strategies and hormonal regulation in conifers

Many conifers in Pinaceae family such as Abies, Pseudotsuga, Pinus and Picea exhibit isohydric stomatal regulation (Steppe et al., 2010). Isohydric species tightly regulate stomatal conductance to maintain relatively constant leaf water potential during drought. These drought responses are largely governed by hormonally mediated signaling, primarily involving abscisic acid (ABA). As soil moisture declines, foliar ABA concentrations increase, inducing early and sustained stomatal closure. This hydraulic strategy reduces the risk of xylem cavitation and hydraulic failure. However, it results in delayed physiological recovery after irrigation because stomatal reopening remains constrained until ABA concentrations decline (Brodribb et al., 2014). In addition to ABA, ethylene plays a critical role in drought-related stress responses in conifers, particularly with respect to needle retention and postharvest quality (Lada and Macdonald, 2015). Drought-induced ethylene production, often acting in coordination with elevated ABA levels, can trigger the formation of an abscission layer at the base of needles (Brodribb and McAdam, 2013). This physiological process signifies that plant-based irrigation strategies such as sap flow are increasingly emphasized in precision irrigation management (Meinzer et al., 2013).

2.4 Internal capacitance and stem water storage

Internal capacitance is the temporary storage of water within stem tissues which plays an important role in conifer water relations. Water stored in the sapwood can buffer demand through transpiration during periods of high evaporative demand, creating a temporal lag between foliar transpiration and measurable sap flow at the stem base (Kumagai et al., 2009; Uddin et al., 2014). In addition, the internal water storage in mature conifers can contribute to a substantial fraction of daily transpiration. For example, in Cryptomeria japonica, approximately 33% of transpired water during peak daytime hours has been shown to originate from internal stem reserves, which are subsequently replenished at night (Kumagai et al., 2009). This buffering capacity complicates irrigation interpretation, as sap flow reductions may lag the onset of physiological stress during short-term drought events (Kumagai et al., 2009; Uddin et al., 2014).

2.5 Root system architecture of conifers

In conifers, root system architecture governs water uptake through the spatial arrangement of coarse and fine roots and the hydraulic properties of the pathways created by roots. Coarse structural roots and deeper axes connect the tree to water stored at depth and provide low-resistance axial transport, whereas fine absorptive roots concentrated in upper soil layers mediate most radial entry of water from soil into the plant (Hacke et al., 2015; Boursiac et al., 2022). Mechanistically, whole-root-system water uptake is co-limited by radial conductance across root tissues and axial conductance through xylem, so changes in branching pattern, root order, and the distribution of conductive tissues alter how efficiently water acquired in different soil layers can be delivered to the stem (Boursiac et al., 2022). This means that conifer sap flow measured in the trunk should be interpreted as the integrated outcome of belowground architecture, soil moisture distribution, and atmospheric demand rather than as a direct measure of shallow root uptake alone (Wullschleger et al., 1998). Evidence from conifers indicates that species able to access deeper soil water can maintain stronger coupling between root uptake and stem transport during dry periods; for example, Douglas-fir showed more effective deep-water acquisition than Norway spruce, helping sustain transpiration when surface soils dried (Nadezhdina et al., 2014). When conifer root systems span wet and dry soil layers, water can move passively through roots along water-potential gradients, producing nocturnal transfer from deeper moist soil to drier shallow horizons or, after rainfall, downward transfer from wet surface soil to depth (Burgess et al., 1998; Burgess et al., 2001; Caldwell et al., 1998). In Pacific Northwest conifer forests, both Douglas-fir and ponderosa pine showed substantial hydraulic redistribution, with reverse root sap flow and partial nocturnal recharge of upper soil layers during summer drought (Brooks et al., 2002). This process can prolong fine-root activity in dry upper horizons, buffer daytime transpiration, and alter the timing and source depth of water later expressed as stem sap flow (Caldwell et al., 1998; Brooks et al., 2002). Accordingly, diel sap flow patterns in conifers show transpiration demand and nighttime reallocation of water within the root system.

3 Sap flow measurement techniques

The selection of sensor methodology is critical in integrating sap flow sensing into irrigation monitoring and control. This is especially true in Christmas tree production, as the thermal properties of tracheid wood and the resinous nature of many conifer species can interfere with certain measurement principles. A comparison of the main sap flow methods and their suitability for conifer species used in Christmas tree production is presented in Table 1.

Table 1. Comparison of different sap flow techniques including principle, strengths and limitations.

Beyond the principles, strengths, and limitations summarized in Table 1, method selection for Christmas tree production should also weigh relative cost, accuracy, installation complexity, and ongoing maintenance burden. Recent studies demonstrate that low-cost and open-source sap flow sensors can approach the accuracy of commercial systems at a small fraction of the cost, which has direct relevance for multi-farm deployment. (Davis et al., 2012) showed that self-made thermal dissipation and heat ratio sensors, built for roughly 2% of the cost of commercial equivalents, produced acceptable sap flow estimates when field-tested against a commercial probe. More recent studies have extended this finding across different methods and species which are listed in Table 2: (Chen et al., 2025) validated a compact, low-cost Granier-style thermal dissipation system on ash trees (Fraxinus spp.), achieving a 72% data collection success rate under field conditions; (Beslity et al., 2022) developed a low cost $150 heat-pulse-based sap flux device with a laboratory-validated adjusted R² of 0.96; (Dangare et al., 2017) constructed a microcontroller-driven thermal dissipation system for a total cost of US $72.64 (excluding labor), achieving an R² of 0.82 against a commercial Dynagage reference sensor; and (Christensen et al., 2021), through Oregon State University’s OPEnS Lab, developed an open-source heat ratio method sensor using flexible printed circuit boards that achieved an R² of 0.90, with error analysis showing that a 2% error in the temperature ratio translated to only a 0.05% overestimation in daily water use. The broader commercial market continues to grow alongside this open-source ecosystem, led by established manufacturers such as Dynamax and METER Group.

Table 2. Summary of studies measuring sap flow in different conifer species, including methods, main findings and references.

3.1 Heat balance method

The heat balance method, also referred to as the stem segment heat balance method, estimates sap flow by applying a known quantity of heat to a defined segment of xylem and quantifying the redistribution of this heat by conductive and convective processes (Figure 2). In this approach, a short stem segment is heated using internal (direct electric) heating, typically delivered through three to five electrodes inserted circumferentially into the xylem tissue. The configuration of the heating elements and temperature sensing points follows the experimental designs described by (Čermák et al., 1973; Kučera et al., 1977).

Temperature gradients within and around the heated stem segment are monitored using a set of copper–constantan thermocouples, positioned to measure axial and radial heat fluxes as well as temperature differences between heated and unheated xylem regions. The thermocouple signals are recorded using a datalogger, and sap flow is subsequently calculated by solving the differential heat balance equation for the heated stem segment (Kučera et al., 1977; Cienciala et al., 1992). The general heat balance for the heated stem segment can be expressed as Equation 1:

(1) Qh=Qf+Qc+Qr+Qs

where, Qh is the electrical heat input supplied to the stem segment, Qf refers to the heat transported by moving sap (convective heat flux), Qc denotes the axial conductive heat loss along the stem, Qr expresses the radial heat loss to the surrounding environment, and

Qs represents heat storage within the stem tissue.

Under steady-state conditions, the storage term (Qs) is often assumed to be negligible, allowing sap flow–related heat transport to be calculated using Equation 2

(2) Qf=Qh−(Qc+Qr)

Sap flow rate (F) is then derived by relating the convective heat flux to the temperature difference between upstream (T₁) and downstream (T₂) thermocouples and the density (ρ) and specific heat of sap (cp) as given by Equation 3:

(3) F=Qh−(Qr+Qc)ρcp(T2−T1)

Herzog et al. (1997) estimated sap flow on mature Norway spruce (Picea abies), comparing two HB method techniques, tissue heat balance (THB) in large diameter stems and stem heat balance (SHB) in small diameter stems. They demonstrated that balancing heat appears to be suitable for quantifying the flow of sap in thin stems by SHB rather than in thick stems as intended by THB, unless complex calibrations are made. Sap flow estimates from the two HB implementations agreed reasonably well, supporting their use as a reference for calibrating other methods such as Thermal Dissipation (TD) and heat pulse techniques. Groot and King (1992) similarly demonstrated the applicability of Dynamax-type heat balance sensors for measuring sap flow in coniferous seedlings, further supporting the method’s suitability for the small-diameter, juvenile stems characteristic of Christmas tree production. Beside conifers, SHB sap flow has also proven effective for tracking plant water status in broadleaf species. In poplar, sap flow-based canopy transpiration was found to be tightly linked to tree phenological stage, vapour pressure deficit, and photosynthetic photon flux density (Navarro et al., 2020), supporting that stem heat balance sap flow sensors as a robust, continuous indicator of plant water status across tree types. In addition, Figure 3 illustrates the sap flow measurement of two Fraser fir trees with different heights conducted by the authors using SHB. Despite differences in magnitude, both trees followed a similar temporal pattern, demonstrating the accuracy and reliability of this method and supporting its applicability.

3.2 Thermal dissipation probe method

The thermal dissipation probe (TDP) method, originally described by Granier, is the most common technique due to its simplicity and continuous monitoring capability. The system uses two probes inserted into the sapwood: a continuously heated probe and an unheated reference probe (Figure 4). Continuous heating is radiated into the stem and the temperature difference between the heated and an unheated upstream point is quantified (ΔT). The temperature difference at which zero flow is assumed needs to be determined (ΔT₀). Sap flux density u (m³ m⁻² s⁻¹) can be empirically determined from the measured temperature difference using Equation 4 (Granier, 1987; Granier et al., 1996a):

(4) u=a.[(ΔT0−ΔT)ΔT]b

The parameters a and b were determined experimentally by (Granier and Bréda, 1996) using five tree species (Pseudotsuga menziesii, Pinus nigra, Quercus robur, Castanea sativa, Prunus malus), and calibrating against columns filled with synthetic fiber and sawdust as: a = 118.99 ×10^-6 (m³ m^-2 s^-1), b = 1.231.

TDP is widely used in conifers because installation is relatively simple and power demand is low, but accuracy depends strongly on species specific calibration and handling of low flow and cold season conditions. Research conducted in two Abies forests by (Chan and Bowling, 2017) concluded that TDP method can be reliably used to examine sap flow during cold periods, recommended species-specific or at least xylem-type-specific calibration for conifers and highlighted the importance of accounting for temperature-dependent baseline drift when operating near freezing. A comparison of Tissue Heat Balance (THB) and TDP in a conifer (Pinus echinata) and two ring-porous oaks (Quercus prinus Willd. and Quercus velutina Lam.) showed close agreement between methods for the pine, while TDP underestimated sap flow by 20–50% in two species of ring-porous oaks (Renninger and Schaefer, 2012).

3.3 Heat ratio method

The Heat Ratio Method (HRM) has emerged as a benchmark technique for conifer research because it employs short-duration heat pulses (typically 2–6 s) rather than continuous heating, thereby reducing sensitivity to background thermal gradients within the stem (Burgess et al., 2001; Vandegehuchte et al., 2015). In HRM, a brief heat pulse is released from a heater probe, and the resulting temperature rise is measured simultaneously at sensors positioned equidistant upstream and downstream of the heater (Vandegehuchte and Steppe, 2012). Sap flow velocity is inferred from the ratio of temperature increases detected by these sensors.

The heat pulse velocity (vh) is calculated as Equation 5:

(5) vh=kxln(ΔT1ΔT2)

where, k is the thermal diffusivity of the sapwood, X refers the distance between the heater and each temperature sensor, ΔT₁ and ΔT₂ are the temperature increases measured at the upstream and downstream sensors, respectively.

The HRM is particularly well suited for Christmas tree plantations for several reasons. First, it enables the measurement of very low sap velocities, including near-zero flow conditions (Burgess et al., 2001), which is critical for identifying nighttime stem refilling and early-season physiological activity such as bud swelling. Second, HRM can detect bidirectional sap flow, allowing quantification of reverse flow associated with hydraulic redistribution or foliar water uptake events (Wang et al., 2018). Finally, because HRM relies on a pulsed heat signal, it is relatively insensitive to natural temperature gradients in the stem, thereby minimizing errors caused by diurnal trunk temperature fluctuations (Vandegehuchte et al., 2015).

3.4 Heat pulse velocity method

The HPV method, originally developed by Huber (1932), as cited in Ishtda et al., 1991), estimates xylem flow velocity by tracking the time required for a heat pulse to travel through sapwood. In this technique, short heat pulses are emitted into the stem at predetermined time intervals (Figure 5). Temperature sensors are inserted equidistantly above and below the heater, typically at distances of x=0.5 – 0.6 cm, and record the arrival times of the heat pulse downstream (td) and upstream (tu). The heat pulse velocity is calculated as Equation 6:

(6) vh=x2(1td1tu)

Where V_h is the heat pulse velocity, x is the distance between the heater and each temperature sensor, and t_d and t_u are the times taken for the heat pulse to reach the downstream and upstream sensors, respectively.

The thermal diffusivity of the sapwood (k) must be determined independently prior to sap flow calculations. To convert heat pulse velocity to actual sap velocity, corrections are applied that account for the thermal properties of sap and xylem tissue, including the density and specific heat capacity of sap and volumetric water content of the sapwood (Vandegehuchte et al., 2015). When properly calibrated, heat pulse techniques can provide accurate sap flow measurements; however, the HPV method is less sensitive at very low flow rates, which limits its effectiveness under near-zero flow conditions (Becker, 1998).

Capurro et al. (2024) presented a novel sap flow sensor that uses a heat pulse technique to measure water flow through plant stems. Testing in a maize field showed reliable daily water use measurements within 5.9–9.4% of gravimetric values and 8% compared to commercial sap flow gauges. The sensor also demonstrated minimal differences (2–4.4%) when compared to transpiration estimates based on reference evapotranspiration and crop coefficients.

3.5 Compensation heat pulse method

The Compensation Heat Pulse (CHP) method is a refinement of the heat pulse approach that improves accuracy under moderate to high sap flow conditions. In this method, upstream and downstream temperature probes are inserted at equal radial depths into the sapwood, with the heater positioned between them. Typically, the upstream probe is placed at a distance xu=5 mm and the downstream probe at xd=10mm from the heater (Swanson and Whitfield, 1981; Swanson, 1994). Following the application of a heat pulse, the upstream probe initially records a higher temperature due to conductive heat transfer. As sap flow transfers heat downstream, the temperature at the downstream probe increases. The time required for the temperatures at the two probes to become equal (t₀) is used to calculate the heat pulse velocity. Heat pulse velocity is determined as Equation 7:

(7) vh=xd−xu2t0

Where x_d and x_u are the distances of the downstream and upstream probes from the heater, t₀ is the compensation time at which both probes register the same temperature.

By relying on the symmetry point of heat redistribution rather than exact arrival times or temperature ratios, the CHP method reduces sensitivity to noise and timing errors, making it particularly useful for steady and moderate sap flow conditions.

3.6 Heat field deformation

The heat field deformation method is used to measure sap flow in trees by detecting changes in the heat field around a linear heater embedded in the xylem (Nadezhdina et al., 2007). This method utilizes two types of sensors: multiple-point and single-point (Nadezhdina et al., 2002). The multiple-point sensor includes a heater and six thermocouples spaced along a stainless-steel needle, allowing for detailed measurements of temperature differences: symmetrical (ΔT_sym) and asymmetrical (ΔT_asym) around the heater. These differences reflect the movement of sap and the physical properties of the xylem. Single-point sensors, which contain only one thermocouple, are used for routine long-term monitoring and are placed at specific depths based on prior measurements from the multiple-point sensors (Nadezhdina et al., 2012).

To estimate whole-tree sap flow, data from multiple radial points on at least two sides of the stem are averaged. The depth of active sapwood is determined by identifying where sap flow approaches zero, or in some cases, by observing changes in water content or xylem color. Sap flow density at each point is multiplied by the area of the corresponding annulus to calculate sap flow volume, and total tree sap flow is obtained by summing these values. Radial integration curves, often modeled using polynomial or Gaussian functions, help generalize sap flow patterns across the xylem radius (Wimmler et al., 2024).

3.7 Transient thermal dissipation and TreeTalker systems

Innovative Internet of Things (IoT) platforms like the “TreeTalker” as an emerging method utilize the transient thermal dissipation (TTD) method to monitor tree health. The TTD systems reduce power consumption by employing cyclic heating for example, 10 minutes of heating followed by 50 minutes of cooling (Dluhosch et al., 2025). This approach allows for the simultaneous measurement of both sap flow and stem volumetric water content (VWC).

4 Sensor design and deployment

Sensor selection and installation in conifers must be explicitly aligned with stem diameter and sapwood depth. Sensor types and deployment protocol must be adjusted according to the developmental stage to minimize bias and tissue disturbance.

4.1 Nursery seedlings and small stems

For conifer seedlings and very small saplings, invasive probes would cause catastrophic damage to the vascular system. In such case, Stem Heat Balance (SHB) sensors are preferred because they provide non-invasive, whole-stem flow measurements without penetrating the xylem (Baker and Van Bavel, 1987) Commercial sensors such as the Dynagage series (Dynamax Inc., Houston, Texas, USA) and Implexx (Edaphic Scientific, Moorabin, Visctoria, Australia) sensors are designed for stems ranging from approximately 1 mm to 30–32 mm in diameter, including herbaceous and juvenile tissues (Dynamax Inc, 2005; Edaphic Scientific, 2025). These devices consist of a flexible heater collar with embedded thermistors that quantify axial and radial heat transport. Proper installation requires selection of smooth internodal sections, application of thermal interface material to eliminate air gaps, insulation to prevent heat loss and careful stabilization to prevent cambial compression, as small installation errors can disproportionately affect measurement accuracy at this scale.

4.2 Juvenile trees

As stem diameter increases, miniaturized invasive probes such as Thermal Dissipation Probes (TDP; Granier-type) or Heat Ratio Method (HRM) sensors can be deployed (Granier, 1987; Burgess et al., 2001). Juvenile conifers typically exhibit sapwood extending across most of the radial profile, enabling probes of 20–30 mm length to integrate flow across the active xylem. At this size, the primary concern is the ratio of sensor volume to stem volume, as the displacement of conductive tissue by the needles can significantly alter local hydraulics.

Installation should utilize a precision drill guide to ensure parallel probe alignment and low-speed drilling to avoid frictional heating, which can char xylem tissue and alter thermal conductivity (Steppe et al., 2010). In rapidly growing saplings, radial expansion can lead to probe misalignment “displacement error”, necessitating periodic inspection and recalibration.

4.3 Mature forest stands (10–50 cm DBH)

In mature conifer stands, standard TDP and HRM systems are widely applied due to their robustness and suitability for long-term monitoring (Lu et al., 2004). At this stage, accurate characterization of the sapwood–heartwood boundary becomes essential, as heartwood is hydraulically inactive and thermally distinct. The standardized protocol for these installations includes several critical steps:

• Sapwood area determination: An increment borer is used to extract a wood core. If the boundary between the sapwood (active) and heartwood (non-active) is not visible, methyl orange dye is applied. Heartwood typically stains red/purple, allowing for the precise measurement of the sapwood band thickness.
• Bark preparation: Loose, corky bark is removed at the instrumentation site to reach the cambium. The bark thickness is measured using a specialized gauge to ensure the sensor’s thermistors are placed at the correct depth below the cambium.
• Alignment and drilling: A pre-drilled metal guide is clamped to the tree. A low-RPM battery-operated drill is used to prevent the frictional burning of the tracheids, which would create a charred barrier to both sap flow and heat transfer.
• Vertical temperature gradient correction: In mature trees, natural vertical temperature gradients (VTG) can interfere with sap flow signals. The Goulden modification is often applied, which involves installing a separate pair of unheated reference probes laterally to the main sensor to subtract the ambient thermal noise from the measured sap flow (Goulden and Field, 1994).

5 Calibration of sensors

Conifer wood exhibits distinct thermal diffusivity and lower axial conductivity compared with vessel-bearing angiosperms, which can bias empirical calibration equations derived from broadleaf species. Species-specific calibration is therefore recommended (Bush et al., 2010; Steppe et al., 2010). Calibration protocols typically compare sensor-derived sap flux density against independent measurements of whole-plant transpiration or water uptake, including weighing lysimeters, gravimetric, or whole-tree potometric methods. Correction coefficients may vary substantially among conifer species due to differences in tracheid diameter, resin canal abundance, and wood density. Failure to account for these anatomical and thermal differences can result in systematic under- or overestimation of transpiration, particularly under low-flow conditions where signal-to-noise ratios are reduced. Table 3 presents the advantages and disadvantages of different methods of calibration for Christmas tree.

Table 3. Comparison of calibration methods for sap flow measurement, including their advantages, disadvantages, and application.

6 Wood anatomy, regression constants, and scaling to whole-tree transpiration

The relationship between the dimensionless flow index such as temperature differential ratio in TDP or HRM systems and sap flux density is generally linear across different xylem types, including tracheid-based conifer wood (Granier, 1987; Burgess et al., 2001). However, regression constants are influenced by wood anatomical and thermal properties, including tracheid dimensions, wood density, and thermal diffusivity (Lu et al., 2004; Steppe et al., 2010). Because conifers conduct water through relatively homogeneous tracheids rather than vessels, heat transport characteristics differ from those of angiosperms, potentially altering calibration slopes and intercepts. As a result, species-specific calibration is often required to avoid systematic over and underestimation of sap flux density, particularly under low-flow conditions (Bush et al., 2010; Vandegehuchte and Steppe, 2013). Scaling from point-based sap flux density to whole-tree transpiration remains a critical methodological step in forest ecohydrology. Single-point measurements must be integrated over the total conductive sapwood area to estimate total tree water use (Čermák et al., 2004; Lu et al., 2004). The point-to-area scaling coefficient (Ki–t) converts localized sap flux density to whole-tree sap flow and varies with species, stem diameter, and radial flow distribution. The circumferential scaling coefficient (Ks–t) accounts for azimuthal variation in sap flux around the stem; in many conifers, circumferential variation is relatively small, and Ks–t often approximates unity (Čermák et al., 2004).

Accurate scaling requires precise determination of sapwood area at the measurement height. This is commonly achieved using increment cores combined with dye staining (e.g., safranin) to distinguish active sapwood from non-conductive heartwood (Čermák et al., 2004; Vandegehuchte and Steppe, 2013). In younger plantation-grown conifers, sapwood may constitute more than 60% of the cross-sectional area, simplifying scaling compared with mature forest trees, where conductive flow may be confined to narrow outer growth rings and radial gradients become more pronounced (Lu et al., 2004; Steppe et al., 2010).

7 Applying sap flow techniques in precision irrigation scheduling and decision support

Evapotranspiration based irrigation scheduling methods are widely used since they provide a standardized approach for estimating crop water requirements under varying climatic conditions (Davis and Dukes, 2010). The FAO-56 Penman-Monteith (PM) equation, the international standard for estimating reference evapotranspiration (ETref) depends on the accurate meteorological data, including air temperature, relative humidity, wind speed, and solar radiation (Allen et al., 1998). However, in many agricultural regions, such comprehensive weather measurements are not readily available at the field scale. Moreover, ETref represents potential atmospheric demand rather than the actual water use of individual plants, which varies with plant health, canopy structure, and soil water availability. This limitation becomes more pronounced under water stress as declining soil moisture results in stomatal closure thereby reducing transpiration. In contrast, sap flow measurements reflect the actual water uptake of plants, captures dynamic responses to environmental conditions such as solar radiation, vapor pressure deficit, and soil moisture status. Sap flow monitoring can provide early detection of plant water stress and support more precise irrigation scheduling by aligning irrigation timing and volume with actual plant water consumption. Some comparative studies between sap flow measurements and evapotranspiration models have shown that sap flow in isolated and orchard trees, was linearly related to the FAO-56 PM grass reference evapotranspiration (mm d⁻¹) (Fernández et al., 2001; Nicolas et al., 2005). Therefore, consideration of the relationship between evapotranspiration models and sap flow measurements is essential for improving irrigation scheduling strategies. Besides this, sap flow sensors integration with other irrigation technologies such as soil moisture sensors, weather station can significantly improve irrigation decision making. The multi-sensor approach enables more robust irrigation scheduling decisions based on soil moisture availability, plant water status, and atmospheric demand. Irrigation scheduling and decision support systems are used widely in horticultural crops. In container nursery production, decision support system has been used where irrigation practices from field scale Christmas tree production due to rootzone depth, soil heterogeneity and environmental conditions. Advanced irrigation decision support tools in other horticultural crops offers a framework which requires adaptation and validation in Christmas tree production in field scale. Therefore, there is need for species-specific irrigation thresholds, soil-plant-weather based sensor integration for decision support system for field grown Christmas tree production.

7.1 Water saving irrigation strategies using sap flow monitoring for conifer trees

Water-saving irrigation techniques are critical in regions where water resources are limited. Several plant-based indicators, including canopy temperature, stem water content, stem diameter variation integrated with sap flow, can be used to assess plant water status and optimize irrigation scheduling. In conifers, stem diameter variation is commonly applied due to its ability to continuously detect mild water stress (Herzog et al., 1995). Integrating stem diameter variation with sap flow measurements can improve water saving irrigation, especially in orchards with variable water stress conditions (Fernández and Cuevas, 2010). An automatic irrigation controller was used to calculate daily irrigation dose from sap flow readings in adequately irrigated trees relative to over-irrigated reference trees was able to keep soil moisture close to field capacity throughout the irrigation period in an olive orchard, avoiding both under- and over-watering (Fernández et al., 2008). In addition, in water-limited regions, hydrology-oriented practices such as thinning are proposed to improve tree growth and water use efficiency (WUE). Fernandes et al. (2016) evaluated the short- and medium-term effects of thinning on tree growth, drought sensitivity, and WUE (from growth and sap flow) in a semiarid Aleppo pine plantation. Thinning increased tree growth and reduced dependence on precipitation compared with control stands. While WUE measured from growth and sap flow increased markedly after thinning (Fernandes et al., 2016). In cotton, comparable sensor-feedback irrigation scheduling has shown to maintain yield while reducing applied water, with deficit-irrigated cotton plots at 75% soil-water replenishment saving a minimum of 16–20% of irrigation water across two seasons relative to fully irrigated plots (O’shaughnessy et al., 2023). Sap flow sensors do not reduce water use on their own; rather, they provide real-time physiological data that allow irrigation decisions to be better matched to actual plant demand, and it is this improved matching that drives gains in water-use efficiency and reduced plant stress. This integration approach improves the reliability of plant water monitoring and effective irrigation management (Manzanera et al., 2017).

7.2 AI and IoT integration for sap flow-based irrigation management

Recent technological innovations, such as Ribbonized sap flow (RSF) sensors, non-invasive micro-sensing methods, the integration of IoT components, Artificial Intelligence (AI), and machine learning, have introduced a new approach to sap flow measurement, overcoming persistent challenges. The RSF sensors, which integrate heaters and thermistors into a single flexible platform, directly address probe misalignment and installation errors while streamlining deployment (Jones et al., 2020). Wearable plant electronics further reduce invasiveness by adhering to stem surfaces rather than penetrating xylem tissue, thereby minimizing physiological disturbance and long-term signal degradation. These structural advancements substantially improve measurement reliability and facilitate the implementation of dense sensor networks.

The integration of IoT technologies introduces another transformative dimension. Real-time, cloud-connected systems enable continuous, autonomous monitoring of sap flow alongside microclimatic variables, allowing early detection of water stress and dynamic assessment of tree–atmosphere interactions (Baseca et al., 2025). Such connectivity improves temporal resolution, supports large-scale deployment through low-power communication networks, and facilitates the development of predictive models. By linking sap flow measurements with environmental data streams, these systems move beyond isolated measurements toward holistic ecosystem monitoring.

The use of AI, machine learning (ML), and deep learning frameworks represent a transformative direction in sap flow research, moving the field from sensor-based measurement toward data-driven inference and predictive ecohydrological modeling. These approaches enable automated data preprocessing, noise filtering, gap-filling, sapwood area scaling, and water-use estimation, addressing long-standing limitations associated with manual calibration and empirical correction. Machine learning algorithms can capture the non-linear relationships between sap flow and its environmental drivers such as temperature, humidity and light intensity (Marks et al., 2024). The SapFlower integrates random forest, Gaussian process regression, Long Short-Term Memory (LSTM) and Bidirectional Long Short-Term Memory (BiLSTM) for cleaning, gap-filling and scaling sap flow from TDP data (Wang and Renninger, 2025). Similarly, deep learning models based on LSTM networks have demonstrated strong performance in predicting hourly sap flow across diverse species and climatic regions in Europe, effectively scaling tree-level measurements to continental assessments (Loritz et al., 2024). Training these models on the extensive SAPFLUXNET dataset significantly improves their temporal and spatial accuracy, underscoring the importance of large-scale deep learning approaches and the continued expansion of global sap flow databases (Loritz et al., 2024).

Collectively, these emerging innovations directly respond to the major weaknesses of traditional sap flow techniques: invasiveness, calibration uncertainty, limited scalability, and insufficient integration with environmental datasets.

8 Limitations of sap flow sensors and associated issues of conifers anatomy

8.1 Anatomical limitations of conifers

In many conifers, the “radial profile” follows a Gaussian distribution, with peak flow occurring in the outer 2–4 cm of sapwood and declining toward the heartwood (Norton et al., 2025). However, in species like silver fir (Abies alba) and Norway spruce, researchers have noted that profiles can be “concave, “ where flow is lower near the cambium and higher at intermediate depths (Ford et al., 2004a). If radial variability is ignored and a uniform flow is assumed, stand-level transpiration can be overestimated by as much as 32% on sunny days and 40% over an entire growing season (Ford et al., 2004b). TDP methods do not consider the radial sap flow patterns while the HPV methods can explore radial changes of xylem sap flow. Using additional independent temperature sensors in HPV helps to read the heat injected by the short pulses dissipated in the vertical direction through mass flow and radial flow (Swanson and Whitfield, 1981). In addition, uncertainty in determining the zero-flow condition during nighttime sap flow, stem refilling, and internal water redistribution exists especially under irrigated conditions, warm nights, or advective environments (Peters et al., 2018). These processes violate the assumption of zero sap flow used to establish reference conditions in thermal-based methods, particularly TDP. Errors in defining zero flow propagate directly into sap flux calculations and introduce systematic bias, especially during periods of low transpiration or drought (Do and Rocheteau, 2002). Moreover, conifers exhibit strong resin-based defense responses to mechanical injury. Probe insertion induces traumatic resinosis, leading to the formation of resin ducts and accumulation of oleoresin around sensors (Nagy et al., 2000). Resin alters local thermal diffusivity and conductivity, causing signal drift and apparent declines in measured sap flow over time even when transpiration remains stable (Gartner and Heinrich, 2009). In addition, resin can physically block tracheids near the probe, reducing local hydraulic conductivity and further biasing sap flow estimates during long-term measurements (Golovin et al., 2022).

8.2 Limitations and considerations for sensors

Despite the advantages, certain practical challenges are faced under field conditions including improper sensors installation, probe misalignment, wounding of plant tissues, specie-specific wood anatomy, uneven distribution of sap wood, active sapwood area variability and diverse environmental conditions. The improper installation may sometime allows the rainwater to infiltrate thereby affecting xylem function. Long term field deployment also depends on reliable power supply which leads to data loss and affects the continuous data monitoring. In addition, many wireless sap flow monitoring systems depend on cellular networks for remote data transmission; however, limited or unreliable coverage in remote orchards can compromise data transfer reliability and may necessitate periodic on-site data retrieval. Also, the increasing trunk diameter with growth can affect sensor fit and measurement accuracy, requiring periodic sensor adjustment or replacement. Other technical and operational limitations of sap flow sensors are discussed in detail in the following sections.

Scaling a single point measurement from a sensor to the whole tree is a major source of error in coniferous trees. A single point measurement fails to capture radial and azimuthal variability in sap flux, inaccuracy in estimation of sapwood area, and limited spatial replication further accumulates bias during upscaling (James et al., 2002; Ford et al., 2004b). In addition, due to the non-linear relationship between sap flux density and the heat transfer coefficient, the Granier method requires a species-specific calibration (Smith and Allen, 1996) and should be performed within a defined portion of the sapwood. Conifer xylem exhibits strong thermal anisotropy and variable diffusivity; application of generic calibration parameters can lead to large errors unless species- and size-specific calibrations are applied. In many conifers, the TDP method has been found to underestimate sap flux density by 10% to 60% if species-specific calibrations are not applied (Lu et al., 2004). Moreover, continuous heating can cause thermal damage to delicate cambial tissues, leading to wound responses that permanently alter the accuracy of the probe (Forster, 2019). Drilling and probe insertion disrupt xylem continuity and reduce local hydraulic conductivity, creating a wound zone around the sensor.

9 Research gaps and future directions

The development of sap flow sensors for conifers must address several technical challenges such as radial variability of sap flow within thick sapwood and the wounding responses caused by invasive probes, which can degrade data quality over time. The empirical equations used in the calibration of sap flow sensors often lack robustness at the species level, particularly in conifers with diverse sapwood characteristics and thermal properties. This has made it difficult to compare results across sites and has also undermined the scaling of sap flow measurements to the whole tree and stand transpiration level. Additionally, large-scale deployment of traditional sensors is labor-intensive and logistically complex, which restricts the ability to achieve high-resolution spatial and temporal monitoring in forest ecosystems.

Several challenges must still be resolved before these technologies can be widely adopted in research and precision agriculture. Future studies should establish standardized benchmarking frameworks by directly comparing emerging non-invasive sap flow sensors with conventional thermal techniques across contrasting plant species, stem diameters, developmental stages, and environmental conditions (Vandegehuchte and Steppe, 2013). Multi-season field trials are also needed to quantify sensor drift, calibration stability, durability, energy demand, and data loss under realistic agricultural and forest conditions. In parallel, large, harmonized, and openly accessible sap flow datasets should be developed to support the training and validation of transferable artificial intelligence (AI) and machine-learning models rather than site-specific models with limited generalizability (Poyatos et al., 2016; Kabala et al., 2025). Greater emphasis should also be placed on integrating sap flow measurements with soil moisture, meteorological, remote sensing (Uni et al., 2025), and plant hydraulic data so that sensor outputs can be assimilated into crop, hydraulic, and land-surface models. Future research should therefore prioritize cross-comparison studies, multi-season performance evaluations, and integration with hydraulic and land-surface models.

10 Conclusions

Christmas trees, as a valuable commercial conifer crop, highlight the critical need for effective irrigation management. Proper irrigation scheduling is essential for conserving water and improving water-use efficiency. Sap flow evaluation has been used as a valuable approach for monitoring plant water use. The evolution of Christmas tree irrigation from empirical to precision management is supported by a robust body of research demonstrating the value of sap flow sensors as a diagnostic tool. While gymnosperm anatomy poses unique challenges for sensor calibration and scaling, the use of pulsed heat methods like HRM and species-specific regression constants have largely overcome these technical hurdles. The integration of IoT platforms and low-cost sensor designs is finally making plant-based monitoring economically feasible for commercial plantations. Ultimately, the adoption of these technologies not only improves the resilience of Christmas trees to climate-induced drought but also enhances the sustainability of industry by protecting valuable water and soil resources from mismanagement and disease.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Wang Xiukang, Yan’an University, China

Reviewed by: María R. Conesa, Spanish National Research Council (CSIC), Spain

Fazli Wahid, The University of Agriculture Swat (UAS), Pakistan