Plant and substrate-based indices for salinity monitoring in Cestrum nocturnum

This study proposes salinity indices based on plant and substrate measurements to define reference thresholds for salinity management in potted crops, using Cestrum nocturnum as a model species. A greenhouse experiment was conducted with plants grown in containers and irrigated with nutrient solutions at three electrical conductivity (EC) levels (2.0, 4.5, and 7.0 dS m - ¹). Plant responses were assessed through vegetative growth, visual quality, flowering intensity, continuous stem diameter variation (maximum daily stem shrinkage, MDS), cumulative evapotranspiration (ETa), and substrate bulk EC monitored with sensors. Increasing salinity reduced vegetative growth, particularly shoot biomass, while enhancing flowering intensity at 4.5 dS m - ¹, indicating a shift from vegetative to reproductive development. The moving average of MDS (avgMDS) responded to salinity, showing both increases and decreases depending on stress intensity, and, when expressed as signal intensity (SI: control/salinity), discriminated between stress levels, establishing alert (1.10) and critical (1.38) thresholds. Salinity decreased ETa by 35% and 65% at 4.5 and 7.0 dS m - ¹, respectively, and ETa-based SI defined alert (1.20) and critical (1.55) thresholds. The hourly moving average of bulk EC (avgECb) enabled continuous assessment of salinity dynamics, minimizing the influence of substrate moisture variability. The use of avgMDS, ETa, and avgECb enables the detection and interpretation of salinity stress by integrating plant physiological responses with substrate conditions, while the combined use of two or more indices improves the robustness of the assessment, providing a quantitative framework for salinity management in potted crops.

The availability of freshwater for irrigation is becoming increasingly scarce, forcing the use of low-quality water for agricultural production, especially saline water. Salinity stress is one of the major challenges to crop sustainability worldwide. Zhang et al. (2025) highlight that soil salinization, exacerbated by climate change, requires the development of precise monitoring strategies to mitigate its effects on plant growth. In this context, nursery production of ornamental plants faces the challenge of maintaining a marketable production within a sustainable cultivation approach. For this purpose, in addition to assessing plant tolerance to salinity, it is essential to design cultivation strategies that improve this tolerance. Among them, efficient irrigation management allows leaching of salts from the root zone, optimizing water use. However, the leaching of salts implies a high economic cost due to the loss of water and nutrients (Oron et al., 2002).

Furthermore, this may have an environmental impact due to contamination of the environment with fertilizers and pesticides. This is why it is important to precisely determine the leaching fraction and timing of application. On the other hand, efficient management of saline irrigation requires the implementation of automatic systems capable of continuously monitoring the physiological responses of plants to high salt concentrations in the substrate. Plant growth is a highly salinity-sensitive indicator, accurately reflecting the plant’s physiological responses to salt stress, including both osmotic stress and ionic damage.

Some studies in fruit trees have shown that stem diameter variations (SDV) are well related to the water status of the plant, and their measurement can facilitate efficient irrigation management (Fernández and Cuevas, 2010). Although dendrometric studies on ornamental shrubs are scarce, Miralles et al. (2010) also demonstrated that SDV is a sensitive indicator of water status. SDV are recorded using dendrometers, from which various indicators are obtained, such as maximum and minimum daily diameters, maximum daily stem shrinkage (MDS) and radial growth rate. Although all these parameters reflect the physiological status of the plant, numerous studies have shown that MDS is particularly sensitive for assessing water stress in fruit trees (Cohen et al., 2001; Ortuño et al., 2006; 2010). Gallardo et al. (2006) highlighted the sensitivity of MDS as an indicator of water status in mature tomato plants when evaporative demand is high.

The close relationship between SDV and the degree of water stress suffered by the plant also suggests their sensitivity when saline water is applied, particularly the osmotic component, which affects water availability to the plant. The relationship between SDV and salt stress has been confirmed in species such as Pistacia lentiscus (Castillo-Campohermoso et al., 2020) and Vigna radiata (Khan et al., 2022), where the osmotic response to the presence of salts affects stem diameter in a similar way to water deficit. However, SDV is strongly affected by environmental factors, especially vapor pressure deficit (VPD) (Goldhamer and Fereres, 2001), which may limit its usefulness as an index of salinity level exclusively. Consequently, the same MDS value can correspond both to a healthy plant exposed to high VPD and to a stressed plant under low VPD conditions (Ortuño et al., 2010). This limitation is more pronounced under greenhouse conditions, where VPD fluctuations are greater than in the open air due to ventilation or humidification, which reinforces the need to look for indices that are less dependent on evaporative demand. In this sense, we propose that the moving average of MDS, by integrating environmental variations, provides a more stable and accurate indicator of salt stress.

As mentioned above, the accumulation of salts in the substrate reduces the osmotic potential, hindering water absorption by the roots and altering the plant’s water balance. As an initial response, the plant activates water conservation mechanisms, such as stomatal closure, which reduces stomatal conductance, limits the entry of water into the vascular system and reduces transpiration (Wang et al., 2025a). This behavior has been described in several species such as pepper, papaya and carnation (Kwon et al., 2019; Hailegnaw et al., 2025). Cerrato et al. (2024) demonstrated in geophytes that differences in evapotranspiration (ET) reduction allow discriminating levels of salinity tolerance, which reinforces its use as a diagnostic tool. Therefore, reduced transpiration represents one of the first physiological responses to osmotic stress, and its magnitude depends on factors such as species, salt concentration and duration of stress (Xing et al., 2021). The use of automatic scales allows the calculation of daily evapotranspiration (ET), from which cumulative ET (ETa) is obtained, an indicator with potential for early detection of the plant’s physiological response.

To estimate the salinity level, the electrical conductivity (EC) of the substrate can be measured. Nowadays, there are several sensors that allow this measurement to be made directly, easily and immediately. However, these sensors record the bulk EC (ECb), which represents the combined EC of the substrate, air and soil solution (Sodini et al., 2024). Monitoring substrate electrical conductivity (ECb) is key to understanding ion dynamics in the root zone. Wang et al. (2022) highlight that sensors enable real-time assessment of salt accumulation, supporting precision irrigation strategies to mitigate osmotic stress. However, as ECb is very sensitive to substrate moisture (Barnard et al., 2025), its value can be easily altered, which limits its usefulness as a reliable indicator of salinity. In potted crops, this sensitivity is accentuated due to the water characteristics of the substrates, which cause sudden fluctuations in moisture and, consequently, a greater variability in ECb. Under these conditions, ECb is no longer a reliable indicator of salinity. This sensitivity of ECb to moisture has prompted the development of methods to mitigate such interference. For example, Incrocci et al. (2019) showed that ECb correlates more closely with salinity when substrate moisture is high. Similarly, Valdés et al. (2014) found that, in Poinsettia, ECb measured immediately after irrigation acts as a good indicator of salinity. Also, in potted Euphorbia sp. plants, Bañón et al. (2022a) evaluated different salinity indices based on ECb measurements taken before and after irrigation, as well as their average between irrigation events. In this context, we hypothesize that the moving average of hourly ECb values may constitute an effective index of substrate salinity, as its ability to integrate moisture fluctuations would allow for a more accurate and stable estimation of salinity over time.

Cestrum nocturnum L., a fast-growing perennial ornamental shrub widely cultivated for its nocturnal flowering and intense aroma (Shaista and Amrita, 2016), was selected as experimental plant material due to its commercial importance, its vigorous growth, which enables efficient monitoring of physiological and dendrometric parameters, and the scarcity of studies on its agronomic management, particularly regarding salt stress, which makes this study a novel contribution to the current scientific literature. The aim of this study is to evaluate the effectiveness of two plant-based salinity indices for the early detection of osmotic stress in Cestrum nocturnum: i) one derived from daily variations in stem diameter, recorded using dendrometers, and ii) another derived from the calculation of daily accumulated evapotranspiration, obtained using automated scales. In addition, a complementary index of substrate salinity is analyzed, calculated from the ECb measured with soil sensors. In parallel, the impact of salinity on growth and development is evaluated in order to link it to the proposed indices and to establish physiologically relevant salinity thresholds.

Plant material, growing conditions, and treatments

Seedlings of Cestrum nocturnum L. produced in a local nursery (Viveros Bermejo S.L., Totana, Spain) were used. To ensure uniformity, plants with a height of 10-12 cm were selected and transplanted into 2.5 L black plastic pots containing a substrate composed of peat, coconut fibre, and perlite in a 60/30/10 (v/v/v) ratio. The experiment was carried out in a greenhouse covered with polycarbonate located at the experimental farm of the Polytechnic University of Cartagena (Cartagena, Murcia, Spain; 37.683°N, 0.951°W). Plants were distributed in the greenhouse following a randomized complete block design, with three blocks of ten plants each, totaling 30 plants per treatment. Transplanting was carried out in the third week of January, and the experiment was concluded in the fourth week of April 2024.

The pots were fitted into the holes of metal grids, supported on concrete blocks that kept the pots 80 cm above the ground. Air temperature and relative humidity inside the greenhouse were monitored using a LOG 32 TH datalogger (Dostmann Electronic GmbH, Wertheim, Germany). These parameters were used to calculate the VPD based on the air temperature and the saturation vapor pressure at that temperature. The daily fluctuations of both mean VPD and temperature throughout the experimental period are presented in Figure 1.

Three saline treatments were established based on the electrical conductivity (EC) of the irrigation solution: a control (2.0 dS m - ¹), moderate salinity (4.5 dS m - ¹), and high salinity (7.0 dS m - ¹). The control solution was prepared by adding a liquid fertilizer (4-1.7-4.5-4-4-1.4 of N-P 2 O 5 -K 2 O-CaO-MgO; Vegetator, Transferencia Agrotecnológico S.L., Murcia, Spain) to the irrigation water (1.2 dS m - ¹) until the EC reached 2.0 dS m - ¹. The moderate and high salinity treatments were then reached by adding sodium chloride (NaCl) to this control nutrient solution until the target ECs of 4.5 and 7.0 dS m - ¹ were achieved, respectively. NaCl was added to the irrigation water on 14 February. The pH of all solutions was adjusted to 6.0 using nitric acid.

Irrigation management

A drip irrigation system with a self-compensating, anti-drainage emitter (Netafim Ltd., Tel Aviv, Israel) of 2.3 L h -1 per pot was used. The water was distributed through a tube of 4 mm diameter and 40 cm length, ending in a plastic arrow inserted into the substrate. There were three independent irrigation sectors, one per treatment. Irrigation was managed by Teros-12 sensors (METER Group, Inc., Pullman, WA, USA) inserted in the pots. Irrigation relays (AZ8222-2C-5DME, Zettler Components Inc., Aliso Viejo, CA, USA) connected to a control multiplier port (SMD-CD16D, Campbell Scientific, Inc., Logan, Utah, USA) were used to activate the valves of the 1000-litre tanks assigned to each treatment. Irrigation instructions were programmed with LoggerNet 3.x software (Campbell Scientific, Inc., Logan, UT, USA).

The CR1000 datalogger monitored the volumetric water content (VWC) of the substrate every hour and, daily at 11:00 h, checked if it was below the set threshold (36%). If this criterion was met, a 20-minute irrigation was triggered; otherwise, the system waited until the next day for a new check. VWC was estimated using a substrate-specific calibration equation developed from permittivity and moisture content measurements according to the methodology of Valdés et al. (2014).

The total amount of water applied was estimated considering dripper flow rate and irrigation time. Leachate was collected weekly in plastic containers under the pots, and its volume and EC were determined. The leaching fraction (LF) was quantified as:

where V leachate is the volume of leached water and V applied is the total volume of water applied to the pot. Water consumption (WC) was calculated as:

Teros-12 dielectric soil sensors were used, which were selected for their ability to simultaneously measure permittivity, temperature and bulk EC (ECb), and for their robustness under high salinity conditions (Qi et al., 2024). The sensors were inserted vertically into randomly selected pots, always at the same cardinal position and distance from the emitter. There was one sensor per pot, and three sensors per treatment. The Teros-12 were connected to a CR1000, programmed to record their outputs every hour, such as ECb.

Determination of stem diameter variations and evapotranspiration

Stem diameter variations (SDV) were monitored using DD-S1 small-diameter dendrometers (Ecomatik GmbH, Dachau, Germany). Three units per treatment were installed on the main stem of the plants at 5 cm above the substrate. The dendrometers were placed on the plants on 23 February, obtaining the maximum and minimum values of stem diameter during 24-hour cycles. Maximum daily stem shrinkage (MDS, µm) was calculated as the difference between the daily maximum and minimum values according to Ortuño et al. (2010). The dendrometers were connected to the CR1000 datalogger, programmed with LoggerNet 3.x software to record hourly measurements.

Programmable balances (Sartorius Analytica, model 52015201; Sartorius AG, Göttingen, Germany) were used to monitor the weight of the pots during the experiment. Each balance has a maximum capacity of 5.2 kg and a resolution of 0.01 g. One balance per pot was used, with a total of three balances per treatment. The scales were connected to a CR1000 configured using LoggerNet 3.x software to automatically record the weight every hour. The connection between the scales and the CR1000 was made via an RS-232 DB-25 cable.

Salinity indexes

Continuous and automatic salinity indices were evaluated, both in the plant and in the substrate. One of them was designed to estimate water stress from the maximum daily stem shrinkage (MDS), calculating a moving average (avgMDS, µm) from the start of irrigation to any time of the crop. Signal intensity (SI, unitless), defined as the ratio between avgMDS under control and saline conditions (SIavgMDS, unitless) was used as a relative indicator of salt stress. SI values greater than 1 indicate the presence of stress, and an increase in SI values reflects a greater physiological deterioration of the plant. The use of SI allows the establishment of quantifiable thresholds of salinity tolerance, facilitating the physiological interpretation of stress in plants. Furthermore, by normalizing avgMDS values between control and salinity, this approach improves comparability between different climatic environments and phenological stages of the crop, making it a robust tool for monitoring salt stress (Naor and Cohen, 2003).

The second index was based on daily evapotranspiration (ET, mm), calculated from the difference in pot weight between consecutive days. On irrigation days, the volume of water retained was discounted. From the daily values, cumulative ET (ETa, mm) was calculated as the moving sum of daily ET over the experimental period. As with MDS, the signal intensity for ETa (SIETa, unitless) was determined to quantify the response of ETa to salinity relative to the control.

To characterize the salinity conditions to which the plants were exposed, a salinity index based on the ECb, obtained directly from the Teros-12 was used. The hourly moving average of ECb values (avgECb) is proposed as an index to integrate the temporal variability of substrate moisture into the interpretation of ECb, providing a continuous and dynamic estimate of salt accumulation.

Growth and visual quality

At the end of the experiment (68 days after salt treatment), growth parameters were determined on five plants per treatment. To calculate the dry weight (DW, g), the shoot was separated from the roots, which were washed with pressurized water using a hose with a flat nozzle. Leaves, stems and roots were dried at 80 °C to constant weight. The number of leaves and shoots were counted and shoot length (cm) was measured. Leaf area (dm 2) was calculated using Easy Leaf Area software (Easlon and Bloom, 2014), based on digital image analysis. In addition, the percentage of defoliation and leaf necrosis was determined, as well as the presence of inflorescences expressed as the percentage of flowering plants (%).

The aesthetic quality of the plants was assessed by means of a visual index on a scale of 0 to 4, considering four parameters: defoliation, necrosis, leaf chlorosis, and flowering. This approach aligns with established protocols for evaluating salt-stressed ornamentals, where visual integrity and the absence of leaf injury are the primary determinants of commercial value (Wang et al., 2025a). A value of 4 indicates good quality (control), while values below 2 correspond to non-commercial plants. Defoliation was scored according to the percentage of fallen leaves: >10% (1), 10-5% (2), 5-2% (3) y<2% (4). Leaf necrosis was assessed as a percentage of total leaves: >20% (1), 20-10% (2), 10-3% (3) y<3% (4). Chlorosis was determined according to relative chlorophyll content (SPAD) values:<30 (1), 30-35 (2), 35-40 (3) y >40 (4). Flowering was assessed according to the percentage of plants with inflorescences:<50% (1), 50-60% (2), 60-70% (3) y >70% (4). The final score was calculated as the average of the four parameters assessed.

Statistical analysis

The data obtained were analyzed by one-way analysis of variance (ANOVA) using Statgraphics Centurion XVI software (StatPoint Technologies, Warrenton, VA, USA). Before conducting the ANOVA, Levene’s test was conducted to assess the homogeneity of variances, confirming that this assumption was met (p > 0.05 for all variables). When significant differences (p < 0.05) were detected, the Least Significant Difference (LSD) test was applied to compare the means of the treatments. All graphics were produced using SigmaPlot v. 14.5 (Systat Software Inc., San Jose, CA, USA).

Salinity and irrigation

Salinity significantly altered irrigation parameters in pot-grown Cestrum nocturnum plants, decreasing their water consumption and altering the balance between water uptake, retention and drainage (Table 1). As salinity in the nutrient solution increased, a decrease in the total volume of water applied was observed, with reductions of 34% (EC4.5) and 56% (EC7.0) compared to the control. These reductions were related to a lower frequency of irrigation, suggesting a decrease in water demand by these plants. This behavior is attributed to the osmotic effect, which reduces water availability, and to ionic toxicity caused by salt accumulation in the tissues (Zhu, 2001; van Zelm et al., 2020). In parallel, volumetric water content (VWC) was higher in EC7.0 (Table 1) as a result of lower water extraction by roots and higher salinity-induced moisture retention in the substrate. Although the total leachate volume decreased, the leaching fraction was higher in the salinity treatments (Table 1). This indicates that proportionally more water is lost through drainage under saline conditions, which is a good strategy to avoid salt accumulation in the root zone (; Grieve et al., 2012; Liu et al., 2024; Wang et al., 2025b). The decrease in applied water and leachate led to a reduction in water consumption of 40% and 59% in the treatments with EC4.5 and EC7.0, respectively. The leachate EC increased progressively with the salinity of the irrigation, which was expected and confirms the continuous accumulation of salts in the substrate. The results indicate that developing a sustainable management strategy for saline water in potted crops requires a systematic study of the factors contributing to salinity parameters for water-saving irrigation approaches. Substrate and plant salinity indexes.

Substrate and plant salinity indexes

The moving average of the daily maximum stem diameter shrinkage (avgMDS)

The moving average of daily maximum stem diameter shrinkage (avgMDS) followed a polynomial trend in spite of being influenced by variations in VPD and substrate moisture, resulting in smoother daily fluctuations compared to individual MDS values (Figure 2A), which are highly sensitive to environmental changes (Gallardo et al., 2006). At the beginning of the study, all treatments exhibited a strong increase in avgMDS, primarily due to limited initial data and vigorous plant growth. By early March, avgMDS values under saline stress conditions declined significantly compared to the control, revealing early symptoms of salt stress. The EC7.0 treatment showed an earlier and more pronounced decline (around March 10) (Figure 2A), highlighting avgMDS as a sensitive indicator of salt-induced physiological responses. This reduction was attributed to a loss of stem elasticity and/or reduced water flow, both of which limited radial growth under saline conditions. Although previous studies have shown that MDS may increase under moderate water stress and decrease under severe stress (Cohen et al., 2001; Moriana et al., 2003; Miralles et al., 2010), avgMDS in this study decreased consistently with increasing salinity (Figure 2A), aligning with findings in Pistacia lentiscus (Castillo-Campohermoso et al., 2020).

The avgMDS curve for EC4.5, which did not compromise plant commercial viability (see Table 2), may represent a tolerable salinity threshold. Importantly, inflection points in avgMDS trends can serve as early warning indicators of salt stress without requiring a control treatment. However, when such changes are not evident, the use of a control and the calculation of signal intensity (SIavgMDS) are necessary to quantify crop responses accurately. As shown in Figure 2B, SIavgMDS provides a clear quantification of stress progression over time, allowing the identification of both early (SI > 1.10) and critical (SI > 1.38) salinity thresholds. These thresholds are valuable for informing salinity management in Cestrum nocturnum cultivation and preventing irreversible damage to plant productivity.

Cumulative daily evapotranspiration (ETa)

Cumulative daily evapotranspiration (ETa) increased throughout the growing season, but its rate and magnitude were influenced by the salinity level of the nutrient solution (Figure 3A). In the EC7 treatment (high salinity), ETa showed a significant reduction three weeks after salt application, marking an early physiological response to salt stress. This supports previous findings in other crops such as grapevine and papaya, where evapotranspiration decreases before visible symptoms appear (Shani and Ben-Gal, 2005; Hailegnaw et al., 2025). In the EC4.5 treatment, ETa reduction was less pronounced and occurred later, becoming significant after five weeks (Figure 3A). By the end of the cycle, the ETa reduction reached 35% in EC4.5 and 65% in EC7.0 compared to the control, indicating the cumulative nature of salinity stress. Similar salinity-induced reductions in transpiration have been documented in other ornamental species (Bai et al., 2021; Xing et al., 2021). In Dianthus caryophyllus, restricting transpiration has been identified as a key salt-tolerance mechanism (Kwon et al., 2019). Furthermore, Cerrato et al. (2024) showed contrasting evapotranspiration resilience between Pancratium maritimum and Eryngium maritimum, highlighting species-specific salinity adaptations.

A 35% ETa reduction in EC4.5 is proposed as a critical salinity threshold for Cestrum nocturnum. In EC7.0, this threshold was crossed around 11 March (Figure 3A, black arrow), suggesting that corrective measures should have been implemented at that time to avoid reaching the final 65% ETa reduction, which is associated with severe morphological and physiological damage (Parida and Das, 2005; Zhang et al., 2025). Unlike avgMDS, which shows distinct inflection points, ETa trends only show slope changes, which makes it more sensitive for early stress detection. The signal intensity of ETa (SIETa) was studied too, which allowed earlier and clearer differentiation between treatments, beginning as early as the first week (Figure 3B). In EC4.5, SIETa progressively increased, reaching a critical value of 1.55, which corresponded to 13 March in EC7.0. By the end of the experiment, SIETa in EC7.0 reached 2.90, indicating severe physiological stress.

Alert thresholds derived from ETa trends correspond to SIETa values of 1.21 for EC4.5 and 1.47 for EC7.0 (Figure 3B, green and red arrows). These values provide practical early-warning references for salinity stress management.

Hourly moving average of the bulk EC (avgECb)

Figure 4A illustrates hourly bulk EC (ECb), showing strong fluctuations primarily driven by moisture variations (Bañón et al., 2021). While ECb patterns differ between treatments, they are too unstable to reliably indicate salinity conditions. Peaks in ECb correspond to irrigation events, which were more frequent in the control due to higher plant growth (Table 2). In contrast, EC7.0 plants, under greater salinity stress, required less irrigation.

After irrigation, ECb values declined, potentially suggesting lower salinity. However, this is misleading: salts remain in the substrate, and their concentration rises as moisture decreases. This highlights a key limitation of instantaneous ECb measurements, due to the confounding influence of substrate moisture (Amente et al., 2000). Additionally, greater salinity caused larger ECb oscillations, indicating a synergistic interaction between moisture and salt levels (Bañón et al., 2021). As a result, a single ECb value can reflect different moisture-salinity combinations, limiting its standalone interpretability.

To address this, the hourly moving average of ECb (avgECb) was used (Figure 4B). This parameter smooths moisture variability and reveals long-term trends in salt accumulation. In the control, avgECb remained relatively stable (from 0.59 to 0.74 dS·m - ¹), while EC7.0 showed a continuous increase to 1.79 dS·m - ¹. EC4.5 also showed an upward trend, ending at 1.39 dS·m - ¹.

The final avgECb value for EC4.5 (1.39 dS·m - ¹) is proposed as the critical salinity threshold for Cestrum nocturnum, since commercial plant quality was maintained at this level. Exceeding this threshold, as observed in EC7.0, could lead to productivity losses and increased stress. In fact, this threshold was crossed in the EC7.0 treatment on March 16 (Figure 4B, arrow), suggesting that irrigation or leaching interventions should have been applied at that time.

While avgECb is useful for monitoring salt accumulation, it offers limited predictive capacity compared to physiological indicators such as avgMDS or ETa, as its trends alone do not provide early warning signs of stress unless used in conjunction with physiological parameters.

Effect of salinity on growth and visual quality

Salinity significantly affected the growth of Cestrum nocturnum, reducing dry weight by 42% and 67% under EC4.5 and EC7.0, respectively (Table 2). This decline reflects limitations in water and nutrient uptake, a well-established indicator of salt stress tolerance (Isayenkov and Maathuis, 2019). Similar effects have been described in other Cestrum species, with C. aurantiacum and C. fasciculatum showing 25-50% reductions in aerial dry weight under moderate salinity (Cassaniti et al., 2009). Based on these results, Cestrum nocturnum could be classified as moderately salt-tolerant. In ornamentals, reduced vegetative growth does not always compromise commercial value, as aesthetics may remain acceptable (Fornes et al., 2007).

Root growth was less affected than the aerial part, resulting in a decreased shoot/root ratio (Table 2), in line with the higher salt tolerance of roots reported by Bernstein and Kafkafi (2002). This strategy helps maintain water uptake under stress (van Zelm et al., 2020).

Plant architecture was also altered by salinity. While shoot number remained stable under EC4.5, shoot length declined by 18%. At EC7.0, both shoot number and length decreased by 20% and 40%, respectively. Interestingly, salinity stimulated flowering: EC4.5 and EC7.0 increased the percentage of flowering plants by 42% and 28%, respectively. This response, also reported in other ornamental shrubs (Bañón et al., 2022a, b), may represent an adaptive mechanism to ensure reproductive success under adverse conditions.

Salinity reduced leaf area by limiting both leaf number and blade size. Leaf area declined by 38% and 63% under EC4.5 and EC7.0, respectively (Table 2), with defoliation reaching 19% in EC7.0. Defoliation is considered a tolerance mechanism, reallocating resources to the remaining leaves (Acosta-Motos et al., 2017). EC7.0 also caused necrosis and chlorosis, associated with ionic toxicity and chlorophyll degradation (Sun et al., 2020). SPAD readings confirmed a reduction in relative chlorophyll content under salt stress (Table 2).

The visual quality index dropped significantly with increasing salinity, from 4.00 in the control to 3.13 in EC4.5 and 1.63 in EC7.0. The marked aesthetic deterioration at EC7.0 compromises marketability, whereas EC4.5 plants retained acceptable visual traits. Thus, C. nocturnum can tolerate up to 4.5 dS m - ¹ without losing commercial viability, but higher salinity levels are unsuitable for its production.

Cestrum nocturnum exhibits intermediate salinity tolerance (4.5 dS m - ¹), characterized by moderate growth reduction and stimulated flowering. However, levels of 7.0 dS m - ¹ significantly impair ornamental value through leaf necrosis, chlorosis, and defoliation.

Plant-based monitoring indices provide a precise assessment of the crop’s physiological status. The moving average of daily maximum stem shrinkage reduces variability induced by environmental factors and substrate moisture, allowing for stress detection without a control reference. In contrast, cumulative evapotranspiration effectively reflects the cumulative impact of saline stress, proving to be a valuable tool for establishing early warning thresholds. Expressing these physiological parameters as signal intensity simplifies the identification of salinization thresholds and offers broader applicability than absolute values.

Calculating the hourly moving average of substrate bulk electrical conductivity increases signal stability by attenuating fluctuations associated with moisture variations, although its sensitivity for early stress detection is lower than that of plant-based indicators.

The integration of stem diameter variations, cumulative evapotranspiration, and the hourly moving average of substrate bulk electrical conductivity captures different dimensions of salt stress, providing a robust quantitative framework for optimizing saline irrigation management and improving the production of ornamental crops grown in containers.