The field experiment was conducted from May to September 2021 in Chayouzhong County, Yinshan Mountain area, China (41°30′ N, 112°64′ E), which has a temperate continental monsoon climate, an average annual temperature of 1.3 °C, an elevation of 1,780 m, a frost-free period of approximately 100 days, and annual evapotranspiration greater than 2,000 mm. The soil at the experimental sites is sandy-loamy, with 26.3 g kg^-1 of organic matter, 2.24 g kg^-1 of total nitrogen, 10.3 mg kg^-1 of available P (Olsen-P), 69.8 mg kg^-1 of exchangeable K, pH of 7.9, and maximum field water holding capacity of 25.61%. The effective precipitation during the potato growth period was 236.6 mm. A small automatic weather station (Spectrum Watchdog 2900ET, USA) was installed in the experimental field to continuously and automatically detect rainfall, wind speed, air temperature, and other meteorological parameters during the potato growth period. The specific meteorological data are presented in Figure 1 .

Five irrigation levels (total irrigation amount during potato growth), namely, 0 mm (W₀), 75 mm (W₁), 150 mm (W₂), 225 mm (W₃), and 300 mm (W₄), were used in this experiment, and the specific irrigation regime (irrigation timing and amount of each irrigation event) for each treatment is shown in Table 1 .

The virus-free seed potato cultivar Kexin No.1 was used in the experiment. The plants were sown in May 2 and harvested in September 9. Drip irrigation was provided with a 0.3-m dripper spacing, and 2.2 L/h dripper flow was used. All of the treatments were arranged in a randomized block design with three replicates. The plot area was 90 m², with 30 cm between potato plants and 90 cm between rows. Urea (300 kg N ha⁻¹) was used as N source, and 30% of it was basal dressed and the 70% was fertigated through drip irrigation. Phosphorus fertilizer (79 kg P ha⁻¹) and potassium fertilizer (224 kg K ha⁻¹) were both broadcasted as basal fertilizers.

The SPSS 25.0 statistical software package was employed to assess the normality of these data and conduct tests for homogeneity of variance. ANOVA with LSD test (P < 0.05) and regression analysis were performed, Pearson correlation coefficients were calculated to assess linear associations between normally distributed variables.

At the early growth stage, the plant dry weights did not show a significant difference among irrigation levels. Afterwards, both plant dry weight and leaf area index (LAI) of potatoes exhibited significant increases in response to higher irrigation rates (p < 0.05) ( Figure 2 ).

Potato yield and its contributing factors demonstrated a remarkable response to irrigation under low-irrigation conditions (from W₀ to W₂). As the irrigation level increased within this range, there was an increase in tuber yield, the percentage of commodity potatoes, tuber number per plant, and the weight of tubers per plant. However, once the irrigation level surpassed 150 mm (W₂), no further increases were observed in either the yield or its contributing factors, as shown in Table 2 .

The irrigation level significantly affected LAD across all measured leaves (L1, L4, and L8). Compared to the non-irrigated treatment (W0), all irrigated treatments (W1–W4) resulted in higher LAD for each leaf. With increasing irrigation levels, the LAD of each leaf showed varying degrees of increase depending on the growth stage—for instance at 35 DAE, the LAD of L₄ under W₄ was 15% higher than that under W₁. By 60 DAE, the LAD under W₄ had increased by 38% compared to that under W₁ ( Table 3 ).

Statistical analysis indicated a significant positive correlation between the LAD and the irrigation level. At each growth stage, the correlation between the LAD of L₄ or L₈ and irrigation level reached a significance level of p < 0.01. Moreover, the correlation coefficient (R) between the LAD and irrigation level was greater for L₄ than for L₁ or L₈ ( Table 4 ).

Under the same irrigation level, the LADs of L₁ exhibited the greatest variation among plants. The variation coefficient of LAD throughout the growth period was L₄ < L₈ < L₁ ( Table 5 ).

Further statistical analysis showed that there was a significant binomial regression relationship (p < 0.05) between the potato yield and the LAD of each leaf at each growth stage, as shown in Figure 3 . All of the regression equations clearly demonstrate that as the LAD increased, the potato yield also increased. However, once LAD reached a critical point, further increments in LAD no longer resulted in increased potato yield.

There were significant linear relationships (p < 0.05) between the LAD of each leaf and the plant water content (PWC) across different growth stages. As the plant water content increased, the LAD exhibited diverse degrees of increase depending on leaf positions. At all growth stages, the coefficient of determination (R ²) of the regression for L₄ was higher than that for other leaf positions ( Figure 4 ).

Just as in the case of the relationship between LAD and PWC, a positive correlation (p < 0.05) between LAD and leaf relative water content (LRWC) was discerned at each growth stage. Throughout each growth stage, the correlation coefficients for L₄ ranged from 0.48 to 0.59, being greater than those for L₁ and L₈ ( Table 6 ).

As the soil water content increased, the LAD increased to varying degrees depending on leaf positions—that is, the LADs of L₁, L₄, and L₈, each to a different extent, reflected the soil water content status. For each leaf position, significant regression (p < 0.05) between LAD and soil water content was discerned. However, the R ² of the regression for L₄ was higher than that for other leaf positions at all growth stages, while the R ² value for L₁ was the lowest ( Figure 5 ).

In this study, we systematically measured LAD at L₁, L₄, and L₈ of potato plants across different growth stages and irrigation levels. The significant positive correlations between LAD and irrigation levels observed for each potato leaf position ( Tables 3 , 4 ) clearly indicate that the LAD of potato responds highly to water supply. Notably, even the first leaf from the plant top (L₁), which is typically incompletely expanded and exhibits substantial morphological variation among plants (Fan, 2019), demonstrated a significant positive correlation with irrigation level ( Table 4 ). The relative water content of leaves (LRWC) is the best index to reflect the water balance of plants (Torres et al., 2019). In this study, the significant correlation observed between LAD and the corresponding leaf’s LRWC ( Table 6 ) indicates that LAD could be an indicator of the water status in potato plants. The positive relationships between LAD and plant water content as well as between LAD and soil water content ( Figures 4 , 5 ) validate that LAD can serve as a reliable indicator to monitor the water status of potato plants—that is, by measuring LAD, the water status of potato plants can be assessed.

The threshold LAD value, defined as the minimal LAD required to achieve maximum yield, is critical for developing LAD-based irrigation protocols. The yield saturation effect observed in the study (in which exceeding a critical LAD value no longer results in higher potato yields) and the significant binomial regression relationship between LAD and potato yield (R ² > 0.92, p < 0.05) ( Figure 3 ) provide two key operational advantages for precision irrigation in potato cultivation: (1) threshold-based irrigation triggering when LAD exceeds optimal ranges and (2) predictive yield modeling through continuous LAD monitoring and thereby further highlighting the substantial agronomic significance of monitoring LAD.

Previous studies have demonstrated that the leaf position has a significant impact on the water status diagnosis accuracy in gramineous crops through temperature measurement (Grant et al., 2016; Wang et al., 2023)—for instance, the LAD at two-thirds of the height of a maize plant could better reflect the moisture characteristics of the crop and soil (Wang et al., 2006), while in sorghum, the three leaves from the top of the plant were found most suitable for leaf temperature monitoring (Wang et al., 2019). They underscore the critical importance of leaf position selection in plant water status diagnostics. As potato has complex pinnately compound leaves and inherent intra-plant variability, precise leaf positioning becomes particularly essential for accurate water assessment. The results that the L₄ exhibited significantly higher correlation coefficients between LAD and irrigation levels compared to L₁ and L₈ ( Table 4 ) suggest a heightened sensitivity of L₄ LAD to irrigation. Furthermore, the significantly higher R ² values for L4 than for other leaf positions in modeling relationships between LAD and both PWC and SWC ( Figures 4 , 5 ), along with the strongest correlations between LAD and LRWC ( Table 6 ), collectively indicate L4’s superior responsiveness to dynamic plant water status changes. Notably, intra-treatment variability analysis shows that L₄ maintained a significantly lower coefficient of variation (CV) in LAD measurements compared to L₁ and L₈ under identical irrigation regimes ( Table 5 ). This is likely attributed to L₄ being the first fully expanded mature leaf with stable morphological structures and physiological functions, whereas L₁ represents developing tissue and L₈ undergoes senescence. Considering its dual advantages of reduced variability and enhanced sensitivity to water availability, L₄ can be proposed as the optimal leaf position for assessing potato plant water status via LAD measurement.

Potato plants have a relatively shallow and sparse root system, leading to inefficient water uptake capabilities (Zarzyńska et al., 2017; Zinta et al., 2022). This characteristic makes precision irrigation techniques particularly crucial for potato cultivation. The findings of this study lay the solid foundation for the development of precision irrigation management in potato production, being of substantial agronomic significance. However, implementing LAD measurements to guide irrigation in potato production relies on establishing appropriate LAD-based irrigation indices. Since the LAD-obtained relative maximum yield varies among years due to the variation of weather conditions, it probably needs several years of data to determine the threshold value of LAD of L₄ for a certain potato production region. Moreover, this study revealed that as irrigation levels increased, the extent of the LAD increase varied with the growth stage. Specifically, the increase was more pronounced during the later growth stages ( Table 3 ). This finding strongly suggests that LAD-based irrigation indices should be customized according to different potato growth stages.