How do soil properties and canopy status influence potato traits while continuously monitoring soil moisture with IoT-based sensors?

How do soil properties and canopy status influence potato traits while continuously monitoring soil moisture with IoT-based sensors? | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article How do soil properties and canopy status influence potato traits while continuously monitoring soil moisture with IoT-based sensors? Stanisław Samborski, Tomasz Gnatowski, Jan Szatyłowicz, Renata Leszczyńska This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7613486/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Many soil-related factors associated with potato yield and quality are implicitly linked to soil water availability (SWA). This soil property also affects potato canopy, which could be indirectly evaluated using Unmanned Aerial Vehicles (UAVs). The objective of this study was to quantify the optimal explanatory soil- and canopy state-related properties that determine potato ( Solanum tuberosum L.) traits: total yield (TY), dry matter (DM) content, and average tuber weight (ATW) over a rainfed production field. SWA was continuously measured using moisture sensors installed at three depths and connected via the Internet to exchange data. Soil water storage (SWS) was derived from soil moisture content (SMC). Thus, the effect of spatial and temporal changes in SWS on TY was also determined. Potatoes were grown under a suboptimal water supply during crucial stages for yield formation. More explanatory variables significantly affected ATW and TY than DM content. SWS, determined between the stages of leaf development and canopy closure, played a more critical role in forming ATW and TY than soil chemical properties. A model solely based on canopy status evaluation between the time of canopy closure and the end of flowering, and SWS determined simultaneously, was a strong alternative for forecasting TY and ATW. In the sampling area with the highest TY, the average SWS did not drop below the level that limits potato growth for most of the growing season. A sampling area with the lowest TY was characterized by the lowest water retention for most of the growing season. potato IoT soil moisture sensors soil properties UAV NDVI Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Introduction Europe is responsible for about 26.2% of the world’s potato production (FAOSTAT 2024). Poland is one of the main potato producers in central Europe. Over the last two decades, the average potato yield in Poland has increased by 0.77 tons per year (Dzwonkowski 2024 ). But, in some years with water shortages and uneven rainfall distribution, the average national potato yield decreases significantly. This is due to water stress – a negative balance of water supply compared to evapotranspiration (Dubois et al. 2020), which is expressed as the Climatic Water Balance (CWB), provided by the Agricultural Drought Monitoring System https://susza.iung.pulawy.pl/en/system (2025). Moreover, water stress also negatively affects tuber quality, leading to their malformations and thus reduction of marketable yield (Abbas and Ranjan 2015). Potato is considered a drought-sensitive crop due to possessing a sparse, shallow, and weak soil-penetrating root system. Thus, according to Zarzyńska et al. ( 2017 ), potato cultivars that develop a deeper root system in drought conditions yield better. Additionally, potatoes yield well in sandy and sandy loams with low to medium water-holding capacity. Due to a wide representation of very light (mainly sands according to USDA, 28% of the area of agricultural land) and light (mainly loamy sands and some sandy loams, 30%) soils (Łopatka, 2017 ), these texture-type soils are commonly used for growing potatoes in Poland. Moisture needs of potato are met at the very early stage by the mother tuber because it is assumed that the role of this potato part as a deliverer of energy stops when the leaf area index (LAI) exceeds 0.75 (Haverkort et al. 2015 ). At the latter growth stages, moisture is supplied either from the reserve of conserved water in the soil or through different irrigation scheduling methods (Rai and Dong 2025 ), or both. According to the Polish Potato Association, Poland's total potato irrigated area currently accounts for approximately 25%. The average potato yields on these farms, which mainly grow potatoes for processing, are around 50–60 t·ha − 1 , thus considerably higher than the country average yield of 31 t·ha − 1 in 2024 (Dzwonkowski, 2024 ). However, many potato fields are only rainfed due to limited access to water supplies and unaffordable irrigation systems, which are out of reach for smaller producers. Current developments in the Internet of Things (IoT), among many applications in agriculture, enable continuous monitoring of soil moisture deficit (Dhal et al. 2024 ), a crucial factor in controlling tuber yield and quality of potatoes (Abbas and Ranjan 2015). However, research studies on probes installed at different depths on potato fields to provide real-time soil moisture readings have been limited (Abbas and Ranjan, 2015). Usually, soil moisture sensors are installed in a limited number, only in areas representing different irrigation treatments (Abbas and Ranjan 2015; Liao et al., 2016 ; Dubois et al., 2020), but not to monitor the spatial variability of soil moisture due to variations in soil texture (ST). Consequently, the same soil moisture level for the field parts of different STs results in varying soil water storage (SWS) and thus water availability for the plants. Therefore, knowledge of spatial SWS variability within a potato field and temporal variability of SWS over the growing season could help explain the variation in potato yield and other traits. It is worth emphasizing that soil SWA for plants. According to Utset et al. ( 2000 ), critical values of matric potential for potatoes range from − 32 kPa to -68 kPa, depending on evapotranspiration and depth of the root system. Results of the study conducted by Matteau et al. ( 2022 ), in a greenhouse on potatoes grown on sandy soils, suggest that the soil matric potential for maximizing potato yield is located between − 10 and − 24 kPa. The results obtained by Redulla et al. ( 2002 ) suggest that many soil-related factors, which are also associated with potato yield and quality, are implicitly linked to SWA. Soil water retention and transport depend on soil properties, climatic variabilities, and water management if irrigation is provided (Alva 2008 ). Still, the majority of farmers apply uniform fertilizer rates across their fields, including potato crops. However, due to common and intrinsic spatial soil variability, actual conditions for growing potatoes are site-specific (Po et al. 2010 , Whelan and Mulcahy 2017 ; Zebarth et al., 2019 and 2021 , Sheng et al., 2023 ). In addition to traditional soil and plant sampling methods used to determine soil variability, sensor-based solutions have also been available. Farooque et al. ( 2019 ) used soil electrical conductivity (EC) measurements to predict potato yields on commercial fields in Canada, assuming that soil EC could be an indirect indicator of soil properties (Corwin and Lesch, 2003 ). Among the remote sensing methods, unmanned aerial vehicles (UAVs) enable the evaluation of soil and plant state using vegetation indices (VIs), which are indirect measures of potato canopy growth and yield. Thus, they can be integrated into potato models to enhance the yield prediction accuracy (Mukiibi et al. 2024 ). The objective of this research was to quantify the optimal explanatory soil- and canopy state-related properties that determine potato traits, i.e., total yield (TY), and quality indicators: dry matter (DM) content, and average tuber weight (ATW) over a production field. The originality of the study refers to continuous monitoring of SWA for potato plants using in-field IoT soil moisture sensors installed at three depths in each sampling area of various STs. Such sensors enabled the measurement of fluctuations in soil moisture levels in both space and time. Based on the integration of SMC, SWS was determined in the 0–40 cm layer (where the majority of the potato root system is concentrated), during critical stages for yield formation, namely with the tuber initiation and tuber bulking stages being the most sensitive to water stress (Thornton 2020 ). Thus, SWS was treated as one of the soil-related explanatory variables under rainfed conditions. Therefore, the second objective of this study was to analyze how the changes in SWS affected TY. Materials and Methods Site Description and Crop Management The climate of the study site has been classified as humid continental with cold winters and hot summers (Błaś and Ojrzyńska 2024 ). The 30-year average annual precipitation is 550 mm, of which 313 mm is received during the potato growing season, from May to the end of September. The average annual air temperature is 94°C, while the average air temperature over the latter months is 174°C ( https://klimat.imgw.pl/pl/climate-normals ). The study was conducted in 2022 on a part of a commercial, rainfed potato production field of 8.88 ha located in central Poland (52°4' 50.94" N, 21°10' 6.73" E), which belongs to the Research Farm of Warsaw University (Wilanów-Obory) of Life Sciences., The field is 87 m above sea level and is characterized by flat topography (range of elevation from 86.91 to 87.96 m). The soils corresponded to Fluvisols, with an average organic matter content of 1.77%, and a pH in KCl solution of 6.02. Potato is grown on this field in a 4-year rotation, and the previous crop was winter rape ( Brassica napus L.). Fertilizer (N, P, and K) application was at preplant: 150 kg ha − 1 (0–0–60) on March 14th, 107 kg ha − 1 (18–46–0) on April 22nd. Three weeks after planting, fertilizer (N, S) was applied at the rate of 120 kg ha − 1 (26–32.5), and ammonium nitrate (33.5%) at the rate of 52 kg ha − 1 . All fertilizers were applied at uniform rates. The field was planted on 9 May 2022 with a semi-early Hermes potato ( Solanum tuberosum L.) cultivar, tolerant to a combination of heat and drought stress (Siano et al. 2024 ). Hermes is the third most popular potato variety grown in Poland. Beds were formed with 75 cm between rows, and the planting distance was 30 cm. The orientation of crop rows was west-east. Management practices were conducted uniformly across the field throughout the growing period. Delineation of soil and tuber sampling areas Twenty soil, and potato sampling areas (Fig. 1 ), were placed in locations of different growing conditions, thus of various potato yield potential, using a yield map of winter rape from 2021, and a historical soil map at a scale of 1:5000. A similar approach to the location of the soil moisture sensors, but within a soybean field, was used in the study by Pan et al. ( 2010 ). The precise location of the sampling areas was adjusted to the centroids of Sentinel-2 imagery pixels, which were used in another project. A GNSS-RTK receiver (Topcon GRS-3, Topcon Corporation, Tokyo, Japan), with a 3 cm accuracy specification, was used for navigation to the centroids. Weather conditions Rainfall was registered by a rain gauge located c . 1.5 km from the research field, average air temperature, solar radiation, relative air humidity, and wind speed were obtained from a weather station placed at the Warsaw University of Life Sciences campus, c .13 km from the research field. The average air temperature from May to the end of August 2022 was 19.2°C, (Fig. 2 ). The total precipitation during the respective period was 257 mm. The rainfall distribution was 84 mm during May and June and 173 mm in July and August. Measured meteorological data were used to compute reference evapotranspiration (ET), using Penman-Monteith formulae (Allen et al., 1998 ) $$\:ET=\frac{0.408\:{\Delta\:}{\:R}_{n}+{\gamma\:}\:\frac{900}{t+273\:}v\:D}{{\Delta\:}+\:\left(1+0.34v\right)}$$ 1 where: ET – reference evapotranspiration [mm d − 1 ], R n – net radiation at the crop surface [MJ m − 2 d − 1 ], ∆ – slope vapour pressure curve [kPa °C − 1 ], γ – psychrometric constant [kPa °C − 1 ], t – mean daily air temperature [°C], v – wind speed [m s − 1 ], D – saturation vapour pressure deficit [kPa]. The climatic water balance (CWB), in other words, a precipitation deficit, was calculated for the potato growing season and decades, as the difference between precipitation (mm) and reference evaporation (mm) (Wierzbicka 2014 ). Potato was grown mainly on medium (ST of sandy loam), and light (ST of loamy sand), soils (Fig. 1 ). According to Chmura et al. ( 2013 ), to obtain maximum yields of semi-early cultivars grown in Poland on medium soils, potato requires 88 mm of rainfall in May and June and 220 mm in the second part (July-August) of the season. The optimal rainfall distribution for growing semi-early potato cultivars on light soils is 143 mm during May and June and 220 mm in July and August. Thus, the experiment was conducted in a year of suboptimal water supply for potatoes cultivated on light soils during the whole season and optimal conditions in only the first two months of potato growth on medium soils. UAV Flight Data Collection The aerial-based data was collected by UAV (Phantom 4 Multispectral, DJI, Shenzhen Dajiang Baiwang Technology Co., Ltd., China), equipped with a highly accurate Global Navigation Satellite System with RTK (Real Time Kinematic) correction. Flights were conducted five times with 2-week intervals at BBCH growth stages (described by Meier 2018 , and given in the brackets), starting from mid-season as recommended by Whelan & Mulcahy ( 2017 ), on June 22nd (BBCH 39 – canopy closure, tuber initiation, the largest tubers of 25 mm); July 6th (BBCH 61 – beginning of flowering, the largest tubers of 40 mm), July 20th (BBCH 69 – end of flowering, tuber size in the range of 35–55 mm); 3 August (BBCH 70–79 – development of fruit), and 17th (BBCH 81–-89 – ripening of fruit and seed). Each UAV flight was conducted perpendicularly to the direction of sunlight, close to solar noon, on cloud-free days, to avoid changing light conditions. The UAV altitude was c. 80 m above ground level, providing a 4 cm pixel − 1 resolution. Images were taken by DJI’s FC6360 camera with 75% front and side overlap. Ortho-photo images for each spectral band (blue, green, red, red-edge, and NIR) and NDVI (Normalized Difference Vegetation Index) were generated using the Pix4DFields computer software (Pix4D S.A., Prilly, Switzerland). NDVI, which is commonly used for monitoring crop canopy but does not require sophisticated calculations, was applied based on red (650 nm ± 16 nm) and near-infrared (840 nm ± 26 nm) bands (Rouse et al., 1973 ). Further processing of the ortho-photo NDVI images was conducted using QGIS software (QGIS.org 2022). A polygon-shaped file with a rectangular region of interest (ROI) of the same size, 0.75 m (width) by 4.0 m (length), was created for each sampling area to extract average NDVI values using zonal statistics. Soil sampling and analyses At harvest on September 12 and 13th, 2022, soil samples for physicochemical analyses were taken before digging the potatoes, and soil resistance (SR) was measured in three locations (Fig. 3 a). The latter was done with the use of a Penetrologger (Eijkelkamp, Giesbeek, the Netherlands) to a depth of 80 cm, and then averaged for the depths of 0–10, 11–20, 21–30, 31–40, 41–50, 51–60, 61–70, and 71–80 cm, and additionally for the plow layer of 0–30 cm (Fig. 3 b). After that, undisturbed soil samples for bulk density (BD) determination were taken at the three locations to a depth of 15 and 25 cm using 100 cm³ cylinders, where most potato tubers develop (Fig. 3 c). Finally, five soil subsamples were taken from each of the three locations (15 subsamples in total), per sampled row for soil chemical and texture analyses (Fig. 3 d). The content of available forms of the following nutrients: P, K, Mg, Fe, Na, Cu, Mn, Zn, and Co in soil samples, sieved through a 2-mm mesh, was determined using the Mehlich-3 method (Sikora & Moore, 2014 ). The Avio 200 ICP optical emission spectrometer (PerkinElmer, Inc., Waltham, MA, USA) was used to determine the contents of the above nutrients. Soil pH was measured in a 1 mol KCl solution using pH-METER CG 842 (Hofheim, SCHOTT, Geräte GmbH, Germany). The total contents of C, N, and S were determined using the Elementar Vario MacroCube analyzer (Elementar-Straße 1, 63505 Langenselbold, Germany). Soil samples sieved through a 2-mm mesh were subjected to particle size distribution analysis using the hydrometric (areometric) method of Casagrande modified by Prószyński (Musierowicz 1950 ). The content of soil separates was determined according to the USDA classification. Soil EC measurements were done on 28 October 2022 with a MSP-3 platform (Veris Technologies, Salina, Kansas, USA), at depths of 30 cm (EC sh ) and 90 cm (EC dp ). For statistical analysis, averaged EC data from a buffer of 20 m in diameter of each sampling area were used. Soil moisture monitoring On 3 June 2022, soil moisture sensors of Yosensi Ltd. company (Białystok, Poland), were horizontally mounted at three depths of 10, 20, and 30 cm (Fig. 4 a), in the middle of each sampling area (Fig. 4 b). Potato plants at that time were at the leaf development stage (BBCH 10–19) – Fig. 2 . The SMC measurements were conducted using capacitive probes operating at a low frequency of 75 MHz. The primary measurable parameter is the variable dielectric permittivity, expressed as voltage in the 0 to 3.3 VDC range, where higher voltage corresponds to higher SMC and vice-versa . The sensor’s measuring element (capacitor linings) is made of laminate. The electronics are vacuum-sealed in epoxy resin, ensuring the sensor is waterproof and resistant to various weather conditions. The probe is 2 mm thick, 30 mm wide, and 150 mm long. The allowable operating temperature of the probe is between − 40℃ and + 60℃. The soil moisture probes are connected to a node (Fig. 4 b) with an in-built wireless communication module, which gathers data, forms the payload, and sends it to a gateway. This device, similar to a router, is equipped with a LoRa (Long Range) concentrator that receives LoRa packets and sends them to the Internet-connected server. Using the calibration equation (developed for the Research Farm of Warsaw University of Life Sciences, located in Obory, unpublished data), the output of the probes, expressed as voltage, was converted to volumetric moisture content measured hourly and then averaged to obtain daily data. Based on the integration of the SMC, an average soil water storage (SWSavg) in the observation period, i.e., between 3 June (25 DAP – days after planting) to 3 August (102 DAP), was determined in the layer of 0–40 cm. The calculated values of SWS were compared with those at characteristic moisture conditions, i.e., water content at field capacity, limiting, and wilting points. The pedotransfer functions proposed by Wösten et al. ( 1999 ) were used to assess the soil water retention curve. Using measured data on the content of soil separates, organic carbon content, and soil bulk density, the parameters of the van Genuchten ( 1980 ) equation that describes soil water retention curve were calculated. Computation of soil moisture at the limiting and wilting points was performed assuming soil matric potential equal to -500 cm for the limiting point and − 15848 cm for the wilting point, respectively. A physically based analytical equation derived by Assouline and Or ( 2014 ) was used to predict SMC at field capacity. For the need of calculating the relationship between descriptive variables and potato traits, SWS was calculated as an average for the following periods: SWSaP1 from BBCH 10–19 to BBCH 39, SWSaP2 from BBCH 39 to BBCH 61, SWSaP3 from BBCH 61 to BBCH 69, SWSaP4 from BBCH 69 to BBCH 70–79, SWSaP5 from BBCH 69 to BBCH 81–89, Additionally, for the same periods, the minimum (SWS min ), maximum (SWS max ), range (SWS range ), and standard deviation (SWSstd) of SWS were determined. Potato sampling and potato trait determination On 31 August, the potato crop was defoliated. Potatoes were manually harvested at each sampling area of 3 m 2 , by collecting tubers along a 4 m distance from a row of 0.75 m width (Fig. 3 a). A total yield (TY) – t·ha − 1 , and the total number of tubers per sample were determined. The ratio of these two traits was used to calculate the average tuber weight (g) – ATW. Dry matter content in tubers, an equivalent of specific gravity, was determined using a hydrometric method (a weight-in-air/weight-in-water method). A potato sample of 3.64 kg was weighed in air and then reweighed in water. DM content was calculated based on the differences of these two weights (Kleinschmidt et al. 1983 ). Statistical analysis The statistical analysis included the calculation of correlations between explanatory and response variables obtained for the twenty sampling areas (Fig. 1 ), with the determination of statistical significance represented by the p-value below 0.05. The set of explanatory variables was described by soil chemical and physical properties, including measured SWS (i.e., SWSaP1, SWSaP2, SWSaP3, SWSaP4, and SWSaP5), in a 40 cm soil layer. Additionally, NDVI values derived from UAV imagery, an indirect indicator of plant biomass and chlorophyll content during plant development, shaped by the soil properties from plant emergence to plant senescence, were also used as the explanatory (descriptive) variables. The response variables were represented by potato traits determined at harvest: TY, ATW, and DM content. For the multidimensional analysis, redundancy analysis (RDA) was employed, enabling the examination of the effects of explanatory variables on the set of response variables. The applied RDA is an extended Principal Component Analysis (PCA) method, which enables the ordination of Y to obtain ordination axes that are linear combinations of the variables in X (Legendre & Legendre, 2012 ). The statistical method of RDA also involved reducing multicollinearity between explanatory variables in the X matrix and the explained potato traits (Y matrix). Multicollinearity reduction refers to removing highly correlated variables from a model to enhance its accuracy. On the other hand, ordination axes are the new variables created through the PCA method that best explain the variation in the original variables. The RDA was performed using the R-package vegan v2.6-10 (Oksanen et al., 2025 ) in the RStudio interface (RStudio Team, 2020). The graphical representation of the data set was prepared using ggplot2 (Wickham, 2016 ). Before the RDA was performed, the whole data set was standardized with the method of standard deviation from the mean as follows: $$\:{Z}_{ij}=\frac{{O}_{ij}-{O}_{avg}}{{\sigma\:}_{i}}$$ 2 where Z ij is the standardized value of the i-th response variable, including also tuber quality for the j-th sampling area; Oij is the original value of the i-th response variable for the j-th sampling area; O avg is the average value of the i-th analysed variable in all sampling areas; and σ is the standard deviation of the i-th variable. The R 2 adj was used to assess the performance of the RDA analyses. The modeling part of the study plays a crucial role in establishing efficient forecasting models for the potato traits (response variables): TY, ATW, and DM content at a field scale. The study considered four models (strategies): M1, M2, M3, and M4, each with a unique approach to understanding and predicting the potato traits. Strategy M1 assumed that selecting the descriptive variables based on the highest correlation ( r value above 0.50) enables a proper prediction of the potato traits using the RDA. This strategy evolved because the number (48) of descriptive variables was higher than the number of sampling areas (n = 20). However, such an extensive range of explanatory variables is not standard practice. Very often, the number of the explained variables is limited only to, e.g., soil physical properties. Therefore, we examined three other models (strategies) for the potato traits. The M2 strategy assumed that the input variables represent only physical soil properties during the model development. The next model (M3) investigated the effect of chemical soil properties on potato traits alone. Finally, the last considered model (M4) assumed that the potato traits depend only on the indirect indicator of plant development (NDVI) and SWS properties. Additionally, the Leave-one-out (LOO) cross-validation (CV) scheme (James et al., 2013 ) was applied to assess the RMSEcv error of the potato traits. The quality of the final model was evaluated using a ranking procedure. This procedure is based on the statistical measures obtained from developing and testing (LOO scheme) of the analyzed models. During the development of the models, the adjusted coefficients of determination were determined (adjusted R 2 ) for the standardized values used in the RDA. The root mean squared error (RMSE) was also calculated, based on untransformed data from the prediction and the original data of the potato traits. Results and discussion Variability of potato traits The average potato yield of all the sampling areas was 45.5 t∙ha -1 , and the yield range was 16.2. to 67.1 t∙ha -1 (Table 1). Table 1. Descriptive statistics for variation in total potato yield (t∙ha -1 ), average tuber weight (g), and dry matter content (%), n = 20, for the commercial potato field in 2022. Variable Average Standard Deviation Coefficient of variation (%) Range Total yield (t·ha -1 ) 49.5 13.0 26.2 16.2-67.1 Average tuber weight (g) 91.9 23.3 25.4 45.1-132 Dry matter content (%) 22.6 0.92 4.06 20.8-24.2 Among the determined potato traits, total yield and ATW were characterized by similar coefficient of variation (CV), 26.2 and 25.4 %, respectively. Zebarth et al. (2019) found a CV of yield of 18% within a 21-ha rain-fed field in Canada. Much greater yield variability within a field, of 3.5-fold, for sixteen fields, was observed by Whelan and Mulcahy (2017) in Tasmania. Quality potato trait – DM content was much less variable among the twentieth sampling areas and had a CV value of 4.06 %. Kleinschmidt et al. (1983) summarized that sandy and heavy clay soils produce potatoes with lower specific gravity (DM content) than medium textured soils. In the case of the sampling locations with sandy (no 2) and loam (no 7), soil texture (Figure 1), the DM content was respectively 22.0% and 22.6%, thus very close to the field average. Beukema and van der Zaag (1990) observed that fields that are homogeneous in soil and fertility produce crops with less variation in DM content than fields with higher spatial variability. Variability of soil chemical and physical properties All the results of the soil analyses are presented as supplementary material in Tables 1s, 2s, and 3s. Among all soil chemical properties (Table 1S), acid soil and low fertility in Mg, Mn, and Zn, which could have limited potato yield, were observed only in soil sampling areas 2 and 9 (Kęsik 2016; Korzeniowska et al. 2021). However, very different yields were achieved in these two areas, respectively 16.2 and 57.7 t∙ha -1 (unpresented data). This may indicate that other factors also limited yield in the sampling area number 2. Sheng et al. (2023) observed that irrigation overshadowed the effect of nutrients on potato yields obtained on production fields in northern China. This field, according to the ST map (Figure 1), is mainly composed of sandy loam (western part) and loamy sand ( eastern part) in a plough layer. Sand and loam texture occur only in single sampling areas, 2 and 7, respectively. Soil resistance (an indicator of compaction status) below the plow layer was characterized by average values of above 2.1 MPa, which could be considered a factor that limits deeper soil penetration by potato roots and causes poor drainage and poor deep water storage (Table 2s), Moebius-Clune et al. (2017). Both soil layers (0-30 and 0-90 cm) were characterized by similar average and range electrical conductivity values. Only in a few sampling areas, EC for the deeper layer was slightly higher than the EC measured in the shallow layer (data not presented). This suggests no significant changes in ST within a soil profile. The average value of soil bulk density at a depth of 15 cm was 1.40 g cm -3 , and it ranged from 1.30 (sampling area no 2) to 1.50 g cm -3 (sampling area no 12) (unpresented data). In the deeper layer (25 cm), the soil was characterized by a higher bulk density, with an average value of 1.49 g cm -3 , ranging from 1.36 g cm -3 (sampling area no 13) to 1.63 g cm -3 (sampling area no 1) (Table 3s). The increase in soil density with depth of potato ridges was confirmed by Antoneli et al. (2025), and this soil property largely depends on ST and tillage practices (Abrougui et al., 2014). Relationship between descriptive variables and potato traits (response variables) Correlations of potato traits with all investigated descriptive variables are presented in Figure 6. In general, a much higher number of the explanatory variables significantly affected ATW and TY than DM content. According to Beukema and van der Zaag (1990), almost all factors influencing DM content also affected tuber yield. Moreover, the effect of various factors upon DM content is very complex. Namely, under certain conditions, the same factor may have a positive effect, and under other conditions (not always noticeable), it may have an adverse impact on this potato trait. Only seven descriptive variables significantly correlated with all investigated potato traits, with r values above 0.50 (Fig. 6). These variables represent soil physical properties (sand and silt content), and chemical properties like soil fertility in Zn and Cu, NDVI derived from UAV images taken on the 3 rd and 17 th of August 2022. SWSaP1, determined for the period from BBCH 10–19 to BBCH 39 (the time from leaf development to canopy closure – tuber initiation), also significantly correlated with all potato traits. Regarding the first model (M1), NDVI registered on 17 August 2022 (during ripening of fruit and seed), due to the high correlation with the other descriptive variables, was redundant (Fig. 8). Five of the analyzed explanatory variables were responsible for DM, AWT, and TY formation. Sand content negatively correlated with all potato traits and thus limited their development. DM strongly and positively depended on the silt content. The first two components of the RDA explained almost 87% of the variation of the dataset. Total yield and ATW were strongly positively related to sand and silt content, soil fertility in zinc and copper, NDVI measured on 3 August 2022, and SWSaP1 from BBCH 10–19 to BBCH 39. However, none of the sampling areas showed a deficiency in soil copper, and only three sampling areas with high soil P content showed a deficiency in soil zinc (unpresented data). The colored points on Figure 8 indicate the field sampling areas with different ATW values. The highest DM had tubers harvested from sampling areas of A6 and A13. Conversely, the lowest DM values were determined for soil and plant sampling areas A5 and A17 (Fig. 7). The results of the RDA of the second model (M2) indicate the negative influence of sand content and bulk density measured at 25 cm on TY and ATW (Fig. 9). In this case, silt content also significantly and positively influenced DM, whereas bulk density at a depth of 15 cm negatively affected this potato trait. Lower soil resistance measured at 0–10, 21–30, 31–40 cm, and 61–70 cm positively influenced TY and ATW. The first two components of the RDA explained almost 85% of the variation of this dataset. The third model (M3) illustrates the impact of chemical properties on potato traits (Fig. 10). A higher content of sodium and copper and higher soil pH values were positively correlated with all dependent variables. On the other hand, the increase of phosphorus content in the soil negatively affected all potato traits. This could be explained by the fact that all sampling areas, taking into account also soil pH, soil was characterized by very high P content. Interestingly, Ravensbergen et al. (2023) did not find any correlation between potato yield response to P application and the amount of phosphorus applied by the farmers in the Netherlands on 46 fields. On the contrary, Leonel et al. (2017) observed that soils with higher phosphorus concentrations allowed the production of tubers with increased DM content. It should be stressed that the performance of model M3 was relatively low, because it explained only 45.5% of the variability in the dataset. Redulla et al. (2002), found that ST components (sand, silt, and clay), impacted yield stronger than soil chemical properties. The strategy representing the fourth model (M4) explained almost 66% of the data set variation (Fig. 11). Higher values of such explanatory variables as NDVI registered by UAV on 22 nd of June and 20 th of July, 2022, SWSaP1, determined for the period from BBCH 10–19 to BBCH 39, positively influenced all potato traits. Lynch et al. (1995) found that early (tuber initiation – 2 and 4 weeks after emergence) and midseason (early tuber sizing – 6 and 8 weeks after emergence) moisture stress had the most significant negative impact on tuber yield in Alberta, Canada.e The tuber initiation stage was the most appropriate for remote sensing data acquisition (Mukiibi et al. 2024). Research results for 46 potato cultivars from different maturity classes grown in Poland show that the crop was most sensitive to drought from 3 to 6 weeks after tuber initiation (Głuska 2004). Summary of the four models’ performance The cross-validation scheme is typically used to evaluate the final model performance, and reflected by the RMSE CVLOO error value (Table 2). The results of the statistical analysis indicate that the value of the adjusted coefficient of determination (R 2 adj.) was highest for the first (M1) model and lowest for the third (M3) model. Table 2. Statistical measures of the four models and the cross-validation scheme. Model Descriptive variables included in the model Adjusted R 2 RMSE RMSE CVLOO TY ATW DM TY ATW DM M1 silt, sand, Cu, Zn, NDVI0308, SWSaP1 0.832 5.85 7.98 0.075 8.59 13.0 0.13 M2 BD15, BD25, sand, silt, SR0010, SR2130, SR3140, SR6170 0.781 5.95 9.04 0.018 10.9 16.0 0.04 M3 pH, Al, Cu, Na, P 0.268 10.2 13.5 0.701 14.0 19.3 1.03 M4 NDVI2206, NDVI2007, SWSaP1 0.595 6.67 9.23 0.680 7.52 10.9 0.81 TY – tuber yield; ATW – average tuber weight, DM – dry matter, RMSE – root mean squared error in the dimension of the original variable, RMSE CV – root mean squared error based on the leave-one-out cross-validation scheme. The results show that forecasting the development of the analyzed potato traits using the M1 model required a diverse input data set, including soil physical and chemical properties, and SWS during the canopy closure period. The effectiveness of the forecast was also influenced by the indirect crop canopy status, expressed by NDVI measured at the berry formation stage (BBCH 70–79). The statistical measures, specifically the RMSE, used for estimating potato traits showed the lowest values for M1, in the case of TY and ATW. However, this error evaluated for DM was slightly lower for M2. A similar level of RMSE CVLOO was obtained in the cross-validation scheme for models M1 and M2. Interestingly, model M4, which was solely based on the indirect measure of canopy status (NDVI) and SWSaP1, was a strong alternative for forecasting TY and ATW, because the RMSE CVLOO values for the latter potato traits were the lowest for this model. Therefore, if the ranking of the models was solely based on the cross-validation results, the performance of model 4 would be ranked the highest. Table 3. The ranking of the model performance is based on statistical measures from Table 2. Models Ranking values Rank Model order M1 1 1 1 2 2 2 2 1.7 1 M2 2 2 2 1 3 3 1 2.0 2 M3 4 4 4 4 4 4 4 4.0 4 M4 3 3 3 3 1 1 3 2.4 3 The final performance (Table 3) of the four models was evaluated using all statistical measures presented in Table 2. Considering all the measures, the model’s performance was ranked: 1, 2, 4, and 3. Effect of soil water storage on potato yield According to Liao et al. (2016), soil temperature at a depth of 20 cm and SMC at 20-30 cm depth were considered significant factors that affected potato yields on commercial fields in Florida, USA. Recommended soil moisture for optimal potato growth depends on the development stage. It should be maintained at 70–80% of field capacity (FC) from planting to early vine growth and 80–90% FC at later growth stages, until maturation, when soil moisture should be held at 60–65% FC (Rai and Dong 2025, after Pavlista 1995 and van Loon 1981). In our research, the average measured value of SWS in the 40 cm soil layer for all sampling areas was equal to 59.7 mm, which lies in the range between the SWS at field capacity and limiting point (Table 3s). This means that in the majority of the sampling areas, the average SMC was in the optimal range. For the comparison of SWS, four (number 2, 4, 8, and 11) soil sampling areas were chosen based on various yields. Fig. 5 shows the results of measured SWS in the soil layer of 40 cm in the four selected sampling areas against the background of calculated values of reference evapotranspiration using equation (1) and measured precipitation values. This figure also shows calculated SWS at field capacity, limiting, and wilting points. During the potato growing season (from May 9 th to September 13 th ), the calculated ET value was 427.7 mm, precipitation was 280 mm, and the CWB value was 147.2 mm, which confirms that, in general, potato was grown under suboptimal water supply. For the same period as given above, the positive precipitation deficit was observed only in the last decade of May and the first decade of July, but after a whole month of June with a negative value of CWB. Consequently, the most significant water deficit occurred during tuber initiation and tuber bulking stages, the most critical for potato yield formation (Thornton 2020). In a sampling area number 2 with the lowest yield of 16.2 t·ha -1 , the soil was characterized by the lowest water retention capacity, as shown by the measured SWS values, which fluctuated most of the growing season around 33 mm. This value was close to the SWS corresponding to the limiting point of 32 mm. Potatoes harvested from this sampling area were also characterized by the lowest AWT and number of tubers per plant (unpresented data). In a soil sampling area number 11 with a yield of 38.8 t·ha -1 , the average SWS was higher (50.7 mm) compared to the SWS measurements done for the previous sampling area. SWS corresponding to a limiting point for this area was 45.4 mm. For soil sampling area number 4, yielding 40.7 t·ha -1 , the average SWS was 67.2 mm, and thus was lower than the SWS at the limiting point of 72 mm. This means that soil moisture was slightly below its optimal level. The highest yield of 67.1 t·ha -1 was obtained at soil sampling area no 8. An average SWS in this area was 97.2 mm and did not drop below the SWS corresponding to the limiting point (77.4 mm) for most of the growing season. This sufficient water supply also produced tubers with the highest average mass. In all of the described soil sampling areas, clear changes in SWS were observed after significant rainfall, which means that the soil moisture sensors worked correctly. It should be noted that the most substantial changes in SWS (water depletion) were observed in the period of SWSaP1 (from BBCH 10–19 to BBCH 39), namely from 25 to 44 DAP. Conclusions The analysis indicates that a greater proportion of the descriptive variables significantly impacted ATW and TY than DM content. This suggests that factors influencing ATW and TY are more pronounced or varied than those affecting DM content, which may have implications for understanding the dynamics of these variables in the study. Among the four tested models describing the effect of descriptive variables on potato traits, model 1, based on sand and silt content, soil fertility in Cu and Zn, and NDVI measured by the use of UAV at the time of fruit development, and soil water storage (SWSaP1), registered for the time between leaf development and canopy closure (BBCH 10–19 and BBCH 39), was ranked the highest using adjusted R 2 , RMSE and RMSE CVLOO . If the ranking of the models was solely based on the cross-validation results, the performance of model 4, incorporating NDVI measured twice, at the time of canopy closure and the end of flowering, and SWSaP1, would be ranked the highest. The sampling area with the lowest yield was characterized by the lowest water retention for most of the growing season, and also by an early deficit of SWS, which started close to BBCH 10–19. On the contrary, in the sampling area with the highest yield, an average SWS did not drop below the SWS that corresponds to the limiting point for most of the growing season. Soil water storage determined for the time from leaf development to canopy closure, included in models 1 and 4, seems to play a more critical role in shaping tuber size and potato yield than soil chemical properties if these are kept at an optimal level. Declarations Conflict of Interest The authors declare that there is no conflict of interest. Funding No funding was received. Acknowledgments The authors thank the management of the Research Farm of the Warsaw University of Life Sciences for providing the field to conduct this study and for their technical support. We also thank Przemysław Rudzki from the Yosensi Ltd. company for providing us with the IoT soil moisture sensors, and Bartłomiej Rochalski from the Research Farm of Warsaw University of Life Sciences in Żelazna for conducting part of the UAV flights. References Abbas H., Ranjan R. Sri. 2015. Effect of soil moisture deficit on marketable yield and quality of potatoes. Canadian Biosystems Engineering, 57: 1.25-1.37 http://dx.doi.org/10.7451/CBE.2015.57.1.25 Abrougui, K., Chehaibi, S., Boukhalfa, H. H., Chenini, I., Douh, B., Nemri, M. 2014. Soil bulk density and potato tuber yield as influenced by tillage systems and working depths. Greener Journal of Agricultural Sciences, 4(2), 46-51. Agricultural Drought Monitoring System (2025) https://susza.iung.pulawy.pl/en Allen R. G., Pereira L. S., Raes D., & Smith M. 1998. Crop evapotranspiration-Guidelines for computing crop water requirements – FAO Irrigation and Drainage Paper 56. FAO, Rome, 300(9), D05109. Alva 2008. Setpoints for Potato Irrigation in Sandy Soils Using Real-Time, Continuous Monitoring of Soil-Water Content in Soil Profile. Journal of Crop Improvement, 21(2), 117–137. https://doi.org/10.1080/15427520701885311 Antoneli,V., Rogiski F. L., Bednarz J. A., Barrena-González J., Bernardo, F. S., & Fernández, M. P. 2025. Assessing Soil Degradation and Hydrological Processes in Conventional Potato Crops. Air, Soil and Water Research, 18, 11786221251344463. Beukema, H.P., van der Zaag, D.E. 1990. Introduction to potato production. Pudoc Wageningen. 208 pp. ISBN 90 220 0963 7. Błaś, M., Ojrzyńska, H. 2024. The Climate of Poland. In: Migoń, P., Jancewicz, K. (eds) Landscapes and Landforms of Poland. World Geomorphological Landscapes. Springer, Cham. https://doi.org/10.1007/978-3-031-45762-3_3 Assouline, S., & Or, D. 2014. The concept of field capacity revisited: Defining intrinsic static and dynamic criteria for soil internal drainage dynamics. Water Resources Research, 50(6), 4787-4802. Chmura, K., H. Dzieżyc, Piotrowski M. 2013. Response of medium early, medium late, and late potatoes to water factor on wheat and rye soil complexes. Infrastructure and Ecology of Rural Areas , 2: 103–113 (in Polish). Corwin, D.L., Lesch, S.M., 2003. Application of soil electrical conductivity to precision agriculture: theory, principles, and guidelines. Agronomy Journal, 95, 455–471. https://doi.org/10.2134/agronj2003.4550 Dhal, S., Wyatt, B. M., Mahanta, S., Bhattarai, N., Sharma, S., Rout, T., Saud, P., & Acharya, B. S. 2024. Internet of Things (IoT) in digital agriculture: An overview. Agronomy Journal, 116, 1144–1163. https://doi.org/10.1002/agj2.21385 Duboisa A., Teytauda F., Verel S. 2020. Short term soil moisture forecasts for potato crop farming: a machine learning approach. Computer and Electronics in Agriculture, 180, 105902. https://doi.org/10.1016/j.compag.2020.105902 Dzwonkowski W. 2024. Rynek ziemniaka. Stan i perspektywy. 51, Analizy Rynkowe. IERiGŻ PIB. pp. 44. (In Polish). Farooque A. A., Zare, M., Abbas, F., Bos, M., Esau, T. & Q. Zaman. 2019. Forecasting potato tuber yield using a soil electromagnetic induction method. European Journal of Soil Science, 1-18. Głuska A. 2004. Reaction of potato ( Solanum tuberosum L.) to rainfall distribution under field conditions and testing drought imposed at different phases of bulking stage in cultivars of different vegetation length in a pot experiment. Zeszyty Problemowe Postępów Nauk Rolniczych, 496, 217-227. (in Polish) Haverkort A.J., Franke A.C., Steyn J.M., Pronk A.A, Caldiz D.O., Kooman P.L. 2015. A Robust Potato Model: LINTUL-POTATO-DSS. Potato Research, 58:313-327. https://doi.org/10.1007/s11540-015-9303-7 Institute of Meteorology and Water Management. National Research Institute. Climate normal 1991–2020. https://klimat.imgw.pl/pl/climate-normals James G., Witten D., Hastie T., Tibshirani R. 2013. An introduction to statistical learning (Vol. 112, No. 1). New York: Springer (177-199). Kęsik K. 2016. Zastosowanie metody Mehlich 3 w systemie doradztwa nawozowego. Studia i raporty IUNG-PIB, ZESZYT 48(2): 95-104, doi: 10.26114/sir.iung.2016.48.07 (In Polish). Kleinschmidt G.D., Kleinkopf G.E., Walz A.J., Zalewski. J.C., 1983. Specific gravity of potatoes. Current Information Series No. 609/ University of Idaho. College of Agriculture. Korzeniowska J., Stanisławska-Glubiak E., Jadczyszyn T., Lipiński W. 2021. Nawożenie upraw rolniczych mikroelementami. Instrukcja upowszechnieniowa nr 249. Wydawnictwo IUNG-PIB, pp. 31. (In Polish). Legendre P, Legendre L. Redundancy analysis. 2012. In: Numerical Ecology. Elsevier, editor. 3 rd ed. Elsevier Science B.V; 629–658 Legendre P., & Legendre L. 2012. Canonical analysis. In Developments in environmental modelling (Vol. 24, pp. 625-710). Elsevier. Leonel M, do Carmo EL, Fernandes AM, Soratto RP, Ebúrneo JAM, Garcia ÉL, dos Santos TPR. 2017. Chemical composition of potato tubers: the effect of cultivars and growth conditions. Journal of Food Science and Technology, 54, 2372–2378. https://doi.org/10.1007/s13197-017-2677-6 Liao X., Su Z., Liu G., Zotarelli L., Cui Y., Snodgrass C. 2016. Impact of soil moisture and temperature on potato production using seepage and center pivot irrigation. Agricultural Water Management, 165, 230-236. http://dx.doi.org/10.1016/j.agwat.2015.10.023 Lynch D. R., Foroud N., Kozub G. C., Farries B. C. 1995. The effect of moisture stress at three growth stages on the yield, components of yield, and processing quality of eight potato varieties. American Potato Journal, 72, 375-385. https://link.springer.com/article/10.1007/BF02849334 Łopatka, A., 2017. Europejski monitoring użytkowania gruntów i baza danych projektu LUCAS. Studia i Raporty IUNG-PIB 51 (5), 73–89. (in Polish). Matteau, J. P., Célicourt, P., Létourneau, G., Gumiere, T., & Gumiere, S. J. 2022. Effects of irrigation thresholds and temporal distribution on potato yield and water productivity in sandy soil. Agricultural Water Management, 264, 107483. Meier, U. 2018. Growth stages of mono- and dicotyledonous plants. BBCH Monograph. http://doi/10.5073/20180906-07461 Moebius-Clune B.N., Moebius-Clune D.J., Gugino B.K., Idowu O.J., Schindelbeck R.R., Ristow A.J., van Es H.M., Thies J.E., Shayler H.A., McBride M.B., Kurtz K.S.M., Wolfe D.W., and Abawi G.S. 2017. Comprehensive Assessment of Soil Health – The Cornell Framework. Edition 3.2, Cornell University, Geneva, NY. Mukiibi A., Machakaire A.T.B., Franke A.C., Steyn J.M. 2024. A Systematic Review of Vegetation Indices for Potato Growth Monitoring and Tuber Yield Prediction from Remote Sensing. Potato Research https://doi.org/10.1007/s11540-024-09748-7 Musierowicz, A. 1950. Granulometric composition of soils and methods of analysis of granulometric composition. (in Polish). Warszawa, 1949. PIWR. Oksanen, J.,Simposon G.L, Blanchet, F.G., Kindt, R., Legendre, P., Minchin R., O’Hara R.B., et al., 2025. Vegan Community Ecology Package Version 2.6-10 January 2025. Pan, L.; Adamchuk, V.I.; Martin D.L.; Schroeder M.A. & Ferguson, R.B. 2010. Analysis of water use efficiency using on-the-go soil sensing and a wireless network. In: Proceedings of the Tenth International Conference on Precision Agriculture, Denver, Colorado, 18-21 July 2010, ed. R. Kholsa. Colorado State University, Fort Collins, Colorado, USA (CD publication, 13 pages). Po E., Snapp S.S, & A. Kravchenko. 2010. Potato Yield Variability across the Landscape. Agronomy Journal, 102(3), 885-894. doi:10.2134/agronj2009.0424. Rai A., Dong Y. 2025. A scoping review of irrigation scheduling methods in potato ( Solanum tuberosum L.) production. American Journal of Potato Research https://doi.org/10.1007/s12230-025-09997-3 Ravensbergen A. P. P., van Ittersum M. K., Kempenaar C., Reidsma P. 2023. Current Phosphorus and Potassium Fertiliser Application Rates Do Not Limit Tuber Yield and Quality in Potato Production Systems in the Netherlands. Potato Research, 66:1033–1058 https://doi.org/10.1007/s11540-022-09613-5 Redulla, C.A., Davenport, J. R., Evans, R. G., Hattendorf, M. J., Alva, A.K., Boydston, R. A. 2002. Relating Potato Yield and Quality to Field Scale Variability in Soil Characteristics. American Journal of Potato, 79:317–323. Rouse, J. W., Haas, R. H., Schell, J. A., Deering, D. W. 1973. Monitoring vegetation systems in the Great Plains with ERTS. In Third ERTS Symposium, DC. 10-14. NASASP-351, MD, pp 309–317. Sheng, Y., He, P., Xu X., Liu Y. 2023. A large-scale assessment on spatial variability of potato yield and soil chemical properties in northern China. Soil and Tillage Research, 231. https://doi.org/10.1016/j.still.2023.105743 RStudio Team (2020). RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL http://www.rstudio.com/. Siano, A. B., Roskruge, N., Kerckhoffs, H., Sofkova-Bobcheva, S. 2024. Effects of Abiotic Stress Associated with Climate Change on Potato Yield and Tuber Quality Under a Multi‑environment Trial in New Zealand. Potato Research, 67: 1603-1624. Sikora, F. J., Moore, K. P. 2014. Soil test methods from the southeastern United States. Southern Cooperative Series Bulletin 419. Clemson, SC, USA: Clemson University. Thornton, M., 2020. Potato Growth and Development, in: Potato Production Systems. Springer International Publishing, Cham, pp. 19–33. https://doi.org/10.1007/978-3-030-39157-7_2 Utset, A., Ruiz, M. E., Garcia, J., & Feddes, R. A. 2000. A SWACROP-based potato root water-uptake function as determined under tropical conditions. Potato Research, 43, 19-29. Van Genuchten, M. T. 1980. A closed‐form equation for predicting the hydraulic conductivity of unsaturated soils. Soil science society of America journal, 44(5), 892-898. Whelan, B. M., Mulcahy, F., 2017. A strategy to instigate SSCM in Australian potato production. In J A Taylor, D Cammarano, A Prashar, A Hamilton (Eds.) Proceedings of the 11th European Conference on Precision Agriculture. Advances in Animal Biosciences , 8 (2), 743-748. doi:10.1017/S2040470017000401 Wickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org. Wierzbicka A. 2014. Zmiany klimatycznego bilansu wodnego w okresie wegetacji ziemniaka w regionie Północnego Mazowsza w latach 1973–2012. Zeszyty Problemowe Postępów Nauk Rolniczych, 576, 207–215. Wösten, J. H. M., Lilly, A., Nemes, A., & Le Bas, C. 1999. Development and use of a database of hydraulic properties of European soils. Geoderma, 90 (3-4), 169-185. Zarzyńska K., Boguszewska-Mańkowska D., Nosalewicz A. 2017. Differences in size and architecture of the potato cultivars root system and their tolerance to drought stress. Plant, Soil and Environment, 63, 4, 159-164. Zebarth B.J., Cambouris A.N., Perron I., Burton D.L., Comeau L.P., & G. Moreau. 2019. Spatial Variation of Soil Health Indices in a Commercial Potato Field in Eastern Canada. Soil Science Society of America Journal, 83, 1786-1798. Zebarth B.J., Fillmore S., Watts S., Barrett R., & L.P. Comeau. 2021. Soil Factors Related to within-Field Yield Variation in Commercial Potato Fields in Prince Edward Island, Canada. American Journal of Potato Research, https://doi.org/10.1007/s12230-021-09825-4. Additional Declarations No competing interests reported. Supplementary Files Tables.docx Supplementary Tables Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7613486","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":523076207,"identity":"91c73183-2c1c-4f60-bf0d-c3f9cb56eea0","order_by":0,"name":"Stanisław Samborski","email":"data:image/png;base64,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","orcid":"","institution":"Warsaw University of Life Sciences","correspondingAuthor":true,"prefix":"","firstName":"Stanisław","middleName":"","lastName":"Samborski","suffix":""},{"id":523076208,"identity":"1c355f6f-2955-4833-b3cb-17c20a0d3172","order_by":1,"name":"Tomasz Gnatowski","email":"","orcid":"","institution":"Warsaw University of Life Sciences","correspondingAuthor":false,"prefix":"","firstName":"Tomasz","middleName":"","lastName":"Gnatowski","suffix":""},{"id":523076209,"identity":"763b188b-32bc-4736-96e8-62b2a004bae0","order_by":2,"name":"Jan Szatyłowicz","email":"","orcid":"","institution":"Warsaw University of Life Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jan","middleName":"","lastName":"Szatyłowicz","suffix":""},{"id":523076210,"identity":"ebf6f78b-6d8a-4aa8-b81e-b6fead7e139f","order_by":3,"name":"Renata Leszczyńska","email":"","orcid":"","institution":"Warsaw University of Life Sciences","correspondingAuthor":false,"prefix":"","firstName":"Renata","middleName":"","lastName":"Leszczyńska","suffix":""}],"badges":[],"createdAt":"2025-09-14 15:23:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7613486/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7613486/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92645513,"identity":"ead51b22-a3e1-481b-a3f8-8a5f6a4bb6ae","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5682910,"visible":true,"origin":"","legend":"","description":"","filename":"Howdosoil...S.Samborskietal14.09.2025submitted.docx","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/180d149208f68a956c354345.docx"},{"id":92645508,"identity":"8ab6e8a0-21a5-48c0-a191-3a30ada72401","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6495,"visible":true,"origin":"","legend":"","description":"","filename":"661bc444b33c442d86e9bdab34927659.json","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/4eb9f69eed76d31c8221bf45.json"},{"id":92646247,"identity":"7afa727d-14c5-4657-9ed5-59edfe4052e1","added_by":"auto","created_at":"2025-10-02 10:00:28","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150641,"visible":true,"origin":"","legend":"","description":"","filename":"661bc444b33c442d86e9bdab349276591enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/9c1fda16f435521d7110b54f.xml"},{"id":92645517,"identity":"2abcf51e-1a07-4bb0-9fe2-f1f3013d74ae","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313536,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/659e5ad2bffd19e0b728ceac.jpeg"},{"id":92645514,"identity":"5a0c6ed7-bbd8-458a-9bf0-54feda0de742","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":98535,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/0ac8255b5783dddcee9aca83.png"},{"id":92646504,"identity":"863ee465-b70a-4aea-b633-f109a1c7cd44","added_by":"auto","created_at":"2025-10-02 10:08:28","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":344218,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage11.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/c03ef996b9d04d88c0dc2744.jpeg"},{"id":92646251,"identity":"d5c7f799-eb7a-42e3-a331-c02d46572e74","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":139361,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/2c6a3ce7dfc766fbdc9e33cc.png"},{"id":92646254,"identity":"b11bfff9-696a-4022-b0f7-0b5784c550c5","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1505548,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/37b8cb4ebf8b02dbbbfe5e18.png"},{"id":92645519,"identity":"5b88900e-d83d-44b9-8e63-e08f3c8bde79","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1434177,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/b12e44af795f0ceaa9d59075.png"},{"id":92646253,"identity":"9f7da4c4-283a-48f1-9024-51705d7a8aa0","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":801119,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/6b11a868b12d19db2a6bebdc.png"},{"id":92646259,"identity":"cd1cc422-06ea-4db1-9ab9-a444b63eef20","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":569545,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/6a17601cb85d0366688ef87f.png"},{"id":92645534,"identity":"7bf3283d-cad6-42f6-bd83-9e39b3a0e0d0","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135916,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/c6f6dfbee41d4315036805c8.png"},{"id":92645539,"identity":"7b47899a-8e9f-4aa7-87d1-fcb074f1d2b7","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134522,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/8bdf69147aab71990e706fa9.png"},{"id":92645520,"identity":"51234ae1-aa6e-4e59-b945-c817e1beff56","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":103037,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/39159075f392e2370f6c041f.png"},{"id":92646258,"identity":"77bf71a3-356b-4ae8-97ee-bfdcbe2166fd","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":79322,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/7401e4e8fe4cd346c677c868.png"},{"id":92645535,"identity":"927f4cf7-76c7-4950-9fa1-a20996bde7eb","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":30437,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/a9093d9bdda7b5bf3dc810b9.png"},{"id":92646260,"identity":"39e33c9d-4c4f-46eb-b876-eac7e93715d7","added_by":"auto","created_at":"2025-10-02 10:00:29","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":274468,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/17ba7be2d32286da1a9abc22.png"},{"id":92645529,"identity":"418ff229-47b3-419f-a34b-ad5d347b2727","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62517,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/8108ba9eef47dead2d310c6a.png"},{"id":92645542,"identity":"ee4f8e00-cc3b-43e6-b1df-112b8042fca8","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":229972,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/e08e4c6ff9d18af97f5d7694.png"},{"id":92645541,"identity":"36bf523f-ee1b-4195-bc69-1173a4fb899f","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":215828,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/7e924fa604d1a05f5e36ce13.png"},{"id":92645533,"identity":"3d9f73b9-3cb2-41da-8283-59475317b418","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":512216,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/56e9b0ecc7f4ff4ba1232b39.png"},{"id":92645525,"identity":"d674d185-641c-4788-b330-1d484741fb2d","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":277665,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/919aa505b98cfb37dbfdb8c0.png"},{"id":92645527,"identity":"90baa715-6cc1-4725-952f-989d691aad3b","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37340,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/1a80629e86d671a28b824255.png"},{"id":92645531,"identity":"e7c4b234-cf3d-4745-bc3c-1f1eb2bc19a8","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":38576,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/d2b3f1d7de885577971db921.png"},{"id":92645532,"identity":"21ea0588-cc06-453a-a848-a9d86505f0dc","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":30804,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/db36cd3281f3de0c7b8b6faf.png"},{"id":92645537,"identity":"a8638fff-91b1-4d96-b2db-304b86f32934","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"xml","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":151112,"visible":true,"origin":"","legend":"","description":"","filename":"661bc444b33c442d86e9bdab349276591structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/e85ba1f519abec99a82d46e3.xml"},{"id":92645543,"identity":"044d592c-8011-43bb-a826-4a6fc81930a0","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"html","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":159773,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/9421e70073d5d3a6e8915801.html"},{"id":92646246,"identity":"35d510e4-3c71-4fa1-9247-627e29b8c7d0","added_by":"auto","created_at":"2025-10-02 10:00:28","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":88561,"visible":true,"origin":"","legend":"\u003cp\u003eMap of soil and plant sampling areas with superimposed information on soil texture\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/1dee3b1c66fa6a9265f212e5.png"},{"id":92645507,"identity":"481019f9-af4b-46fd-867d-d8a7cb1cdd63","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":91541,"visible":true,"origin":"","legend":"\u003cp\u003eDecade averages of maximum, minimum, and mean air temperatures and precipitation with dates of planting, installation of soil moisture sensors (3\u003csup\u003erd\u003c/sup\u003e of June), growth stages determination for NDVI acquisition from May to August of 2022, using UAV, and harvest.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/4ebd1c6f1e1ddd713b805cbb.png"},{"id":92645510,"identity":"1e829f29-91ad-4706-a859-d40bc70556d9","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":494541,"visible":true,"origin":"","legend":"\u003cp\u003eIllustration of a single 3 m\u003csup\u003e2\u003c/sup\u003e (4 m x 0.75 m), soil and potato sampling area: a) part of a four-meter potato ridge, white horizontal, dotted arrow indicates sampling area length, three vertical, short arrows indicate where soil samples in three replications were taken and resistance measured, b) measurement of soil resistance using a Penetrologger, c) taking undisturbed soil samples for bulk density determination, d) five soil subsamples taken for chemical and texture analysis\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/9e4cfad6a118b983f3f8dedc.png"},{"id":92645522,"identity":"33a31fda-ef26-42ab-bd63-b73532a531a0","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":499034,"visible":true,"origin":"","legend":"\u003cp\u003eInstallation of: a) soil moisture sensors at three depths of 10, 20, and 30 cm, b) a node with three soil sensors connected\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/69de18fe3db9fb7472a80d11.png"},{"id":92646248,"identity":"d862e7c3-4f02-4820-85fe-8de99c4ecc5f","added_by":"auto","created_at":"2025-10-02 10:00:28","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":933430,"visible":true,"origin":"","legend":"\u003cp\u003eMeasured soil water storage in the soil layer of 40 cm in four (p2, p4, p8 and p11), selected soil sampling areas (of various yield), calculated SWS at field capacity, limiting and wilting points, during the growing season (days after planting – DAP), against the background of reference evapotranspiration and precipitation.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/cdfce0acc87c0f94cce78450.png"},{"id":92645515,"identity":"dd8e9a5a-56ff-442b-bc21-e2480ccd0b4f","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":439824,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation coefficients between descriptive variables and three potato traits. The dotted, red vertical lines indicate the limits of the statistically significant \u003cem\u003er\u003c/em\u003e-values and \u003cem\u003ep\u003c/em\u003e-values lower than or equal to 0.05. The solid vertical lines indicate the \u003cem\u003er \u003c/em\u003evalues equal to 0. The labels on the left side of the marker (dot) refer to the \u003cem\u003er\u003c/em\u003e-value, and the labels on the right side indicate the \u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/3f8da9194bd888e49b8c7b0f.png"},{"id":92645516,"identity":"cd820d34-226c-4f6c-ae15-d7a69b3a662c","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":144092,"visible":true,"origin":"","legend":"\u003cp\u003eThe heat map of the correlation coefficients between descriptive variables and potato traits (ATW, DM, and TY), which were higher than 0.50. The black labels refer to the correlation coefficient value, whereas the yellow labels indicate the p-value\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/98e9ac98b0bdba1a4b231e25.png"},{"id":92646509,"identity":"04f2d71a-c358-4fa1-9b97-d87ba2734ab5","added_by":"auto","created_at":"2025-10-02 10:08:28","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":128007,"visible":true,"origin":"","legend":"\u003cp\u003eRedundancy analysis (RDA) graph explaining the influence of the best correlated descriptive variables out of all considered variables in Figure 6, on potato traits. Labels A\u003csub\u003ei\u003c/sub\u003e indicate the location of the twentieth sampling areas in model M1.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/428b3067ea6f5b71a4630d95.png"},{"id":92645524,"identity":"0ad5ad2f-3221-41b9-856c-a972d02d1299","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":129466,"visible":true,"origin":"","legend":"\u003cp\u003eRedundancy analysis (RDA) graph explaining the influence of the physical properties on potato traits. Labels A\u003csub\u003ei\u003c/sub\u003e indicate the location of the twentieth sampling areas in model M2.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/41646d5f376ba55feb13e4ad.png"},{"id":92646250,"identity":"9d80a3a8-284d-4e33-b21f-1b300f1e3794","added_by":"auto","created_at":"2025-10-02 10:00:28","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":106270,"visible":true,"origin":"","legend":"\u003cp\u003eRedundancy analysis (RDA) graph explaining the influence of the chemical properties on potato traits. Labels A\u003csub\u003ei\u003c/sub\u003e indicate the location of the twentieth sampling areas in model M3.\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/bc9c43c072264e201ff20537.png"},{"id":92645526,"identity":"b0c403d4-b4fc-46a9-9738-621e8d05516c","added_by":"auto","created_at":"2025-10-02 09:52:29","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":98346,"visible":true,"origin":"","legend":"\u003cp\u003eRedundancy analysis (RDA) graph explaining the influence of soil water storage and NDVI on potato traits. Labels A\u003csub\u003ei\u003c/sub\u003e indicate the location of the twentieth sampling areas in model M4.\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/190dea8fe3168b9a11180cc2.png"},{"id":97896015,"identity":"a8e7dc08-1f5b-4d1c-9d2d-13ad846e61b5","added_by":"auto","created_at":"2025-12-10 15:35:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3610293,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/d67e89b0-7269-4c6c-915a-fbf889a80b6a.pdf"},{"id":92645505,"identity":"b560891c-52f5-45e7-ae27-8b74497f9fd4","added_by":"auto","created_at":"2025-10-02 09:52:28","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":17488,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Tables\u003c/p\u003e","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7613486/v1/6dc2f2450cdf8c7bcd95effc.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"How do soil properties and canopy status influence potato traits while continuously monitoring soil moisture with IoT-based sensors?","fulltext":[{"header":"Introduction","content":"\u003cp\u003eEurope is responsible for about 26.2% of the world\u0026rsquo;s potato production (FAOSTAT 2024). Poland is one of the main potato producers in central Europe. Over the last two decades, the average potato yield in Poland has increased by 0.77 tons per year (Dzwonkowski \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). But, in some years with water shortages and uneven rainfall distribution, the average national potato yield decreases significantly. This is due to water stress \u0026ndash; a negative balance of water supply compared to evapotranspiration (Dubois et al. 2020), which is expressed as the Climatic Water Balance (CWB), provided by the Agricultural Drought Monitoring System \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://susza.iung.pulawy.pl/en/system\u003c/span\u003e\u003cspan address=\"https://susza.iung.pulawy.pl/en/system\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025). Moreover, water stress also negatively affects tuber quality, leading to their malformations and thus reduction of marketable yield (Abbas and Ranjan 2015). Potato is considered a drought-sensitive crop due to possessing a sparse, shallow, and weak soil-penetrating root system. Thus, according to Zarzyńska et al. (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), potato cultivars that develop a deeper root system in drought conditions yield better. Additionally, potatoes yield well in sandy and sandy loams with low to medium water-holding capacity. Due to a wide representation of very light (mainly sands according to USDA, 28% of the area of agricultural land) and light (mainly loamy sands and some sandy loams, 30%) soils (Łopatka, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), these texture-type soils are commonly used for growing potatoes in Poland. Moisture needs of potato are met at the very early stage by the mother tuber because it is assumed that the role of this potato part as a deliverer of energy stops when the leaf area index (LAI) exceeds 0.75 (Haverkort et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). At the latter growth stages, moisture is supplied either from the reserve of conserved water in the soil or through different irrigation scheduling methods (Rai and Dong \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), or both. According to the Polish Potato Association, Poland's total potato irrigated area currently accounts for approximately 25%. The average potato yields on these farms, which mainly grow potatoes for processing, are around 50\u0026ndash;60 t\u0026middot;ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, thus considerably higher than the country average yield of 31 t\u0026middot;ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in 2024 (Dzwonkowski, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). However, many potato fields are only rainfed due to limited access to water supplies and unaffordable irrigation systems, which are out of reach for smaller producers. Current developments in the Internet of Things (IoT), among many applications in agriculture, enable continuous monitoring of soil moisture deficit (Dhal et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), a crucial factor in controlling tuber yield and quality of potatoes (Abbas and Ranjan 2015). However, research studies on probes installed at different depths on potato fields to provide real-time soil moisture readings have been limited (Abbas and Ranjan, 2015). Usually, soil moisture sensors are installed in a limited number, only in areas representing different irrigation treatments (Abbas and Ranjan 2015; Liao et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Dubois et al., 2020), but not to monitor the spatial variability of soil moisture due to variations in soil texture (ST). Consequently, the same soil moisture level for the field parts of different STs results in varying soil water storage (SWS) and thus water availability for the plants. Therefore, knowledge of spatial SWS variability within a potato field and temporal variability of SWS over the growing season could help explain the variation in potato yield and other traits. It is worth emphasizing that soil SWA for plants. According to Utset et al. (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2000\u003c/span\u003e), critical values of matric potential for potatoes range from \u0026minus;\u0026thinsp;32 kPa to -68 kPa, depending on evapotranspiration and depth of the root system. Results of the study conducted by Matteau et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), in a greenhouse on potatoes grown on sandy soils, suggest that the soil matric potential for maximizing potato yield is located between \u0026minus;\u0026thinsp;10 and \u0026minus;\u0026thinsp;24 kPa. The results obtained by Redulla et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) suggest that many soil-related factors, which are also associated with potato yield and quality, are implicitly linked to SWA. Soil water retention and transport depend on soil properties, climatic variabilities, and water management if irrigation is provided (Alva \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Still, the majority of farmers apply uniform fertilizer rates across their fields, including potato crops. However, due to common and intrinsic spatial soil variability, actual conditions for growing potatoes are site-specific (Po et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Whelan and Mulcahy \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Zebarth et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e and \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Sheng et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In addition to traditional soil and plant sampling methods used to determine soil variability, sensor-based solutions have also been available. Farooque et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) used soil electrical conductivity (EC) measurements to predict potato yields on commercial fields in Canada, assuming that soil EC could be an indirect indicator of soil properties (Corwin and Lesch, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Among the remote sensing methods, unmanned aerial vehicles (UAVs) enable the evaluation of soil and plant state using vegetation indices (VIs), which are indirect measures of potato canopy growth and yield. Thus, they can be integrated into potato models to enhance the yield prediction accuracy (Mukiibi et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe objective of this research was to quantify the optimal explanatory soil- and canopy state-related properties that determine potato traits, i.e., total yield (TY), and quality indicators: dry matter (DM) content, and average tuber weight (ATW) over a production field. The originality of the study refers to continuous monitoring of SWA for potato plants using in-field IoT soil moisture sensors installed at three depths in each sampling area of various STs. Such sensors enabled the measurement of fluctuations in soil moisture levels in both space and time. Based on the integration of SMC, SWS was determined in the 0\u0026ndash;40 cm layer (where the majority of the potato root system is concentrated), during critical stages for yield formation, namely with the tuber initiation and tuber bulking stages being the most sensitive to water stress (Thornton \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Thus, SWS was treated as one of the soil-related explanatory variables under rainfed conditions. Therefore, the second objective of this study was to analyze how the changes in SWS affected TY.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eSite Description and Crop Management\u003c/h2\u003e\u003cp\u003eThe climate of the study site has been classified as humid continental with cold winters and hot summers (Błaś and Ojrzyńska \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The 30-year average annual precipitation is 550 mm, of which 313 mm is received during the potato growing season, from May to the end of September. The average annual air temperature is 94\u0026deg;C, while the average air temperature over the latter months is 174\u0026deg;C (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://klimat.imgw.pl/pl/climate-normals\u003c/span\u003e\u003cspan address=\"https://klimat.imgw.pl/pl/climate-normals\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The study was conducted in 2022 on a part of a commercial, rainfed potato production field of 8.88 ha located in central Poland (52\u0026deg;4' 50.94\" N, 21\u0026deg;10' 6.73\" E), which belongs to the Research Farm of Warsaw University (Wilan\u0026oacute;w-Obory) of Life Sciences., The field is 87 m above sea level and is characterized by flat topography (range of elevation from 86.91 to 87.96 m). The soils corresponded to Fluvisols, with an average organic matter content of 1.77%, and a pH in KCl solution of 6.02. Potato is grown on this field in a 4-year rotation, and the previous crop was winter rape (\u003cem\u003eBrassica napus\u003c/em\u003e L.). Fertilizer (N, P, and K) application was at preplant: 150 kg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (0\u0026ndash;0\u0026ndash;60) on March 14th, 107 kg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (18\u0026ndash;46\u0026ndash;0) on April 22nd. Three weeks after planting, fertilizer (N, S) was applied at the rate of 120 kg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (26\u0026ndash;32.5), and ammonium nitrate (33.5%) at the rate of 52 kg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. All fertilizers were applied at uniform rates.\u003c/p\u003e\u003cp\u003eThe field was planted on 9 May 2022 with a semi-early \u003cem\u003eHermes\u003c/em\u003e potato (\u003cem\u003eSolanum tuberosum\u003c/em\u003e L.) cultivar, tolerant to a combination of heat and drought stress (Siano et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Hermes is the third most popular potato variety grown in Poland. Beds were formed with 75 cm between rows, and the planting distance was 30 cm. The orientation of crop rows was west-east. Management practices were conducted uniformly across the field throughout the growing period.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eDelineation of soil and tuber sampling areas\u003c/h3\u003e\n\u003cp\u003eTwenty soil, and potato sampling areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), were placed in locations of different growing conditions, thus of various potato yield potential, using a yield map of winter rape from 2021, and a historical soil map at a scale of 1:5000. A similar approach to the location of the soil moisture sensors, but within a soybean field, was used in the study by Pan et al. (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe precise location of the sampling areas was adjusted to the centroids of Sentinel-2 imagery pixels, which were used in another project. A GNSS-RTK receiver (Topcon GRS-3, Topcon Corporation, Tokyo, Japan), with a 3 cm accuracy specification, was used for navigation to the centroids.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eWeather conditions\u003c/h3\u003e\n\u003cp\u003eRainfall was registered by a rain gauge located \u003cem\u003ec\u003c/em\u003e. 1.5 km from the research field, average air temperature, solar radiation, relative air humidity, and wind speed were obtained from a weather station placed at the Warsaw University of Life Sciences campus, \u003cem\u003ec\u003c/em\u003e.13 km from the research field. The average air temperature from May to the end of August 2022 was 19.2\u0026deg;C, (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The total precipitation during the respective period was 257 mm. The rainfall distribution was 84 mm during May and June and 173 mm in July and August. Measured meteorological data were used to compute reference evapotranspiration (ET), using Penman-Monteith formulae (Allen et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e1998\u003c/span\u003e)\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:ET=\\frac{0.408\\:{\\Delta\\:}{\\:R}_{n}+{\\gamma\\:}\\:\\frac{900}{t+273\\:}v\\:D}{{\\Delta\\:}+\\:\\left(1+0.34v\\right)}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere:\u003c/p\u003e\u003cp\u003e\u003cem\u003eET\u003c/em\u003e\u0026ndash; reference evapotranspiration [mm d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e],\u003c/p\u003e\u003cp\u003e\u003cem\u003eR\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e \u0026ndash; net radiation at the crop surface [MJ m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e],\u003c/p\u003e\u003cp\u003e\u003cem\u003e∆\u003c/em\u003e \u0026ndash; slope vapour pressure curve [kPa \u0026deg;C\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e],\u003c/p\u003e\u003cp\u003eγ \u0026ndash; psychrometric constant [kPa \u0026deg;C\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e],\u003c/p\u003e\u003cp\u003e\u003cem\u003et\u003c/em\u003e \u0026ndash; mean daily air temperature [\u0026deg;C],\u003c/p\u003e\u003cp\u003e\u003cem\u003ev\u003c/em\u003e \u0026ndash; wind speed [m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e],\u003c/p\u003e\u003cp\u003e\u003cem\u003eD\u003c/em\u003e \u0026ndash; saturation vapour pressure deficit [kPa].\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe climatic water balance (CWB), in other words, a precipitation deficit, was calculated for the potato growing season and decades, as the difference between precipitation (mm) and reference evaporation (mm) (Wierzbicka \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003ePotato was grown mainly on medium (ST of sandy loam), and light (ST of loamy sand), soils (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). According to Chmura et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), to obtain maximum yields of semi-early cultivars grown in Poland on medium soils, potato requires 88 mm of rainfall in May and June and 220 mm in the second part (July-August) of the season. The optimal rainfall distribution for growing semi-early potato cultivars on light soils is 143 mm during May and June and 220 mm in July and August. Thus, the experiment was conducted in a year of suboptimal water supply for potatoes cultivated on light soils during the whole season and optimal conditions in only the first two months of potato growth on medium soils.\u003c/p\u003e\n\u003ch3\u003eUAV Flight Data Collection\u003c/h3\u003e\n\u003cp\u003eThe aerial-based data was collected by UAV (Phantom 4 Multispectral, DJI, Shenzhen Dajiang Baiwang Technology Co., Ltd., China), equipped with a highly accurate Global Navigation Satellite System with RTK (Real Time Kinematic) correction. Flights were conducted five times with 2-week intervals at BBCH growth stages (described by Meier \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e, and given in the brackets), starting from mid-season as recommended by Whelan \u0026amp; Mulcahy (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), on June 22nd (BBCH 39 \u0026ndash; canopy closure, tuber initiation, the largest tubers of 25 mm); July 6th (BBCH 61 \u0026ndash; beginning of flowering, the largest tubers of 40 mm), July 20th (BBCH 69 \u0026ndash; end of flowering, tuber size in the range of 35\u0026ndash;55 mm); 3 August (BBCH 70\u0026ndash;79 \u0026ndash; development of fruit), and 17th (BBCH 81\u0026ndash;-89 \u0026ndash; ripening of fruit and seed). Each UAV flight was conducted perpendicularly to the direction of sunlight, close to solar noon, on cloud-free days, to avoid changing light conditions. The UAV altitude was c. 80 m above ground level, providing a 4 cm pixel\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e resolution. Images were taken by DJI\u0026rsquo;s FC6360 camera with 75% front and side overlap. Ortho-photo images for each spectral band (blue, green, red, red-edge, and NIR) and NDVI (Normalized Difference Vegetation Index) were generated using the Pix4DFields computer software (Pix4D S.A., Prilly, Switzerland). NDVI, which is commonly used for monitoring crop canopy but does not require sophisticated calculations, was applied based on red (650 nm\u0026thinsp;\u0026plusmn;\u0026thinsp;16 nm) and near-infrared (840 nm\u0026thinsp;\u0026plusmn;\u0026thinsp;26 nm) bands (Rouse et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e1973\u003c/span\u003e). Further processing of the ortho-photo NDVI images was conducted using QGIS software (QGIS.org 2022). A polygon-shaped file with a rectangular region of interest (ROI) of the same size, 0.75 m (width) by 4.0 m (length), was created for each sampling area to extract average NDVI values using zonal statistics.\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003eSoil sampling and analyses\u003c/h2\u003e\u003cp\u003eAt harvest on September 12 and 13th, 2022, soil samples for physicochemical analyses were taken before digging the potatoes, and soil resistance (SR) was measured in three locations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). The latter was done with the use of a Penetrologger (Eijkelkamp, Giesbeek, the Netherlands) to a depth of 80 cm, and then averaged for the depths of 0\u0026ndash;10, 11\u0026ndash;20, 21\u0026ndash;30, 31\u0026ndash;40, 41\u0026ndash;50, 51\u0026ndash;60, 61\u0026ndash;70, and 71\u0026ndash;80 cm, and additionally for the plow layer of 0\u0026ndash;30 cm (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). After that, undisturbed soil samples for bulk density (BD) determination were taken at the three locations to a depth of 15 and 25 cm using 100 cm\u0026sup3; cylinders, where most potato tubers develop (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Finally, five soil subsamples were taken from each of the three locations (15 subsamples in total), per sampled row for soil chemical and texture analyses (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe content of available forms of the following nutrients: P, K, Mg, Fe, Na, Cu, Mn, Zn, and Co in soil samples, sieved through a 2-mm mesh, was determined using the Mehlich-3 method (Sikora \u0026amp; Moore, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The Avio 200 ICP optical emission spectrometer (PerkinElmer, Inc., Waltham, MA, USA) was used to determine the contents of the above nutrients. Soil pH was measured in a 1 mol KCl solution using pH-METER CG 842 (Hofheim, SCHOTT, Ger\u0026auml;te GmbH, Germany). The total contents of C, N, and S were determined using the Elementar Vario MacroCube analyzer (Elementar-Stra\u0026szlig;e 1, 63505 Langenselbold, Germany).\u003c/p\u003e\u003cp\u003eSoil samples sieved through a 2-mm mesh were subjected to particle size distribution analysis using the hydrometric (areometric) method of Casagrande modified by Pr\u0026oacute;szyński (Musierowicz \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e1950\u003c/span\u003e). The content of soil separates was determined according to the USDA classification.\u003c/p\u003e\u003cp\u003eSoil EC measurements were done on 28 October 2022 with a MSP-3 platform (Veris Technologies, Salina, Kansas, USA), at depths of 30 cm (EC\u003csub\u003esh\u003c/sub\u003e) and 90 cm (EC\u003csub\u003edp\u003c/sub\u003e). For statistical analysis, averaged EC data from a buffer of 20 m in diameter of each sampling area were used.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eSoil moisture monitoring\u003c/h3\u003e\n\u003cp\u003eOn 3 June 2022, soil moisture sensors of Yosensi Ltd. company (Białystok, Poland), were horizontally mounted at three depths of 10, 20, and 30 cm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea), in the middle of each sampling area (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). Potato plants at that time were at the leaf development stage (BBCH 10\u0026ndash;19) \u0026ndash; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The SMC measurements were conducted using capacitive probes operating at a low frequency of 75 MHz. The primary measurable parameter is the variable dielectric permittivity, expressed as voltage in the 0 to 3.3 VDC range, where higher voltage corresponds to higher SMC and \u003cem\u003evice-versa\u003c/em\u003e. The sensor\u0026rsquo;s measuring element (capacitor linings) is made of laminate. The electronics are vacuum-sealed in epoxy resin, ensuring the sensor is waterproof and resistant to various weather conditions. The probe is 2 mm thick, 30 mm wide, and 150 mm long. The allowable operating temperature of the probe is between \u0026minus;\u0026thinsp;40℃ and +\u0026thinsp;60℃. The soil moisture probes are connected to a node (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) with an in-built wireless communication module, which gathers data, forms the payload, and sends it to a gateway. This device, similar to a router, is equipped with a LoRa (Long Range) concentrator that receives LoRa packets and sends them to the Internet-connected server. Using the calibration equation (developed for the Research Farm of Warsaw University of Life Sciences, located in Obory, unpublished data), the output of the probes, expressed as voltage, was converted to volumetric moisture content measured hourly and then averaged to obtain daily data. Based on the integration of the SMC, an average soil water storage (SWSavg) in the observation period, i.e., between 3 June (25 DAP \u0026ndash; days after planting) to 3 August (102 DAP), was determined in the layer of 0\u0026ndash;40 cm. The calculated values of SWS were compared with those at characteristic moisture conditions, i.e., water content at field capacity, limiting, and wilting points. The pedotransfer functions proposed by W\u0026ouml;sten et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1999\u003c/span\u003e) were used to assess the soil water retention curve. Using measured data on the content of soil separates, organic carbon content, and soil bulk density, the parameters of the van Genuchten (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e1980\u003c/span\u003e) equation that describes soil water retention curve were calculated. Computation of soil moisture at the limiting and wilting points was performed assuming soil matric potential equal to -500 cm for the limiting point and \u0026minus;\u0026thinsp;15848 cm for the wilting point, respectively. A physically based analytical equation derived by Assouline and Or (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) was used to predict SMC at field capacity.\u003c/p\u003e\u003cp\u003eFor the need of calculating the relationship between descriptive variables and potato traits, SWS was calculated as an average for the following periods:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eSWSaP1 from BBCH 10\u0026ndash;19 to BBCH 39,\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSWSaP2 from BBCH 39 to BBCH 61,\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSWSaP3 from BBCH 61 to BBCH 69,\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSWSaP4 from BBCH 69 to BBCH 70\u0026ndash;79,\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSWSaP5 from BBCH 69 to BBCH 81\u0026ndash;89,\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eAdditionally, for the same periods, the minimum (SWS\u003csub\u003emin\u003c/sub\u003e), maximum (SWS\u003csub\u003emax\u003c/sub\u003e), range (SWS\u003csub\u003erange\u003c/sub\u003e), and standard deviation (SWSstd) of SWS were determined.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003ePotato sampling and potato trait determination\u003c/h3\u003e\n\u003cp\u003eOn 31 August, the potato crop was defoliated. Potatoes were manually harvested at each sampling area of 3 m\u003csup\u003e2\u003c/sup\u003e, by collecting tubers along a 4 m distance from a row of 0.75 m width (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). A total yield (TY) \u0026ndash; t\u0026middot;ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and the total number of tubers per sample were determined. The ratio of these two traits was used to calculate the average tuber weight (g) \u0026ndash; ATW. Dry matter content in tubers, an equivalent of specific gravity, was determined using a hydrometric method (a weight-in-air/weight-in-water method). A potato sample of 3.64 kg was weighed in air and then reweighed in water. DM content was calculated based on the differences of these two weights (Kleinschmidt et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1983\u003c/span\u003e).\u003c/p\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eThe statistical analysis included the calculation of correlations between explanatory and response variables obtained for the twenty sampling areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), with the determination of statistical significance represented by the p-value below 0.05. The set of explanatory variables was described by soil chemical and physical properties, including measured SWS (i.e., SWSaP1, SWSaP2, SWSaP3, SWSaP4, and SWSaP5), in a 40 cm soil layer. Additionally, NDVI values derived from UAV imagery, an indirect indicator of plant biomass and chlorophyll content during plant development, shaped by the soil properties from plant emergence to plant senescence, were also used as the explanatory (descriptive) variables. The response variables were represented by potato traits determined at harvest: TY, ATW, and DM content.\u003c/p\u003e\u003cp\u003eFor the multidimensional analysis, redundancy analysis (RDA) was employed, enabling the examination of the effects of explanatory variables on the set of response variables. The applied RDA is an extended Principal Component Analysis (PCA) method, which enables the ordination of Y to obtain ordination axes that are linear combinations of the variables in X (Legendre \u0026amp; Legendre, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The statistical method of RDA also involved reducing multicollinearity between explanatory variables in the X matrix and the explained potato traits (Y matrix). Multicollinearity reduction refers to removing highly correlated variables from a model to enhance its accuracy. On the other hand, ordination axes are the new variables created through the PCA method that best explain the variation in the original variables. The RDA was performed using the R-package vegan v2.6-10 (Oksanen et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) in the RStudio interface (RStudio Team, 2020). The graphical representation of the data set was prepared using ggplot2 (Wickham, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Before the RDA was performed, the whole data set was standardized with the method of standard deviation from the mean as follows:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{Z}_{ij}=\\frac{{O}_{ij}-{O}_{avg}}{{\\sigma\\:}_{i}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere Z\u003csub\u003eij\u003c/sub\u003e is the standardized value of the i-th response variable, including also tuber quality for the j-th sampling area; Oij is the original value of the i-th response variable for the j-th sampling area; O\u003csub\u003eavg\u003c/sub\u003e is the average value of the i-th analysed variable in all sampling areas; and σ is the standard deviation of the i-th variable. The R\u003csup\u003e2\u003c/sup\u003e adj was used to assess the performance of the RDA analyses.\u003c/p\u003e\u003cp\u003eThe modeling part of the study plays a crucial role in establishing efficient forecasting models for the potato traits (response variables): TY, ATW, and DM content at a field scale. The study considered four models (strategies): M1, M2, M3, and M4, each with a unique approach to understanding and predicting the potato traits. Strategy M1 assumed that selecting the descriptive variables based on the highest correlation (\u003cem\u003er\u003c/em\u003e value above 0.50) enables a proper prediction of the potato traits using the RDA. This strategy evolved because the number (48) of descriptive variables was higher than the number of sampling areas (n\u0026thinsp;=\u0026thinsp;20). However, such an extensive range of explanatory variables is not standard practice. Very often, the number of the explained variables is limited only to, e.g., soil physical properties. Therefore, we examined three other models (strategies) for the potato traits. The M2 strategy assumed that the input variables represent only physical soil properties during the model development. The next model (M3) investigated the effect of chemical soil properties on potato traits alone. Finally, the last considered model (M4) assumed that the potato traits depend only on the indirect indicator of plant development (NDVI) and SWS properties.\u003c/p\u003e\u003cp\u003eAdditionally, the Leave-one-out (LOO) cross-validation (CV) scheme (James et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) was applied to assess the RMSEcv error of the potato traits. The quality of the final model was evaluated using a ranking procedure. This procedure is based on the statistical measures obtained from developing and testing (LOO scheme) of the analyzed models. During the development of the models, the adjusted coefficients of determination were determined (adjusted R\u003csup\u003e2\u003c/sup\u003e) for the standardized values used in the RDA. The root mean squared error (RMSE) was also calculated, based on untransformed data from the prediction and the original data of the potato traits.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results and discussion","content":"\u003ch2\u003eVariability of potato traits\u003c/h2\u003e\n\u003cp\u003eThe average potato yield of all the sampling areas was 45.5 t∙ha\u003csup\u003e-1\u003c/sup\u003e, and the yield range was 16.2. to 67.1 t∙ha\u003csup\u003e-1\u003c/sup\u003e (Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1. Descriptive statistics for variation in total potato yield (t∙ha\u003csup\u003e-1\u003c/sup\u003e), average tuber weight (g), and dry matter content (%), n = 20, for the commercial potato field in 2022.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"634\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29.0221%;\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.9338%;\"\u003e\n \u003cp\u003eAverage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003eStandard Deviation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003eCoefficient of variation (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.2492%;\"\u003e\n \u003cp\u003eRange\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.0221%;\"\u003e\n \u003cp\u003eTotal yield (t\u0026middot;ha\u003csup\u003e-1\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.9338%;\"\u003e\n \u003cp\u003e49.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e13.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e26.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.2492%;\"\u003e\n \u003cp\u003e16.2-67.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.0221%;\"\u003e\n \u003cp\u003eAverage tuber weight (g)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.9338%;\"\u003e\n \u003cp\u003e91.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e23.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e25.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.2492%;\"\u003e\n \u003cp\u003e45.1-132\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 29.0221%;\"\u003e\n \u003cp\u003eDry matter content (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.9338%;\"\u003e\n \u003cp\u003e22.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e0.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.3975%;\"\u003e\n \u003cp\u003e4.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13.2492%;\"\u003e\n \u003cp\u003e20.8-24.2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eAmong the determined potato traits, total yield and ATW were characterized by similar coefficient of variation (CV), 26.2 and 25.4 %, respectively. Zebarth et al. (2019) found a CV of yield of 18% within a 21-ha rain-fed field in Canada. Much greater yield variability within a field, of 3.5-fold, for sixteen fields, was observed by Whelan and Mulcahy (2017) in Tasmania. Quality potato trait \u0026ndash; DM content was much less variable among the twentieth sampling areas and had a CV value of 4.06 %. Kleinschmidt et al. (1983) summarized that sandy and heavy clay soils produce potatoes with lower specific gravity (DM content) than medium textured soils. In the case of the sampling locations with sandy (no 2) and loam (no 7), soil texture (Figure 1), the DM content was respectively 22.0% and 22.6%, thus very close to the field average. Beukema and van der Zaag (1990) observed that fields that are homogeneous in soil and fertility produce crops with less variation in DM content than fields with higher spatial variability.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eVariability of soil chemical and physical properties\u003c/h2\u003e\n\u003cp\u003eAll the results of the soil analyses are presented as supplementary material in Tables 1s, 2s, and 3s. Among all soil chemical properties (Table 1S), acid soil and low fertility in Mg, Mn, and Zn, which could have limited potato yield, were observed only in soil sampling areas 2 and 9 (Kęsik 2016; Korzeniowska et al. 2021). However, very different yields were achieved in these two areas, respectively 16.2 and 57.7 t∙ha\u003csup\u003e-1\u003c/sup\u003e (unpresented data). This may indicate that other factors also limited yield in the sampling area number 2. Sheng et al. (2023) observed that irrigation overshadowed the effect of nutrients on potato yields obtained on production fields in northern China.\u003c/p\u003e\n\u003cp\u003eThis field, according to the ST map (Figure 1), is mainly composed of sandy loam (western part) and loamy sand ( eastern part) in a plough layer. Sand and loam texture occur only in single sampling areas, 2 and 7, respectively. Soil resistance (an indicator of compaction status) below the plow layer was characterized by average values of above 2.1 MPa, which could be considered a factor that limits deeper soil penetration by potato roots and causes poor drainage and poor deep water storage (Table 2s), Moebius-Clune et al. (2017). Both soil layers (0-30 and 0-90 cm) were characterized by similar average and range electrical conductivity values. Only in a few sampling areas, EC for the deeper layer was slightly higher than the EC measured in the shallow layer (data not presented). This suggests no significant changes in ST within a soil profile.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe average value of soil bulk density at a depth of 15 cm was 1.40 g cm\u003csup\u003e-3\u003c/sup\u003e, and it ranged from 1.30 (sampling area no 2) to 1.50 g cm\u003csup\u003e-3\u003c/sup\u003e (sampling area no 12) (unpresented data). In the deeper layer (25 cm), the soil was characterized by a higher bulk density, with an average value of 1.49 g cm\u003csup\u003e-3\u003c/sup\u003e, ranging from 1.36 g cm\u003csup\u003e-3\u003c/sup\u003e (sampling area no 13) to 1.63 g cm\u003csup\u003e-3\u003c/sup\u003e (sampling area no 1) (Table 3s). The increase in soil density with depth of potato ridges was confirmed by Antoneli et al. (2025), and this soil property largely depends on ST and tillage practices (Abrougui et al., 2014).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eRelationship between descriptive variables and potato traits (response variables)\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eCorrelations of potato traits with all investigated descriptive variables are presented in Figure 6. In general, a much higher number of the explanatory variables significantly affected ATW and TY than DM content. According to Beukema and van der Zaag (1990), almost all factors influencing DM content also affected tuber yield. Moreover, the effect of various factors upon DM content is very complex. Namely, under certain conditions, the same factor may have a positive effect, and under other conditions (not always noticeable), it may have an adverse impact on this potato trait.\u003c/p\u003e\n\u003cp\u003eOnly seven descriptive variables significantly correlated with all investigated potato traits, with \u003cem\u003er\u0026nbsp;\u003c/em\u003evalues above 0.50 (Fig. 6). These variables represent soil physical properties (sand and silt content), and chemical properties like soil fertility in Zn and Cu, NDVI derived from UAV images taken on the 3\u003csup\u003erd\u003c/sup\u003e and 17\u003csup\u003eth\u003c/sup\u003e of August 2022. SWSaP1, determined for the period from BBCH 10\u0026ndash;19 to BBCH 39 (the time from leaf development to canopy closure \u0026ndash; tuber initiation), also significantly correlated with all potato traits.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRegarding the first model (M1), NDVI registered on 17 August 2022 (during ripening of fruit and seed), due to the high correlation with the other descriptive variables, was redundant (Fig. 8). Five of the analyzed explanatory variables were responsible for DM, AWT, and TY formation. Sand content negatively correlated with all potato traits and thus limited their development. DM strongly and positively depended on the silt content. The first two components of the RDA explained almost 87% of the variation of the dataset. Total yield and ATW were strongly positively related to sand and silt content, soil fertility in zinc and copper, NDVI measured on 3 August 2022, and SWSaP1 from BBCH 10\u0026ndash;19 to BBCH 39. However, none of the sampling areas showed a deficiency in soil copper, and only three sampling areas with high soil P content showed a deficiency in soil zinc (unpresented data). The colored points on Figure 8 indicate the field sampling areas with different ATW values. The highest DM had tubers harvested from sampling areas of A6 and A13. Conversely, the lowest DM values were determined for soil and plant sampling areas A5 and A17 (Fig. 7).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;The results of the RDA of the second model (M2) indicate the negative influence of sand content and bulk density measured at 25 cm on TY and ATW (Fig. 9). In this case, silt content also significantly and positively influenced DM, whereas bulk density at a depth of 15 cm negatively affected this potato trait. Lower soil resistance measured at 0\u0026ndash;10, 21\u0026ndash;30, 31\u0026ndash;40 cm, and 61\u0026ndash;70 cm positively influenced TY and ATW. The first two components of the RDA explained almost 85% of the variation of this dataset.\u003c/p\u003e\n\u003cp\u003eThe third model (M3) illustrates the impact of chemical properties on potato traits (Fig. 10). A higher content of sodium and copper and higher soil pH values were positively correlated with all dependent variables. On the other hand, the increase of phosphorus content in the soil negatively affected all potato traits. This could be explained by the fact that all sampling areas, taking into account also soil pH, soil was characterized by very high P content. Interestingly, Ravensbergen et al. (2023) did not find any correlation between potato yield response to P application and the amount of phosphorus applied by the farmers in the Netherlands on 46 fields. On the contrary, Leonel et al. (2017) observed that soils with higher phosphorus concentrations allowed the production of tubers with increased DM content. It should be stressed that the performance of model M3 was relatively low, because it explained only 45.5% of the variability in the dataset.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRedulla et al. (2002), found that ST components (sand, silt, and clay), impacted yield stronger than soil chemical properties.\u003c/p\u003e\n\u003cp\u003eThe strategy representing the fourth model (M4) explained almost 66% of the data set variation (Fig. 11). Higher values of such explanatory variables as NDVI registered by UAV on 22\u003csup\u003end\u003c/sup\u003e of June and 20\u003csup\u003eth\u003c/sup\u003e of July, 2022, SWSaP1, determined for the period from BBCH 10\u0026ndash;19 to BBCH 39, positively influenced all potato traits. Lynch et al. (1995) found that early (tuber initiation \u0026ndash; 2 and 4 weeks after emergence) and midseason (early tuber sizing \u0026ndash; 6 and 8 weeks after emergence) moisture stress had the most significant negative impact on tuber yield in Alberta, Canada.e The tuber initiation stage was the most appropriate for remote sensing data acquisition (Mukiibi et al. 2024). Research results for 46 potato cultivars from different maturity classes grown in Poland show that the crop was most sensitive to drought from 3 to 6 weeks after tuber initiation (Głuska 2004).\u003c/p\u003e\n\u003cp\u003eSummary of the four models\u0026rsquo; performance\u003c/p\u003e\n\u003cp\u003eThe cross-validation scheme is typically used to evaluate the final model performance, and reflected by the RMSE\u003csub\u003eCVLOO\u0026nbsp;\u003c/sub\u003eerror value (Table 2).\u0026nbsp;The results of the statistical analysis indicate that the value of the adjusted coefficient of determination (R\u003csup\u003e2\u003c/sup\u003e adj.) was highest for the first (M1) model and lowest for the third (M3) model.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2. Statistical measures of the four models and the cross-validation scheme.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"649\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\n \u003cp\u003eDescriptive variables included in the model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 58px;\"\u003e\n \u003cp\u003eAdjusted R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 166px;\"\u003e\n \u003cp\u003eRMSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 190px;\"\u003e\n \u003cp\u003eRMSE\u003csub\u003eCVLOO\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eTY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 58px;\"\u003e\n \u003cp\u003eATW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eDM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eTY\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 70px;\"\u003e\n \u003cp\u003eATW\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eDM\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eM1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\n \u003cp\u003esilt, sand, Cu, Zn, NDVI0308, SWSaP1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.832\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e5.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e7.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e8.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70px;\"\u003e\n \u003cp\u003e13.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eM2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\n \u003cp\u003eBD15, BD25, sand, silt, SR0010, SR2130, SR3140, SR6170\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.781\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e5.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e9.04\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.018\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e10.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70px;\"\u003e\n \u003cp\u003e16.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.04\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eM3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\n \u003cp\u003epH, Al, Cu, Na, P\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.268\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e10.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e13.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.701\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e14.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70px;\"\u003e\n \u003cp\u003e19.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e1.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003eM4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 169px;\"\u003e\n \u003cp\u003eNDVI2206, NDVI2007, SWSaP1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e0.595\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e6.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 58px;\"\u003e\n \u003cp\u003e9.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.680\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e7.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 70px;\"\u003e\n \u003cp\u003e10.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 54px;\"\u003e\n \u003cp\u003e0.81\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTY \u0026ndash; tuber yield; ATW \u0026ndash; average tuber weight, DM \u0026ndash; dry matter, RMSE \u0026ndash; root mean squared error in the dimension of the original variable, RMSE\u003csub\u003eCV\u003c/sub\u003e \u0026ndash; root mean squared error based on the leave-one-out cross-validation scheme.\u003c/p\u003e\n\u003cp\u003eThe results show that forecasting the development of the analyzed potato traits using the M1 model required a diverse input data set, including soil physical and chemical properties, and SWS during the canopy closure period. The effectiveness of the forecast was also influenced by the indirect crop canopy status, expressed by NDVI measured at the berry formation stage (BBCH 70\u0026ndash;79). The statistical measures, specifically the RMSE, used for estimating potato traits showed the lowest values for M1, in the case of TY and ATW. However, this error evaluated for DM was slightly lower for M2. A similar level of RMSE\u003csub\u003eCVLOO\u003c/sub\u003e was obtained in the cross-validation scheme for models M1 and M2. Interestingly, model M4, which was solely based on the indirect measure of canopy status (NDVI) and SWSaP1, was a strong alternative for forecasting TY and ATW, because the RMSE\u003csub\u003eCVLOO\u003c/sub\u003e values for the latter potato traits were the lowest for this model. Therefore, if the ranking of the models was solely based on the cross-validation results, the performance of model 4 would be ranked the highest.\u003c/p\u003e\n\u003cp\u003eTable 3. The ranking of the model performance is based on statistical measures from Table 2.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003eModels\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"7\" valign=\"top\" style=\"width: 326px;\"\u003e\n \u003cp\u003eRanking values\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eRank\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 106px;\"\u003e\n \u003cp\u003eModel order\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003eM1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e1.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 106px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003eM2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 106px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003eM3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e4.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 106px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 113px;\"\u003e\n \u003cp\u003eM4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 45px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e2.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 106px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe final performance (Table 3) of the four models was evaluated using all statistical measures presented in Table 2. Considering all the measures, the model\u0026rsquo;s performance was ranked: 1, 2, 4, and 3.\u003c/p\u003e\n\u003ch2\u003eEffect of soil water storage on potato yield\u003c/h2\u003e\n\u003cp\u003eAccording to Liao et al. (2016), soil temperature at a depth of 20 cm and SMC at 20-30 cm depth were considered significant factors that affected potato yields on commercial fields in Florida, USA. Recommended soil moisture for optimal potato growth depends on the development stage. It should be maintained at 70\u0026ndash;80% of field capacity (FC) from planting to early vine growth and 80\u0026ndash;90% FC at later growth stages, until maturation, when soil moisture should be held at 60\u0026ndash;65% FC (Rai and Dong 2025, after Pavlista 1995 and van Loon 1981). In our research, the average measured value of SWS in the 40 cm soil layer for all sampling areas was equal to 59.7 mm, which lies in the range between the SWS at field capacity and limiting point (Table 3s). This means that in the majority of the sampling areas, the average SMC was in the optimal range.\u003c/p\u003e\n\u003cp\u003eFor the comparison of SWS,\u0026nbsp;four (number 2, 4, 8, and 11) soil sampling areas were chosen based on various yields. Fig. 5 shows the results of measured SWS in the soil layer of 40 cm in the four selected sampling areas against the background of calculated values of reference evapotranspiration using equation (1) and measured precipitation values. This figure also shows calculated SWS at field capacity, limiting, and wilting points. During the potato growing season (from May 9\u003csup\u003eth\u003c/sup\u003e to September 13\u003csup\u003eth\u003c/sup\u003e), the calculated ET value was 427.7 mm, precipitation was 280 mm, and the CWB value was 147.2 mm, which confirms that, in general, potato was grown under suboptimal water supply. For the same period as given above, the positive precipitation deficit was observed only in the last decade of May and the first decade of July, but after a whole month of June with a negative value of CWB. Consequently, the most significant water deficit occurred during tuber initiation and tuber bulking stages, the most critical for potato yield formation (Thornton 2020).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn a sampling area number 2 with the lowest yield of 16.2 t\u0026middot;ha\u003csup\u003e-1\u003c/sup\u003e, the soil was characterized by the lowest water retention capacity, as shown by the measured SWS values, which fluctuated most of the growing season around 33 mm. This value was close to the SWS corresponding to the limiting point of 32 mm. Potatoes harvested from this sampling area were also characterized by the lowest AWT and number of tubers per plant (unpresented data). In a soil sampling area number 11 with a yield of 38.8 t\u0026middot;ha\u003csup\u003e-1\u003c/sup\u003e, the average SWS was higher (50.7 mm) compared to the SWS measurements done for the previous sampling area. SWS corresponding to a limiting point for this area was 45.4 mm. For soil sampling area number 4, yielding 40.7 t\u0026middot;ha\u003csup\u003e-1\u003c/sup\u003e, the average SWS was 67.2 mm, and thus was lower than the SWS at the limiting point of 72 mm.\u0026nbsp;This means that soil moisture was slightly below its optimal level. The highest yield of 67.1 t\u0026middot;ha\u003csup\u003e-1\u003c/sup\u003e was obtained at soil sampling area no 8. An average SWS in this area was 97.2 mm and did not drop below the SWS corresponding to the limiting point (77.4 mm) for most of the growing season. This sufficient water supply also produced tubers with the highest average mass.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn all of the described soil sampling areas, clear changes in SWS were observed after significant rainfall, which means that the soil moisture sensors worked correctly. It should be noted that the most substantial changes in SWS (water depletion) were observed in the period of SWSaP1 (from BBCH 10\u0026ndash;19 to BBCH 39), namely from 25 to 44 DAP.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe analysis indicates that a greater proportion of the descriptive variables significantly impacted ATW and TY than DM content. This suggests that factors influencing ATW and TY are more pronounced or varied than those affecting DM content, which may have implications for understanding the dynamics of these variables in the study. Among the four tested models describing the effect of descriptive variables on potato traits, model 1, based on sand and silt content, soil fertility in Cu and Zn, and NDVI measured by the use of UAV at the time of fruit development, and soil water storage (SWSaP1), registered for the time between leaf development and canopy closure (BBCH 10\u0026ndash;19 and BBCH 39), was ranked the highest using adjusted R\u003csup\u003e2\u003c/sup\u003e, RMSE and RMSE\u003csub\u003eCVLOO\u003c/sub\u003e. If the ranking of the models was solely based on the cross-validation results, the performance of model 4, incorporating NDVI measured twice, at the time of canopy closure and the end of flowering, and SWSaP1, would be ranked the highest.\u003c/p\u003e\u003cp\u003eThe sampling area with the lowest yield was characterized by the lowest water retention for most of the growing season, and also by an early deficit of SWS, which started close to BBCH 10\u0026ndash;19. On the contrary, in the sampling area with the highest yield, an average SWS did not drop below the SWS that corresponds to the limiting point for most of the growing season. Soil water storage determined for the time from leaf development to canopy closure, included in models 1 and 4, seems to play a more critical role in shaping tuber size and potato yield than soil chemical properties if these are kept at an optimal level.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there is no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funding was received.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank the management of the Research Farm of the Warsaw University of Life Sciences for providing the field to conduct this study and for their technical support. We also thank Przemysław Rudzki from the Yosensi Ltd. company for providing us with the IoT soil moisture sensors, and Bartłomiej Rochalski from the Research Farm of Warsaw University of Life Sciences in Żelazna for conducting part of the UAV flights.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbbas H., Ranjan R. Sri. 2015. Effect of soil moisture deficit on marketable yield and quality of potatoes. Canadian Biosystems Engineering, 57: 1.25-1.37 http://dx.doi.org/10.7451/CBE.2015.57.1.25\u003c/li\u003e\n\u003cli\u003eAbrougui, K., Chehaibi, S., Boukhalfa, H. H., Chenini, I., Douh, B., Nemri, M. 2014. Soil bulk density and potato tuber yield as influenced by tillage systems and working depths. Greener Journal of Agricultural Sciences, 4(2), 46-51.\u003c/li\u003e\n\u003cli\u003eAgricultural Drought Monitoring System (2025) https://susza.iung.pulawy.pl/en \u003c/li\u003e\n\u003cli\u003eAllen R. G., Pereira L. S., Raes D., \u0026amp; Smith M. 1998. Crop evapotranspiration-Guidelines for computing crop water requirements \u0026ndash; FAO Irrigation and Drainage Paper 56. FAO, Rome, 300(9), D05109.\u003c/li\u003e\n\u003cli\u003eAlva 2008. Setpoints for Potato Irrigation in Sandy Soils Using Real-Time, Continuous Monitoring of Soil-Water Content in Soil Profile. Journal of Crop Improvement, 21(2), 117\u0026ndash;137. https://doi.org/10.1080/15427520701885311 \u003c/li\u003e\n\u003cli\u003eAntoneli,V., Rogiski F. L., Bednarz J. A., Barrena-Gonz\u0026aacute;lez J., Bernardo, F. S., \u0026amp; Fern\u0026aacute;ndez, M. P. 2025. Assessing Soil Degradation and Hydrological Processes in Conventional Potato Crops. Air, Soil and Water Research, 18, 11786221251344463.\u003c/li\u003e\n\u003cli\u003eBeukema, H.P., van der Zaag, D.E. 1990. Introduction to potato production. Pudoc Wageningen. 208 pp. ISBN 90 220 0963 7. \u003c/li\u003e\n\u003cli\u003eBłaś, M., Ojrzyńska, H. 2024. The Climate of Poland. In: Migoń, P., Jancewicz, K. (eds) Landscapes and Landforms of Poland. World Geomorphological Landscapes. Springer, Cham. https://doi.org/10.1007/978-3-031-45762-3_3\u003c/li\u003e\n\u003cli\u003eAssouline, S., \u0026amp; Or, D. 2014. The concept of field capacity revisited: Defining intrinsic static and dynamic criteria for soil internal drainage dynamics. Water Resources Research, 50(6), 4787-4802.\u003c/li\u003e\n\u003cli\u003eChmura, K., H. Dzieżyc, Piotrowski M. 2013. Response of medium early, medium late, and late potatoes to water factor on wheat and rye soil complexes. \u003cem\u003eInfrastructure and Ecology of Rural Areas\u003c/em\u003e, 2: 103\u0026ndash;113 (in Polish).\u003c/li\u003e\n\u003cli\u003eCorwin, D.L., Lesch, S.M., 2003. Application of soil electrical conductivity to precision agriculture: theory, principles, and guidelines. Agronomy Journal, 95, 455\u0026ndash;471. https://doi.org/10.2134/agronj2003.4550\u003c/li\u003e\n\u003cli\u003eDhal, S., Wyatt, B. M., Mahanta, S., Bhattarai, N., Sharma, S., Rout, T., Saud, P., \u0026amp; Acharya, B. S. 2024. Internet of Things (IoT) in digital agriculture: An overview. Agronomy Journal, 116, 1144\u0026ndash;1163. https://doi.org/10.1002/agj2.21385\u003c/li\u003e\n\u003cli\u003eDuboisa A., Teytauda F., Verel S. 2020. Short term soil moisture forecasts for potato crop farming: a machine learning approach. Computer and Electronics in Agriculture, 180, 105902. https://doi.org/10.1016/j.compag.2020.105902 \u003c/li\u003e\n\u003cli\u003eDzwonkowski W. 2024. Rynek ziemniaka. Stan i perspektywy. 51, Analizy Rynkowe. IERiGŻ PIB. pp. 44. (In Polish).\u003c/li\u003e\n\u003cli\u003eFarooque A. A., Zare, M., Abbas, F., Bos, M., Esau, T. \u0026amp; Q. Zaman. 2019. Forecasting potato tuber yield using a soil electromagnetic induction method. European Journal of Soil Science, 1-18.\u003c/li\u003e\n\u003cli\u003eGłuska A. 2004. Reaction of potato (\u003cem\u003eSolanum tuberosum\u003c/em\u003e L.) to rainfall distribution under field conditions and testing drought imposed at different phases of bulking stage in cultivars of different vegetation length in a pot experiment. Zeszyty Problemowe Postęp\u0026oacute;w Nauk Rolniczych, 496, 217-227. (in Polish)\u003c/li\u003e\n\u003cli\u003eHaverkort A.J., Franke A.C., Steyn J.M., Pronk A.A, Caldiz D.O., Kooman P.L. 2015. A Robust Potato Model: LINTUL-POTATO-DSS. Potato Research, 58:313-327. https://doi.org/10.1007/s11540-015-9303-7\u003c/li\u003e\n\u003cli\u003eInstitute of Meteorology and Water Management. National Research Institute. Climate normal 1991\u0026ndash;2020. https://klimat.imgw.pl/pl/climate-normals \u003c/li\u003e\n\u003cli\u003eJames G., Witten D., Hastie T., Tibshirani R. 2013. An introduction to statistical learning (Vol. 112, No. 1). New York: Springer (177-199).\u003c/li\u003e\n\u003cli\u003eKęsik K. 2016. Zastosowanie metody Mehlich 3 w systemie doradztwa nawozowego. Studia i raporty IUNG-PIB, ZESZYT 48(2): 95-104, doi: 10.26114/sir.iung.2016.48.07 (In Polish).\u003c/li\u003e\n\u003cli\u003eKleinschmidt G.D., Kleinkopf G.E., Walz A.J., Zalewski. J.C., 1983. Specific gravity of potatoes. Current Information Series No. 609/ University of Idaho. College of Agriculture.\u003c/li\u003e\n\u003cli\u003eKorzeniowska J., Stanisławska-Glubiak E., Jadczyszyn T., Lipiński W. 2021. Nawożenie upraw rolniczych mikroelementami. Instrukcja upowszechnieniowa nr 249. Wydawnictwo IUNG-PIB, pp. 31. (In Polish). \u003c/li\u003e\n\u003cli\u003eLegendre P, Legendre L. Redundancy analysis. 2012. In: Numerical Ecology. Elsevier, editor. 3\u003csup\u003erd\u003c/sup\u003e ed. Elsevier Science B.V; 629\u0026ndash;658\u003c/li\u003e\n\u003cli\u003eLegendre P., \u0026amp; Legendre L. 2012. Canonical analysis. In Developments in environmental modelling (Vol. 24, pp. 625-710). Elsevier.\u003c/li\u003e\n\u003cli\u003eLeonel M, do Carmo EL, Fernandes AM, Soratto RP, Eb\u0026uacute;rneo JAM, Garcia \u0026Eacute;L, dos Santos TPR. 2017. Chemical composition of potato tubers: the effect of cultivars and growth conditions. Journal of Food Science and Technology, 54, 2372\u0026ndash;2378. https://doi.org/10.1007/s13197-017-2677-6 \u003c/li\u003e\n\u003cli\u003eLiao X., Su Z., Liu G., Zotarelli L., Cui Y., Snodgrass C. 2016. Impact of soil moisture and temperature on potato production using seepage and center pivot irrigation. Agricultural Water Management, 165, 230-236. http://dx.doi.org/10.1016/j.agwat.2015.10.023\u003c/li\u003e\n\u003cli\u003eLynch D. R., Foroud N., Kozub G. C., Farries B. C. 1995. The effect of moisture stress at three growth stages on the yield, components of yield, and processing quality of eight potato varieties. American Potato Journal, 72, 375-385. https://link.springer.com/article/10.1007/BF02849334\u003c/li\u003e\n\u003cli\u003eŁopatka, A., 2017. Europejski monitoring użytkowania grunt\u0026oacute;w i baza danych projektu LUCAS. Studia i Raporty IUNG-PIB 51 (5), 73\u0026ndash;89. (in Polish).\u003c/li\u003e\n\u003cli\u003eMatteau, J. P., C\u0026eacute;licourt, P., L\u0026eacute;tourneau, G., Gumiere, T., \u0026amp; Gumiere, S. J. 2022. Effects of irrigation thresholds and temporal distribution on potato yield and water productivity in sandy soil. Agricultural Water Management, 264, 107483.\u003c/li\u003e\n\u003cli\u003eMeier, U. 2018. Growth stages of mono- and dicotyledonous plants. BBCH Monograph. http://doi/10.5073/20180906-07461 \u003c/li\u003e\n\u003cli\u003eMoebius-Clune B.N., Moebius-Clune D.J., Gugino B.K., Idowu O.J., Schindelbeck R.R., Ristow A.J., van Es H.M., Thies J.E., Shayler H.A., McBride M.B., Kurtz K.S.M., Wolfe D.W., and Abawi G.S. 2017. Comprehensive Assessment of Soil Health \u0026ndash; The Cornell Framework. Edition 3.2, Cornell University, Geneva, NY.\u003c/li\u003e\n\u003cli\u003eMukiibi A., Machakaire A.T.B., Franke A.C., Steyn J.M. 2024. A Systematic Review of Vegetation Indices for Potato Growth Monitoring and Tuber Yield Prediction from Remote Sensing. Potato Research https://doi.org/10.1007/s11540-024-09748-7\u003c/li\u003e\n\u003cli\u003eMusierowicz, A. 1950. Granulometric composition of soils and methods of analysis of granulometric composition. (in Polish). Warszawa, 1949. PIWR.\u003c/li\u003e\n\u003cli\u003eOksanen, J.,Simposon G.L, Blanchet, F.G., Kindt, R., Legendre, P., Minchin R., O\u0026rsquo;Hara R.B., et al., 2025. Vegan Community Ecology Package Version 2.6-10 January 2025.\u003c/li\u003e\n\u003cli\u003ePan, L.; Adamchuk, V.I.; Martin D.L.; Schroeder M.A. \u0026amp; Ferguson, R.B. 2010. Analysis of water use efficiency using on-the-go soil sensing and a wireless network. In: Proceedings of the Tenth International Conference on Precision Agriculture, Denver, Colorado, 18-21 July 2010, ed. R. Kholsa. Colorado State University, Fort Collins, Colorado, USA (CD publication, 13 pages). \u003c/li\u003e\n\u003cli\u003ePo E., Snapp S.S, \u0026amp; A. Kravchenko. 2010. Potato Yield Variability across the Landscape. Agronomy Journal, 102(3), 885-894. doi:10.2134/agronj2009.0424.\u003c/li\u003e\n\u003cli\u003eRai A., Dong Y. 2025. A scoping review of irrigation scheduling methods in potato (\u003cem\u003eSolanum tuberosum\u003c/em\u003e L.) production. American Journal of Potato Research https://doi.org/10.1007/s12230-025-09997-3 \u003c/li\u003e\n\u003cli\u003eRavensbergen A. P. P., van Ittersum M. K., Kempenaar C., Reidsma P. 2023. Current Phosphorus and Potassium Fertiliser Application Rates Do Not Limit Tuber Yield and Quality in Potato Production Systems in the Netherlands. Potato Research, 66:1033\u0026ndash;1058 https://doi.org/10.1007/s11540-022-09613-5\u003c/li\u003e\n\u003cli\u003eRedulla, C.A., Davenport, J. R., Evans, R. G., Hattendorf, M. J., Alva, A.K., Boydston, R. A. 2002. Relating Potato Yield and Quality to Field Scale Variability in Soil Characteristics. American Journal of Potato, 79:317\u0026ndash;323. \u003c/li\u003e\n\u003cli\u003eRouse, J. W., Haas, R. H., Schell, J. A., Deering, D. W. 1973. Monitoring vegetation systems in the Great Plains with ERTS. In Third ERTS Symposium, DC. 10-14. NASASP-351, MD, pp 309\u0026ndash;317.\u003c/li\u003e\n\u003cli\u003eSheng, Y., He, P., Xu X., Liu Y. 2023. A large-scale assessment on spatial variability of potato yield and soil chemical properties in northern China. Soil and Tillage Research, 231. https://doi.org/10.1016/j.still.2023.105743 \u003c/li\u003e\n\u003cli\u003eRStudio Team (2020). RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL http://www.rstudio.com/.\u003c/li\u003e\n\u003cli\u003eSiano, A. B., Roskruge, N., Kerckhoffs, H., Sofkova-Bobcheva, S. 2024. Effects of Abiotic Stress Associated with Climate Change on Potato Yield and Tuber Quality Under a Multi‑environment Trial in New Zealand. Potato Research, 67: 1603-1624.\u003c/li\u003e\n\u003cli\u003eSikora, F. J., Moore, K. P. 2014. Soil test methods from the southeastern United States. Southern Cooperative Series Bulletin 419. Clemson, SC, USA: Clemson University.\u003c/li\u003e\n\u003cli\u003eThornton, M., 2020. Potato Growth and Development, in: Potato Production Systems. Springer International Publishing, Cham, pp. 19\u0026ndash;33. https://doi.org/10.1007/978-3-030-39157-7_2 \u003c/li\u003e\n\u003cli\u003eUtset, A., Ruiz, M. E., Garcia, J., \u0026amp; Feddes, R. A. 2000. A SWACROP-based potato root water-uptake function as determined under tropical conditions. Potato Research, 43, 19-29.\u003c/li\u003e\n\u003cli\u003eVan Genuchten, M. T. 1980. A closed‐form equation for predicting the hydraulic conductivity of unsaturated soils. Soil science society of America journal, 44(5), 892-898.\u003c/li\u003e\n\u003cli\u003eWhelan, B. M., Mulcahy, F., 2017. A strategy to instigate SSCM in Australian potato production. In J A Taylor, D Cammarano, A Prashar, A Hamilton (Eds.) Proceedings of the 11th European Conference on Precision Agriculture. \u003cem\u003eAdvances in Animal Biosciences\u003c/em\u003e, 8 (2), 743-748. doi:10.1017/S2040470017000401\u003c/li\u003e\n\u003cli\u003eWickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org.\u003c/li\u003e\n\u003cli\u003eWierzbicka A. 2014. Zmiany klimatycznego bilansu wodnego w okresie wegetacji ziemniaka w regionie P\u0026oacute;łnocnego Mazowsza w latach 1973\u0026ndash;2012. Zeszyty Problemowe Postęp\u0026oacute;w Nauk Rolniczych, 576, 207\u0026ndash;215.\u003c/li\u003e\n\u003cli\u003eW\u0026ouml;sten, J. H. M., Lilly, A., Nemes, A., \u0026amp; Le Bas, C. 1999. Development and use of a database of hydraulic properties of European soils. Geoderma, 90 (3-4), 169-185.\u003c/li\u003e\n\u003cli\u003eZarzyńska K., Boguszewska-Mańkowska D., Nosalewicz A. 2017. Differences in size and architecture of the potato cultivars root system and their tolerance to drought stress. Plant, Soil and Environment, 63, 4, 159-164. \u003c/li\u003e\n\u003cli\u003eZebarth B.J., Cambouris A.N., Perron I., Burton D.L., Comeau L.P., \u0026amp; G. Moreau. 2019. Spatial Variation of Soil Health Indices in a Commercial Potato Field in Eastern Canada. Soil Science Society of America Journal, 83, 1786-1798.\u003c/li\u003e\n\u003cli\u003eZebarth B.J., Fillmore S., Watts S., Barrett R., \u0026amp; L.P. Comeau. 2021. Soil Factors Related to within-Field Yield Variation in Commercial Potato Fields in Prince Edward Island, Canada. American Journal of Potato Research, https://doi.org/10.1007/s12230-021-09825-4.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"potato, IoT, soil moisture sensors, soil properties, UAV, NDVI","lastPublishedDoi":"10.21203/rs.3.rs-7613486/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7613486/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMany soil-related factors associated with potato yield and quality are implicitly linked to soil water availability (SWA). This soil property also affects potato canopy, which could be indirectly evaluated using Unmanned Aerial Vehicles (UAVs). The objective of this study was to quantify the optimal explanatory soil- and canopy state-related properties that determine potato (\u003cem\u003eSolanum tuberosum\u003c/em\u003e L.) traits: total yield (TY), dry matter (DM) content, and average tuber weight (ATW) over a rainfed production field. SWA was continuously measured using moisture sensors installed at three depths and connected via the Internet to exchange data. Soil water storage (SWS) was derived from soil moisture content (SMC). Thus, the effect of spatial and temporal changes in SWS on TY was also determined.\u003c/p\u003e\u003cp\u003ePotatoes were grown under a suboptimal water supply during crucial stages for yield formation. More explanatory variables significantly affected ATW and TY than DM content. SWS, determined between the stages of leaf development and canopy closure, played a more critical role in forming ATW and TY than soil chemical properties. A model solely based on canopy status evaluation between the time of canopy closure and the end of flowering, and SWS determined simultaneously, was a strong alternative for forecasting TY and ATW. In the sampling area with the highest TY, the average SWS did not drop below the level that limits potato growth for most of the growing season. A sampling area with the lowest TY was characterized by the lowest water retention for most of the growing season.\u003c/p\u003e","manuscriptTitle":"How do soil properties and canopy status influence potato traits while continuously monitoring soil moisture with IoT-based sensors?","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-02 09:52:23","doi":"10.21203/rs.3.rs-7613486/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"fe55ca73-c693-4f55-9468-3d305f349435","owner":[],"postedDate":"October 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-12-09T00:53:20+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-02 09:52:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7613486","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7613486","identity":"rs-7613486","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}