Comprehensive Analysis of Soil Physicochemical Properties and Optimization Strategies for “Yantai Fuji 3” Apple Orchards

Abstract

Based on an integrated analysis, this study summarized the current status of soil quality in Yantai apple orchards, developed a multivariate regulation model for key soil physicochemical properties, and proposed optimized fertilization strategies to improve soil quality in the region. The study analyzed the physicochemical properties of the topsoil (0–30 cm) in 19 representative apple orchards across Yantai, including indicators like pH, organic matter (OM), major nutrient ions, and salinity indicators, using standardized measurements and multivariate statistical methods, including descriptive statistics analysis, frequency distribution analysis, canonical correlation analysis, stepwise regression equation analysis, and regression fit model analysis. The results demonstrated that in apple orchards across the Yantai region, reductions in pH were significantly mitigated under the combined increased OM and exchangeable calcium (Ca). Exchangeable potassium (EK) rose in response to the joint elevation of OM and available nitrogen (AN), and AN was also positively influenced by EK, while OM also exhibited a promotive effect on Olsen phosphorus (OP). Furthermore, Ca increased with higher pH. AN and EK jointly contributed to the increases in electrical conductivity (EC) and chloride ions (Cl), while elevated exchangeable sodium (Na) and soluble salts (SS) were primarily driven by EK. Accordingly, enhancing organic and calcium source fertilizers is recommended to boost OM and Ca levels, reduce acidification, and maintain EC within optimal limits. By primarily reducing potassium’s application, followed by nitrogen and phosphorus source fertilizers, the supply of macronutrients can be optimized, and the accumulation of Na, Cl, and SS can be controlled. Collectively, the combined analysis of soil quality status and the multivariate regulation model clarified the optimized fertilization strategies, thereby establishing a solid theoretical and practical foundation for recognizing the necessity of soil testing and formula fertilization, the urgency of improving soil quality, and the scientific rationale for nutrient input management in Yantai apple orchards.

1. Introduction

China is a global leader in both apple cultivation and consumption [1]. Yantai apples are recognized for their unique “crisp and refreshing” texture. In 2022, the brand value of Yantai apples reached RMB 15.837 billion, securing its status as China’s top-ranked fruit brand for 15 consecutive years [2]. However, many orchards depend on empirical fertilization practices, resulting in imbalanced fertilizer structures and decreased economic and environmental efficiency [3,4]. Given the challenges of soil acidification, salinization, compaction, and fertility degradation, coupled with unclear key controlling factors [5], statistical and multivariate analyses of orchard soil physicochemical properties [6], along with fertilization structure optimization [7], can help enhance land resource utilization, improve soil quality, and increase fertilizer efficiency [8]. This approach boosts apple yield and quality [9], supporting safe production and sustainable ecological development [10].

Soil is a critical component in the growth and development of fruit trees [11,12]. Evaluating the abundance–deficiency index of soil nutrients is crucial for evaluating soil quality and developing soil testing and formula fertilization strategies [13,14]. The evaluation system for soil productivity encompasses 115 indicators, such as pH, OM (organic matter), AN (available nitrogen), OP (Olsen phosphorus), EK (exchangeable potassium), Ca (exchangeable calcium), EC (electrical conductivity), Na (exchangeable sodium), SS (soluble salts), and Cl (chloride ions), all of which are critical for assessing soil productivity [15,16]. Macronutrients (N, P, K) in apple orchard soils are crucial for evaluating soil quality and determining fertilization strategies [17,18]. Research has highlighted that calcium deficiency can cause significant problems, including reduced yield and a bitter pit [19,20]. Optimal levels of EC, Na, and SS are vital for maintaining soil structure stability, hydrological properties, and fertility [21,22,23,24,25,26]. Elevated Cl levels may reflect the quantity and quality of agricultural inputs, such as fertilizers, pesticides, and water sources [27,28]. In Yantai apple orchards, over-application and improper use of fertilizers, combined with the natural susceptibility of coastal and mountainous areas to salinization, may lead to primary or secondary salinization. Although these indicators have often been overlooked in traditional soil quality models, they are increasingly recognized for their diagnostic value and practical application potential in orchard ecosystems. All of these indicators have the advantages of simple testing equipment, easy data acquisition, and convenient field measurement, with great potential for widespread application [22], and they offer significant prospects for improving nutrient management [29].

Research on soil fertility in apple orchards primarily focuses on the interconnected framework of changes in yield, quality, soil physicochemical properties, chemical fertilizer reduction, and substitution with organic fertilizers [3]. Statistical methods, including correlation analysis, principal component analysis, and regression analysis, have been employed to evaluate soil quality, examining the interactions among soil fertility, fruit yield and quality, and post-harvest storage [16,30]. Studies by Kai et al. [3], Zhuang et al. [4], Zhang et al. [31,32], and Mota et al. [33] suggest that integrating organic fertilizers with reduced chemical fertilizers and trace elements, such as iron, zinc, and boron, can significantly enhance fruit tree yield and quality. Furthermore, research by Zheng et al. revealed low levels of OM and AN in apple orchard soils in Shaanxi Province [34], whereas Zhang et al. identified total nitrogen deficiencies in orchards on the Loess Plateau [35]. Cao et al. demonstrated a significant correlation between citrus mineral nutrition and soil nutrients in Zhejiang Province [36]. Yildiz et al. found significant relationships between the soil physicochemical properties and the nutrient content of plant leaves in apple orchards [37]. Ganai et al. found that pH exhibited a significant and positive correlation with Ca, while OM showed a positive and significant correlation with AN OP and EK in apple orchards [38]. Nevertheless, systematic research on the evaluation, comprehensive analysis, and optimization of soil physicochemical properties in Yantai apple orchards remains limited; this study fills that gap by incorporating novel indicators (Na, SS, Cl⁻, EC) and constructing a rigorous multivariate analytical framework to support precision soil management.

This study investigates Yantai apple orchards, utilizing standardized measurement methods to analyze soil physicochemical properties, such as pH, OM, AN, OP, EK, Ca, Na, EC, SS, and Cl. Descriptive statistics analysis, frequency distribution, and curve fitting analysis were used to assess multivariate differences among the orchards, followed by a comprehensive evaluation of soil quality standards based on the soil quality classification standard for apple orchards. Correlations between the physicochemical properties were identified (canonical correlation analysis), and regression fitting analysis was conducted to identify key soil properties that significantly influence soil quality (establishment of the regression equation and the regression fitting model). We constructed a logically coherent analytical framework that includes canonical correlation analysis, stepwise regression, regression fit models, and a multivariate regulation model of soil physicochemical properties, and their associations with orchard input indicators facilitated the recognition of critical inputs. Based on this analysis, strategies for optimizing soil quality in Yantai apple orchards were explored, promoting the simultaneous enhancement of economic and environmental benefits and the sustainable growth of the contemporary apple industry.

2. Materials and Methods

2.1. Study Area

Yantai is situated in the northeastern part of the Shandong Peninsula and features a temperate continental climate. It is influenced by the East Asian monsoon, with an annual average temperature of 13 °C, annual precipitation of 722.2 mm, and a frost-free period lasting 210 days. Soil samples were collected from 19 representative “Yantai Fuji 3” apple orchards during their 7th to 9th production years. These orchards are characterized by brown earth soils with a loamy texture and planting densities ranging from 750 to 1200 plants per hectare. The sampling locations included Dayao Street (Muping District); Qishan Town and Fushan Town (Zhaoyuan City); Panshidian Town and Fangyuan Street (Haiyang City); Chengxiang Street and Longwangzhuang Street (Laiyang City); Zhuqiao Town (Laizhou City); Liujiagou Town, Nanwang Street, Xiaomengjia Town, and Heishan Township (Penglai District); Qijia Town, Zhuyouguan Town, and Dongjiang Street (Longkou City); Tangjiabo Town (Qixia City); Zangjiazhuang Town and Gaotuan Town (Fushan District); and Chaoshui Town (Development Zone) (Figure 1).

2.2. Soil Sample Collection

In July 2024, during the early fruit expansion stage, soil samples were collected using an “S”-shaped grid sampling method. Ten grid points were established along the “S”-shaped path within each orchard, and one apple tree was selected at the center of each sampling point. Using professional topsoil samplers (SQ-802C, Global Clearing (Beijing) Technology Co., Ltd., Beijing, China), soil from a depth of 0–30 cm was collected at the drip line of each selected tree. A total of 10 representative apple trees were sampled per orchard, ensuring scientific rigor, systematic spatial distribution, and objectivity in the sampling process. Impurities, such as roots and gravel, were removed, and the quartering method was applied by discarding two opposite quarters and repeating the process until approximately 1 kg of soil was collected, with relevant information marked before returning to the laboratory.

2.3. Soil Physicochemical Property Determination

Air-dried soil samples were passed through a 2 mm sieve for further use. A 5 g subsample was taken and mixed with deionized water at a 2.5:1 soil-to-water ratio, stirred for 5 min, and left to stand for 1 h before filtration, and the pH was measured using a pH meter (PH610, Beijing Wiggens Technology Co., Ltd., Beijing, China) [39]. Soil EC was determined using the electrode method with a 5:1 soil-to-water ratio (DDSJ-308F, Shanghai INASE Scientific Instrument Co., Ltd., Shanghai, China). The suspension was shaken for 30 min, allowed to settle for another 30 min, and then filtered prior to measurement [40]. Additionally, 2 g of samples was analyzed for AN using the alkali-diffusion method, with boric acid (CAS: 10043-35-3, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) as the absorbent solution titrated by sodium hydroxide (CAS: 1310-73-2, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) from an alkali burette [41]. Soil OP was extracted using 0.5 mol/L NaHCO₃ (pH 8.5) (CAS: 144-55-8, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China), shaken for 30 min, and filtered immediately, followed by determination using the molybdenum blue colorimetric method (UV-1100B, Shanghai ZiQi Laboratory Equipment Co., Ltd., Shanghai, China) [22,42]. Soil EK, Ca, and Na were measured using atomic absorption spectrophotometry with a 1 mol/L CH₃COONH₄ extractant (CAS: 631-61-8, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) at pH 7 (TAS-990, Beijing PERSEE General Instrument Co., Ltd., Beijing, China). K and Na extracts were obtained after 30 min of shaking and immediate filtration. For Ca, samples were stirred and centrifuged repeatedly until no calcium ion reaction was detected, followed by filtration and volume adjustment [22,43]. Soil Cl was analyzed from a 5 g sample using the silver nitrate titration method with a 5:1 soil-to-water ratio. Samples were shaken for 30 min, filtered under vacuum, and titrated with a standard AgNO₃ solution (CAS: 7697-37-2, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) [44]. Soil SS was determined using the residue-drying method with a 5:1 soil-to-water ratio. The suspension was shaken for 30 min and filtered under vacuum, and the filtrate was evaporated to dryness [45,46]. Soil OM was analyzed from a 0.2 g sample; to analyze samples in the analyzer, you should ground them smaller than 0.15 mm using the elemental analyzer method (vario MAX cube, Elementar Trading (Shanghai) Co., Ltd., Shanghai, China) [47]. Detailed procedures for instrument operation followed the respective manufacturers’ manuals.

2.4. Soil Quality Classification Standard for Apple Orchards

The second national soil survey conducted in China [48], particularly in Shandong Province [49], serves as a key reference for classifying the physicochemical properties of orchard soils. To account for changes in agricultural practices over time, our study incorporated historical analyses of Yantai orchard soils from various periods, emphasizing regional variations in soil properties [49,50,51]. For the apple orchard soils in Yantai, we relied on our research team’s significant findings and extensive laboratory analysis experience [52,53,54,55,56]. To enhance the quality of apple orchard soils, we developed the “Yantai Apple Orchard Soil Physicochemical Classification Standard” (

Table 1. Classification standard of soil quality indexes of Yantai apple orchards.

Index Description *	OM g/kg	AN mg/kg	OP mg/kg	EK mg/kg	Ca mg/kg	EC µs/cm	Index Description	Na mg/kg	SS g/kg	Cl mg/kg
Lowest	<10	<60	<10	<100	<1000	<120	Low	<120	<3	<180
Low	10–20	60–90	10–20	100–150	1000–2000	120–350
Middle	20–30	90–120	20–30	150–200	2000–3000	350–1000
High	30–40	120–150	30–40	200–250	3000–4000	1000–2000	High	≥120	≥3	≥180
Highest	≥40	≥150	≥40	≥250	≥4000	≥2000

* Strong acid: <4.5; acidity: 4.5–5.5; weak acid: 5.5–6.5; neutrality: 6.5–7.5; weak alkali: 7.5–8.5; alkalinity: 8.5–9.5; strong alkali: >9.5. Note: “<” excludes the value, “≥” includes the value, and “−” contains the value of the latter only.

). Specifically, soil pH was categorized into seven levels according to Liu et al. [57]. Soil OM, AN, OP, and EK were divided into five levels according to studies by Zhang et al., Zheng et al., and Li et al. [58,59,60,61]. Soil Ca and EC were classified into five levels based on Zhang et al. and Luo et al. [60,62]. Soil SS, Cl, and Na were grouped into two levels according to research by Jiang et al., Bao et al., He et al., Ma et al., Cheng et al., and local standards [63,64,65,66,67,68]. Specific physicochemical indicators, such as AN, OP, EK, and EC, should not be maintained at excessively high levels. Similarly, elevated levels of Na, SS, and Cl are undesirable.

2.5. Data Processing

Standardized data collection methods were applied to ensure data reliability and accuracy. Multivariate statistical analyses comprised frequency distribution analysis, nonlinear frequency fitting, canonical correlation analysis, root mean squared error (RMSE), Akaike Information Criterion (AIC), AIC corrected form (AICc), and establishment of the regression equation and regression fit model. Data processing, statistical analyses, and graph plotting were conducted using OriginPro 2024 (version number: 10.1.0.178) and Microsoft Office Professional Plus 2013 (version number: 15.0.4420.1017), all on a 64-bit Windows platform. Specifically, for the frequency distribution analysis, column plots were created based on the “Bin Center” and “Relative Frequency” columns. For nonlinear frequency fitting analysis, the peak function option was set to “Gauss”. Canonical correlation analysis used Pearson correlation coefficients to assess linear relationships between variables. In the establishment of the regression equation, the “Parameters” and “ANOVA” table data were used for outputting “Regression Equations.” The univariate 2D linear regression fitting model was developed by creating scatter plots for different physicochemical properties and filling in the corresponding “Confidence Bands” The bivariate 3D plane regression fitting model selects “Nonlinear Surface Fit” in the analysis and then manually inputs the corresponding bivariate linear equation in the “Function Body” section before proceeding with the fitting process.

3. Results

3.1. Descriptive Statistics Analysis of Soil Physicochemical Properties in Yantai Apple Orchards

The pH, OM, AN, OP, EK, Ca, EC, SS, and Cl in the topsoil were measured, and descriptive statistical analysis was conducted (

Table 2. Descriptive statistics for orchard soil physicochemical properties in Yantai.

Index *	pH	OM g/kg	AN mg/kg	OP mg/kg	EK mg/kg	Ca mg/kg	EC µs/cm	Na mg/kg	SS g/kg	Cl mg/kg
Mean ± SE	6.45 ±0.21	22.42 ±3.36	142.21 ±21.25	106.91 ±17.81	371.78 ±63.37	1650.63 ±202.09	550.22 ±137.83	204.55 ±19.83	2.94 ±0.3	22.5 ±7.72
CV %	14.04	65.40	65.14	72.63	74.30	53.37	109.19	42.26	44.32	149.64
AV	4.73–7.69	3.81–50.19	20.98– 312.46	31.05–284.84	41.41–844.59	376.68– 3465.70	59.6–2260	83.16– 421.30	0.83–5.1	7.13– 142.37

* Abbreviations: pH (–log₁₀[H⁺], where [H⁺] represents soil hydrogen ion activity), OM (soil organic matter), AN (soil available nitrogen), OP (soil Olsen phosphorus), EK (soil exchangeable potassium), Ca (soil exchangeable calcium), EC (soil electrical conductivity), Na (soil exchangeable sodium), SS (soil soluble salts), Cl (soil chloride ions). These abbreviations are consistently used throughout all figures and tables in this study. This row of data in the table presents the average value of 19 observations (mean), along with the standard error (SE). The second row presents the coefficient of variation (CV) of 19 observations. The third row presents the amplitude of variation (AV) of 19 observations.

). A quality assessment based on the mean values of each indicator showed that the overall soil pH was slightly acidic, OM was at a moderate to low level, AN was relatively high, OP and EK were at the highest levels, Ca was low, EC was moderate, Na was relatively high, SS was at a moderately low level, and Cl was low in Yantai apple orchards. The coefficient of variation (CV) for soil physicochemical properties in the Yantai region was found to be relatively high (ranging from 14.0% to 149.64%), and the variability ranking of the ten indicators was as follows: pH < Na < SS < Ca < AN < OM < OP < EK < EC < Cl. Based on the standard error (SE) and amplitude of variation (AV), a broad distribution of soil quality grades was observed, and the soil quality in some orchards was relatively low. This condition was characterized by acidic pH, elevated levels of Na, SS, and Cl, and insufficient levels of AN, OP, EK, OM, and Ca. The importance of promoting awareness of soil testing and formula fertilization in Yantai apple orchards to minimize differences in soil quality was underscored by these findings.

3.2. Frequency Distribution Analysis of Soil Quality Classification Indexes in the Yantai Apple Orchard Region

Based on the classification standards (Table 1), the soil quality distribution in apple orchards in Yantai was analyzed using frequency distribution and curve fitting methods (Figure 2). It was shown that pH was most frequently distributed in the neutrality, acidity, and weak acid categories, with a cumulative frequency of 89.5% of the samples (Figure 2a). Similarly, Ca was mainly concentrated in the low, middle, and lowest categories, representing 94.7% (Figure 2b), and EC was primarily distributed across the middle, low, and lowest categories, with a cumulative frequency of 89.4% (Figure 2c). In contrast, the highest frequencies for AN and EK were observed at both the lowest and the highest levels, with AN at 68.5% and EK at 84.1% (Figure 2d,f). Notably, convergence of the nonlinear fitting for EK was not achieved after three iterations. OP was entirely distributed within the high and highest levels (100%). Similarly, convergence of its curve fitting was also not achieved after three iterations (Figure 2e). A bimodal distribution was exhibited by OM, with the lowest frequency at the highest level and the highest frequencies at the lowest and high levels (52.6%) (Figure 2g). Na was predominantly found in the high category (Figure 2h), whereas SS and Cl were mostly concentrated in the low category (Figure 2i,j). In summary, in the apple orchards of Yantai, the frequency distributions of pH, Ca, and EC followed unimodal curve patterns; AN and EK followed single-valley curve fittings; OP presented an increasing trend; and OM presented a bimodal pattern. The soil in these orchards was found to be slightly acidic, with relatively low levels of OM, Ca, and EC and relatively high but unevenly distributed levels of AN, OP, and EK. Additionally, negative soil quality indicators—Na and SS—were present at excessively high levels, which might have posed adverse effects. It was indicated by these results that the frequency distribution characteristics of different soil physicochemical property grades varied in the apple orchards of Yantai, as reflected in the presence of values across various grades with uneven distribution. The broad distribution patterns of soil quality and the derived results were found to be consistent with the previous descriptive statistics analysis.

3.3. Canonical Correlation Analysis of Soil Physicochemical Properties in Yantai Apple Orchards

Canonical correlation analysis was performed to investigate the correlative relationships between pairs of soil physicochemical variables in Yantai orchards, focusing on those with significant or highly significant correlations (Figure 3). It was indicated by the results that Ca was significantly positively correlated only with pH (p < 0.05), whereas pH was significantly positively correlated with OM (p < 0.05). Significant positive correlations of soil OM with Na and Cl (p < 0.05), and highly significant correlations with OP, EK, EC, and SS (p < 0.01), were observed. Among these variables, significant positive correlations of EK with AN and Na (p < 0.05), and highly significant correlations with EC, SS, and Cl (p < 0.01), were observed. Similarly, highly significant positive correlations of SS and AN with Cl (p < 0.01) were observed. Additionally, highly significant positive correlations of SS and AN with Cl (p < 0.01) were observed. Overall, numerous significant or highly significant positive correlations, along with a few negative ones, were displayed among soil physicochemical properties in Yantai apple orchards. It was suggested by these findings that multiple significant or highly significant synergistic and antagonistic interactions existed among the variables. However, such complex relationships, particularly those involving more than two sets of variables, could not be fully explained by canonical correlation analysis alone. Therefore, canonical correlation analysis could be used as a foundation for identifying significantly related sets of independent variables. These variables could then be utilized in subsequent multivariate regression analyses and model fitting, thereby laying the groundwork for a more in-depth exploration of the complex interactions among soil physicochemical properties.

3.4. Establishment of the Stepwise Regression Equation for Soil Physicochemical Properties in the Yantai Apple Orchard Region

The complex multivariate correlations among soil physicochemical properties were explored, building on the assessment of current soil quality in Yantai apple orchards and addressing the limitations of canonical correlation analysis. All regression equations were statistically significant (p < 0.05) (

Table 3. Establishment of regression equation affecting soil physicochemical properties in Yantai apple orchards.

Soil Qualities (y)	Soil Physicochemical Properties (x)	Regression Equations 1	F 2	RMSE 3	AIC 4	AICc 5
pH	OM, Ca	y = 0.02445x2 + 0.0004366x6 + 5.18536	5.864 *	0.847236	53.62019	55.22019
	OM	y = 0.03118x2 + 5.75525	5.806 *	1.010764	58.3265	59.0765
	Ca	y = 0.000541x6 + 5.56124	6.504 *	0.962426	56.46435	57.21435
OM	pH, OP, EK, EC, Na, SS, Cl	y = 3.10842x1 + 0.04746x4 + 0.02802x5 + 0.00119x7 − 0.00772x8 + 1.49338x9 − 0.01977x10 − 16.15609	10.026 ***	54.71341	221.9998	236.3998
	pH, OP, EK	y = 2.96924x1 + 0.05151x4 + 0.0317x5 − 14.03545	29.821 ****	56.66747	215.3333	218.1904
	OP, EK	y = 0.05486x4 + 0.0346x5 + 3.69217	38.425 ****	54.22059	211.656	213.256
	pH, EK	y = 3.11539x1 + 0.04369x5 − 13.93099	40.393 ****	66.20909	219.2467	220.8467
	pH, OP	y = 3.92747x1 + 0.14244x4 − 18.15811	24.661 ****	89.71605	230.7924	232.3924
	pH	y = 8.16583x1 − 30.28457	5.806 *	286.5772	272.924	273.674
	OP	y = 0.15835x4 + 5.48982	40.290 ****	88.7981	228.4015	229.1515
	EK	y = 0.04756x5 + 4.73612	69.276 ****	59.14407	212.9588	213.7088
AN	EK, EC, Cl	y = −0.06334x3 + 0.16938x7 − 0.70903x10 + 88.51688	6.123 **	5784.364	391.1104	393.9675
	EK	y = 0.16535x5 + 80.73892	5.460 *	12,268.21	415.6808	416.4308
OP	OM, EK, EC, SS	y = 2.15557x2 + 0.16751x5 − 0.02201x7 + 0.85025x9 + 5.91856	10.789 ***	3167.341	370.2243	374.8397
	OM, EK	y = 2.28702x2 + 0.12761x5 + 8.19773	23.234 ***	3123.858	365.699	367.299
	OM	y = 4.44127x2 + 7.34313	40.290 ****	2897.308	360.8381	361.5881
	EK	y = 0.23639x5 + 19.02934	41.045 ***	4086.369	373.9053	374.6553
EK	OM, AN, OP, EC, Na, SS, Cl	y = 7.02255x2 + 0.51911x3 + 1.73416x4 − 0.13412x7 − 0.13803x8 + 18.54982x9 + 3.43519x10 − 74.69012	15.424 ****	12,337.42	427.8946	442.2946
	OM, AN, OP	y = 10.23906x2 + 0.63917x3 + 1.17234x4 − 74.01117	33.409 ****	16,399.39	430.7096	433.5668
	AN, OP	y = 0.87921x3 + 2.71944x4 − 43.99942	29.774 ****	28,719.76	450.0026	451.6026
	OM, OP	y = 12.11929x2 + 1.0723x4 − 14.57107	39.037 ****	22,969.26	441.5123	443.1123
	OM, AN	y = 15.54414x2 + 0.59538x3 − 61.38013	41.317 ****	17,524.59	431.2314	432.8314
	OM	y = 16.88168x2 − 6.69701	69.276 ****	23,047.28	439.6412	440.3912
	AN	y = 1.47021x3 + 162.69976	5.460 *	110,188.5	499.0977	499.8477
	OP	y = 2.99137x4 + 51.96166	41.045 ****	43,038.65	463.3741	464.1241
Ca	pH	y = 511.47475x1 − 1650.53851	6.503 *	988,446.8	582.4675	583.2175
EC	OM, AN, OP, EK, Na, SS, Cl	y = 1.33274x2 + 1.82796x3 + 2.92141x4 − 0.42446x5 − 0.48058x8 + 60.87424x9 + 11.9421x10 − 243.51569	23.832 ****	30,621.06	462.4385	476.8385
	OM, AN, OP, EK	y = 3.95122x2 + 2.98755x3 − 0.39564x4 + 1.07747x5 − 321.51118	10.426 ****	160,923.1	519.4896	524.105
	AN, OP, EK	y = 2.94944x3 − 0.20299x4 + 1.22617x5 − 303.38725	14.770 ****	159,205.1	517.0817	519.9389
	OM, OP, EK	y = −0.55038x2 − 2.03912x4 + 2.17407x5 − 27.70535	7.743 ***	216,098.1	528.6922	531.5493
	OM, AN, EK	y = 3.19697x2 + 3.04201x3 + 1.01082x5 − 329.86525	14.845 ****	166,134.6	518.7007	521.5578
	OM, AN, OP	y = 14.98354x2 + 3.67624x3 − 0.86753x4 − 401.25628	12.650 ***	361,255.1	548.2186	551.0757
	OM, AN	y = 18.90927x2 + 3.64383x3 − 391.9094	19.876 ****	225,973.8	528.3903	529.9903
	OM, OP	y = 25.79778x2 + 0.29214x4 − 59.38383	6.227 **	444,310.2	554.0822	555.6822
	OM, EK	y = -5.2139x2 + 1.91386x5 − 44.42151	11.504 ***	252,359.4	532.5868	534.1868
	AN, OP	y = 4.02751x3 + 3.1315x4 − 357.33798	16.983 ***	264,345.8	534.3502	535.9502
	AN, EK	y = 2.98437x3 + 1.17241x5 − 310.06965	23.609 ****	162,224.6	515.7957	517.3957
	OP, EK	y = -2.06932x4 + 2.15503x5 − 29.73744	12.387 ***	363,322.2	546.4354	548.0354
	OM	y = 27.09524x2 − 57.23863	13.209 **	442,144.4	551.8966	552.6466
	AN	y = 4.70805x3 − 119.31884	18.939 ***	333,987.1	541.2363	541.9863
	OP	y = 4.37717x4 + 82.2413	8.002 *	568,552.2	561.4519	562.2019
	EK	y = 1.66587x5 − 69.11515	24.126 ***	257,703.9	531.3832	532.1332
Na	EK, SS, Cl	y = 0.04139x4 + 18.39293x9 + 0.85299x10 + 115.89144	3.682 *	7144.524	399.1355	401.9927
	EK	y = 0.17083x4 + 141.03401	7.214 *	7757.603	398.264	399.014
SS	OM, OP, EK, EC, Na, Cl	y = 0.02458x2 − 0.00165x4 + 0.001061x5 + 0.00106x7 + 0.00281x8 − 0.00727x10 + 1.18	3.860 *	0.94854	65.91208	76.0939
	OM, OP, EK	y = 0.02292x2 − 0.00143x4 + 0.00279x5 + 1.54253	6.752 **	1.094979	65.36759	68.22474
	OM, OP	y = 0.05669x2 + 0.00156x4 + 1.50193	8.501 **	1.205665	67.02684	68.62684
	OM, EK	y = 0.01966x2 + 0.0026x5 + 1.53084	10.722 **	1.091904	63.26074	64.86074
	OP, EK	y = -0.0001687x4 + 0.00358x5 + 1.62715	10.310 **	1.088537	63.1434	64.7434
	OM	y = 0.06363x2 + 1.5134	17.880 ***	1.228649	65.74443	66.49443
	OP	y = 0.01054x4 + 1.81315	11.073 **	1.358076	69.55028	70.30028
	EK	y = 0.00354x5 + 1.62394	21.907 ***	1.088511	61.14248	61.89248
Cl	OM, AN, EK, EC, Na, SS	y = −0.57947x2 − 0.05806x3 + 0.01001x5 + 0.06129x7 + 0.04744x8 − 2.15476x9 + 2.93581	9.170 ***	336.9845	289.0811	299.2629
	OM, AN, EK	y = -0.44025x2 + 0.13221x3 + 0.0745x5 − 14.13053	5.045 *	1310.039	334.6765	337.5337
	OM, AN	y = 0.71785x2 + 0.17657x3 − 18.70358	6.328 **	1739.418	343.4493	345.0493
	OM, EK	y = −0.80581x2 + 0.11375x5 − 1.7245	5.045 *	1516.646	338.2414	339.8414
	AN, EK	y = 0.14015x3 + 0.05225x5 − 16.85656	7.854 **	1398.37	335.1560	336.756
	OM	y = 1.11452x2 − 2.4863	5.238 *	2554.424	356.0518	356.8018
	AN	y = 0.21697x3 − 8.35532	9.410 **	1993.585	346.6319	347.3819
	EK	y = 0.07542x5 − 5.54089	10.546 **	1754.151	341.7698	342.5198

¹ x₁–x₁₀ are pH, OM, AN, OP, EK, Ca, EC, Na, SS, and Cl, respectively. ² The significance level (p) of the regression equation is indicated by an asterisk, with “*”, “**”, “***”, and “****” representing significant correlation at the 0.05, 0.01, 0.001, and 0.0001 levels. ^3,4,5 Root mean square error (RMSE), Akaike Information Criterion (AIC), and its corrected version (AICc), respectively; smaller values indicate better model performance and greater significance.

), indicating the reliability of the multivariate regression relationships identified. Priority was given to selecting soil physicochemical variables closely associated with orchard inputs, such as pH, OM, AN, OP, EK, and Ca, so that the selection of independent variables could be standardized. Simultaneously, root mean square error (RMSE), Akaike Information Criterion (AIC), and its corrected version (AICc) were introduced so that comprehensive evaluation and optimization of the regression models could be ensured. When Na was set as the dependent variable (y), the regression equation based on the significantly correlated independent variables (x₂, x₅, x₇, x₉, x₁₀) identified through canonical correlation analysis was not found to be statistically significant (p > 0.05), which was noteworthy. Consequently, the independent variables were redefined as EC, SS, and Cl, from which the following equation was derived: y = 0.04139x₇ + 18.39293x₉ + 0.85299x₁₀ + 115.89144. Significant univariate or multivariate regression relationships were demonstrated by the results between dependent and independent variables. Among these, the regression equations involving the following variable combinations—pH with OM and Ca (p < 0.05); OM with OP and EK (p < 0.0001); AN with EK (p < 0.05); OP with OM (p < 0.0001); EK with OM and AN (p < 0.0001); Ca with pH (p < 0.05); EC with AN and EK (p < 0.0001); Na with EK (p < 0.05); SS with EK (p < 0.001); and Cl with AN and EK (p < 0.01)—showed the closest alignment with both observed values and real-world conditions. An optimal balance between goodness of fit and scientific validity was struck by these regression equations for Yantai apple orchards, revealing pronounced and complex univariate and multivariate regression relationships among various soil physicochemical properties, thereby allowing a strong theoretical foundation to be provided for optimizing orchard fertilization strategies.

3.5. Analysis of Regression Fit Model for Soil Physicochemical Properties in Yantai Apple Orchards

Univariate and multivariate regression equations were established to explore the relationships among soil physicochemical properties. 2D linear fitting models and 3D planar fitting models were subsequently constructed, incorporating key statistical indicators, including significance levels, to elucidate the integrated and visualized relationships among variables, ultimately resulting in the establishment of a multivariate regulation model of soil physicochemical properties for Yantai apple orchards. Clear linear or planar relationships among soil physicochemical properties were revealed by the results. Specifically, an increase in pH was promoted by simultaneous increases in OM and Ca (Figure 4a); an increase in OM was promoted by the joint rise in OP and EK (Figure 4b); increases in AN, Na, and SS were facilitated by a unilateral increase in EK (Figure 4c,h,j); an increase in OP was linked to a single increase in OM (Figure 4d); the rise of EK was promoted by combined increases in OM and AN (Figure 4e); the increase of Ca was facilitated by a single increase in pH (Figure 4f); and significant increases in EC and Cl resulted from the joint increase in AN and EK (Figure 4g,j). All corresponding regression fit models were found to be statistically significant (p < 0.05, 0.0001, 0.05, 0.0001, 0.0001, 0.05, 0.0001, 0.05, 0.001, 0.01) (Table 3), with comprehensive integration into the multivariate regulation model of soil physicochemical properties (Figure 4k). These findings indicated that soil physicochemical properties were influenced not only by the direct input of specific elements through orchard management but also by the synergistic effects of one or two other soil physicochemical properties. This enhanced the predictive power and interpretability of soil quality assessment and management in Yantai apple orchards.

4. Discussion

4.1. Current Status and Quality Assessment of Soil Physicochemical Properties in Yantai Apple Orchards

Orchard soil quality, which is influenced by site conditions, plays a crucial role in nutrient supply, which is essential for crop yield and quality. Descriptive statistics and frequency distribution analyses revealed that soil physicochemical properties in Yantai’s apple orchards vary significantly. This suggests wide variation in soil quality grades. The average pH is 6.45, which is slightly acidic and favorable for apple growth [57,69]. However, some orchards experience severe acidification (26.3%), with pH levels dropping to 4.73, negatively affecting nutrient availability [69]. The benefits of OM for yield and quality are well-recognized, and increasing its application is generally recommended [3,53,54]. The average OM in Yantai apple orchards is moderate at 22.42 g/kg, Although 36.8% of the orchards exhibit relatively high organic matter content, 63.2% remain at moderately low levels, with the lowest level reaching 26.3%. This indicates that the majority of apple orchards in the Yantai region still have low soil OM levels, reflecting generally insufficient application of organic fertilizers. In Yantai orchards, increases in soil AN (21.63–38.05%), OP (5.60–54.69%), and EK (14.74–125.92%) are positively correlated with improvements in apple yield, which rose by 7.47–33.80% [53,54,55].

However, the practice of high-input, high-output cultivation has led to declines in soil quality and nutrient absorption efficiency [5,70]. Excessive nutrient levels can also reduce the effectiveness of other nutrients and lead to the accumulation of harmful substances [8,71] (as discussed in Section 4.2 and Section 4.3). Nutrient levels in Yantai orchards are alarmingly high. The fluctuations exhibited by AN and EK are primarily observed at both low and high extremes, while OP is consistently found at high levels, indicating an excessive and uneven distribution of macronutrient inputs. Compared to previous studies on Yantai orchards [49,50,53,54,55,72], the levels of pH, OM, and macronutrients have significantly increased, providing a better foundation for improved yield and quality. These advancements are attributed to an enhanced understanding of soil improvement and fertility enhancement in orchards, along with the widespread adoption of new fertilizers, organic alternatives, innovative technologies, and micronutrient conditioners in soil testing and formula fertilization practices.

The distribution of Ca follows a “low at both ends, high in the middle” pattern, indicating insufficient calcium content and application in orchard soils. This deficiency could lead to physiological disorders, such as apple bitter pit [19,20]. The EC level is moderate to low, indicating uneven distribution. This finding indicates a potential to mitigate soil acidification, reduce the accumulation of macronutrients, such as nitrogen, phosphorus, and potassium, and improve levels of organic matter, calcium, and electrical conductivity (EC) in orchard soils. The Na level is high and requires special attention due to potential sodium carryover from agricultural inputs. Additionally, both SS and Cl are at low levels (57.9% and 94.7%); however, attention should be given to rising trends in certain orchards. The findings of this study provide valuable data for soil testing and fertilization strategies to improve soil quality in Yantai apple orchards.

4.2. Multivariate Statistical Analysis and Regression Fit Model Analysis of Soil Physicochemical Properties in the Yantai Apple Orchard Region

For Yantai apple orchards, the canonical correlation analysis, stepwise regression equation analysis, and regression fit model analysis are interconnected, forming a coherent scientific framework. This suggests a significant plane-fitting relationship between Ca, OM, and pH, emphasizing their interdependence and mutual influence. This finding is partially consistent with the results reported by Feng et al. and Cao et al. in citrus orchards, where pH was negatively correlated with OM but positively with Ca [73,74]. It more closely aligns with Zhang et al.’s study in peach orchards, which reported a positive correlation between pH and both OM and Ca [75]. Additionally, our results are consistent with those of Ganai et al., who also reported a significant positive correlation between pH and Ca in apple orchards [38]. Soil OM is significantly correlated with pH, OP, EK, EC, Na, SS, and Cl. Soil OM is significantly correlated with pH, OP, EK, EC, Na, SS, and Cl, with EK, OP, and OM showing the most significant plane-fitting relationship. This is consistent with the studies by Wang et al. and Ganai et al. on apple orchards, as well as those by Feng et al. and Cao et al. on citrus orchards and Zhang et al. on vineyards and peach orchards [38,73,74,75,76,77]. Soil AN is significantly correlated with EK, EC, and Cl, with EK as the independent variable and AN as the dependent variable showing a significant linear regression relationship (p < 0.05). This is consistent with the studies by Feng et al. and He et al. in citrus orchards, Zhang et al. and Wang et al. in apple orchards, and Zhang et al. in vineyards orchards [31,73,76,77,78] and consistent with the fact that K + promotes the absorption and transportation of available nitrogen in plant roots [79,80]. Soil OP is significantly correlated with OM, EK, EC, and SS, with only OM as the independent variable showing the highest fitting degree in linear regression (RMSE, AIC, AICc). This result is consistent with the studies by Wu et al. in pomelo orchards, Zhang et al. in peach orchards, Ganai et al. Zhang et al. and Wang et al. in apple orchards, Feng et al. in citrus orchards, and Zhang et al. in vineyards [31,38,73,75,76,77,81] but inconsistent with He et al.’s study of citrus orchards, which found a significant negative correlation between OP and EK [78]. Soil EK is significantly correlated with OM, AN, OP, EC, Na, SS, and Cl, with OM and AN exhibiting the most significant positive planar effect on EK. This aligns with the studies of Zhang et al. on vineyards and Feng et al. on citrus orange orchards [73,77], and it is consistent with the significant positive correlation between EK and OP in the work of Zhang et al., Wang et al., and Ganai et al. on apple orchards, Zhang et al. on peach orchards, and Wu et al. on pomelo orchards [31,38,75,76,81]. Soil Ca is significantly positively correlated with pH, with only pH exerting a significant positive linear effect on Ca, exhibiting a non-significant negative correlation with AN, Na, and Cl. This is in line with the inhibitory effect of Ca²⁺ on the uptake and translocation of nitrogen in plant roots [79], which contrasts with the findings of Xing et al., who reported that calcium improved nitrogen utilization efficiency in apples [82]. Soil EC is significantly correlated with OM, AN, OP, EK, Na, SS, and Cl, with EK and AN exerting the largest significant positive plane-fitting effect on EC. This aligns with the findings of Mazur et al. [83], where EC showed a significant correlation with AN and EK, 84, and Chaudhary et al. [84], where EC exhibited a correlation with OP in a maize–wheat cropping system. Soil Na is significantly correlated with OM, EK, EC, SS, and Cl, with only EK demonstrating the highest fitting degree of a significant positive linear effect on Na. Soil SS is significantly correlated with OM, OP, EK, Na, and Cl, showing a similar pattern to AN and Na, where only EK exerts the largest significant positive linear effect on SS. This aligns with the findings of Singh et al., where a significant correlation was observed between SS and Na in field soils [85]. Soil Cl is significantly correlated with OM, AN, EK, EC, Na, and SS, with EK and AN exerting a significant positive plane-fitting effect on Cl. This confirms Chen et al.’s study indicating a positive correlation between Cl and EK, as well as a highly significant correlation with Na, particularly in soils with higher SS [86].

Soil quality results from the complex interactions among various physicochemical properties. In addition to the established relationship between changes in soil properties and orchard inputs [87], this study further explores the interrelationships among soil physicochemical characteristics across different orchards. Notably, significant univariate linear and multivariate planar regression relationships were observed among multiple variables (Figure 4a–k). First, strong correlations were observed among key nutrient indicators. For instance, exchangeable potassium (EK) exhibited a significant positive linear correlation with available nitrogen (AN), and organic matter (OM) showed a similar trend with available phosphorus (OP). Furthermore, OM and AN jointly influenced EK in a significant planar regression model, while EK and OP significantly affected OM. These relationships may result from the synchronized application of multiple macronutrient fertilizers, which can lead to synergistic or antagonistic effects on nutrient availability. Secondly, soil calcium (Ca) showed no significant correlations with any nutrient-related indicators and was only significantly associated with pH in a univariate linear regression model. This may be due to fertilization practices emphasizing macronutrients while neglecting calcium supplementation. This also indicates that the application and recycling of macronutrients may not effectively improve calcium availability in the soil. Moreover, an antagonistic relationship may exist between soil acidification and calcium availability, with OM potentially playing a synergistic role in this regulatory process [79,88]. Further analysis showed that electrical conductivity (EC) was significantly correlated with all soil physicochemical properties, except pH and Ca. Soluble salt content (SS) and chloride (Cl) showed positive correlations with most indicators, while sodium (Na) was significantly correlated with only three. Stepwise regression models indicated that increases in EC and Cl were best explained by elevated levels of AN and EK, forming a strong bivariate planar regression model. In contrast, increases in Na and SS were primarily associated with a univariate linear rise in EK. This may be due to the overapplication of nitrogen- and potassium-based low-quality fertilizers, particularly those containing high levels of Na and Cl. Finally, among all orchard input-related indicators, OM was the most frequently involved explanatory variable, showing significant linear or multivariate associations with positive soil quality indicators (pH, OM, AN, OP, EK, Ca), followed by EK. Conversely, among negative soil quality indicators (Na, SS, Cl), EK showed the strongest associations, followed by AN. These findings offer a theoretical basis for improving soil quality in Yantai apple orchards and suggest a strategy to optimize fertilization practices for achieving both high yield and fruit quality. These results provide a theoretical foundation for improving soil quality in Yantai apple orchards and propose a feasible strategy to optimize fertilization practices, ensuring both high yield and quality.

4.3. Soil Quality Improvement Program for Yantai Apple Orchards: Research Summary and Future Research Directions

This study builds on a clear understanding of soil quality in Yantai derived from descriptive statistics and frequency distribution analyses (Section 4.1). It aims to improve the overall quality of soil physicochemical properties. To achieve this, canonical correlation, stepwise regression equations, and regression model fitting analyses were applied to identify univariate linear and multivariate planar regression relationships in Yantai (Section 4.2) and, subsequently, to develop a multivariate regulation model of soil physicochemical properties. Based on these results, optimized fertilization strategies were developed to prevent tree growth from being significantly limited by the most restrictive nutrients or adversely affected by antagonistic interactions [89]. In Yantai apple orchards, the first priority should be to raise agricultural growers’ awareness of soil testing and formula fertilization, which will contribute to the overall improvement of soil quality while reducing the variation in soil quality levels across orchards. Applying organic and calcium-based fertilizers can increase organic matter (OM) and calcium (Ca) content while mitigating soil acidification. Applying organic and calcium-based fertilizers can directly increase soil organic matter (OM) and calcium (Ca) content, while simultaneously alleviating soil acidification. Increased OM further enhances the availability of phosphorus (OP) and exchangeable potassium (EK). For example, supplementing organic fertilizers with calcium-rich alkaline materials, such as lime, Ca-Mg-P fertilizers, Si-Ca-K-Mg fertilizers, and wood ash, can neutralize soil acidity, supply calcium, and reduce the demand for phosphorus- and potassium-based fertilizers. Additionally, these practices help prevent soil compaction and improve physical soil structure [90,91,92]. Reducing the application of nitrogen, phosphorus, and potassium source fertilizers can decrease both total and available nitrogen, phosphorus, and potassium in the soil [87]. Moreover, a decrease in EK further reduces the availability of alkali-hydrolyzable nitrogen (AN) and increases levels of the negative soil quality indicator chloride (Cl). Conversely, reduced AN levels also lead to lower EK. The simultaneous decline in AN and EK leads to reductions in other negative soil quality indicators, including soluble salts (SS) and sodium (Na). Increased application of organic fertilizers helps restore EK levels and stabilizes electrical conductivity (EC) within acceptable limits. Fertilizer residue issues should also be addressed to prevent secondary salinization. For instance, orchards in mountainous or coastal areas with high chloride, such as in Yantai, should avoid using chloride-based fertilizers (e.g., potassium chloride and magnesium chloride) and refrain from applying sodium-containing or high-sodium fertilizers in areas with high sodium content.

In summary, a lack of understanding of soil quality and the adoption of suboptimal fertilization strategies in Yantai apple orchards may lead to nutrient deficiencies, resource waste, and environmental pollution. Additionally, antagonistic interactions among nutrients can further reduce nutrient use efficiency. To address these challenges, this study introduced several key innovations. First, this study incorporated sodium (Na), soluble salts (SS), chloride (Cl⁻), and electrical conductivity (EC)—alongside conventional soil physicochemical parameters—into the soil quality classification standard for Yantai apple orchards for the first time. Using descriptive statistics analysis, frequency distribution, and curve fitting analysis, the study provided a comprehensive overview of current soil conditions in Yantai orchards. This analysis delivers a robust data foundation for developing optimized fertilization strategies. These indicators are notable for their ease of testing, accessible data acquisition, and convenient field measurement. This expansion broadens the methodological toolkit for orchard soil quality assessment and offers strong potential for practical, large-scale application.

To address the pressing soil quality issues in Yantai apple orchards, this study presents several key innovations that enhance both theoretical understanding and practical management of orchard soils. (1) Indicator system innovation: This is the first study to integrate non-traditional salinity-related indicators, including sodium (Na⁺), chloride (Cl⁻), electrical conductivity (EC), and soluble salts (SS), into a localized soil quality classification system tailored for Yantai apple orchards. Unlike conventional systems focused primarily on macronutrients (N, P, K), this expanded indicator framework captures salinization risk and ionic imbalance, which are particularly relevant in coastal or intensively fertilized regions. Excessive fertilizer application in Yantai has led to severe soil acidification and decreased availability of secondary and micronutrients, ultimately reducing both apple yield and fruit quality. The selected indicators are easy to measure in the field and offer a practical, scalable method for assessing soil degradation [22,29]. (2) Model framework innovation: This study developed a coherent and region-specific multivariate regulation model of soil physicochemical properties by sequentially applying canonical correlation analysis, stepwise regression analysis, and multidimensional regression fitting. For the first time in this research domain, root mean square error (RMSE), Akaike Information Criterion (AIC), and its corrected version (AICc) were introduced as comprehensive metrics to evaluate and optimize the regression models. These methods identified not only direct one-to-one relationships but also synergistic multivariate interactions, such as OM–Ca–pH, that are crucial for understanding complex nutrient dynamics in orchard soils. The use of 2D univariate and 3D bivariate visualization techniques transforms abstract statistical relationships into actionable insights, supporting both scientific understanding and practical decision making in fertilization strategies [13,14]. (3) Practical innovation in precision fertilization: Moving beyond the conventional “supplement what is lacking” approach, the proposed precision fertilization strategy incorporates multivariate interactions and feedback mechanisms. This allows for mitigation of antagonistic nutrient effects, reduction of excess chemical inputs, and restoration of balanced soil functionality. The model was validated using field data from Yantai and represents the first region-specific trial in which fertilization was guided by multidimensional soil interaction models. This approach is data-driven, operationally feasible, and adaptable to similar agro-ecological regions. Together, these innovations respond directly to the growing challenge of soil degradation and productivity decline in Yantai’s apple orchards, offering a science-based and regionally adapted solution for sustainable soil improvement and brand development. Beyond these methodological innovations, the study also demonstrated strong experimental rigor. All data were collected from a uniform soil type (brown earth) across Yantai districts, the same production years (7th to 9th years of the full fruit-bearing period), the same apple cultivar (“Yantai Fuji 3”), and the same phenological stage (early fruit expansion). Soil sampling followed a standardized protocol, including orchard selection, tree location, sampling tools, and lab procedures, all described in Section 2.3. This strict methodology ensured the authenticity and reliability of the data, making the results highly applicable to similar agroecosystems.

However, several limitations should be noted. First, although the study analyzed 19 orchards across Yantai and 10 major soil parameters, the sample size remains limited for robust statistical inference. Thus, expanding the number of test orchards is necessary to enhance the reliability of CV, AV, and multivariate analyses. Second, the study focused on macronutrients and key physicochemical indicators, omitting critical aspects, such as micronutrients, microbial activity, heavy metals, and organic pollutants. This omission limits the comprehensiveness of soil quality assessment and model development. Additionally, it does not explore the relationships between soil parameters and biological outcomes, such as yield, nutrient content, or fruit quality, factors essential for practical implementation and broader adoption. Moreover, although the classification system is grounded in both theoretical foundations and years of field experience, some references are outdated, and data gaps remain regarding specific soil characteristics in Yantai apple orchards. These shortcomings may introduce bias and warrant further validation. While the multivariate analysis was rigorous, it oversimplified the spatiotemporal dynamics and nutrient interactions among soil parameters. The stepwise regression models primarily selected input-related variables (e.g., pH, OM, AN, OP, EK, Ca), limiting insights into more complex interactions. Although fertilization strategies were optimized based on these models, no cost–benefit or feasibility analysis was conducted, potentially limiting their real-world application. Finally, the study did not investigate soil quality across different climates, terrains, soil types, cropping systems, or irrigation conditions. Future research should expand sampling, refine evaluation criteria, integrate a broader set of variables, and develop high-dimensional regression models based on the existing one- and two-dimensional frameworks. Efforts to increase regional replication, conduct production trials, monitor environmental quality, and standardize soil testing under varied conditions will be essential for achieving both agronomic effectiveness and economic feasibility. Overall, this study lays a methodological foundation for exploring multivariate interactions in orchard soils and offers valuable insights for improving nutrient management efficiency, supporting the innovation of the “Yantai Apple” brand and guiding future scientific research.

5. Conclusions

This study investigated the physicochemical properties and quality distribution of soils in Yantai apple orchards, identifying key factors limiting soil quality improvement. These include variations in soil physicochemical characteristics, low pH, and excessive levels of macronutrients, Na, and SS, as well as insufficient levels of OM, Ca, and EC. Multivariate statistical analysis was used to develop univariate 2D linear regression models and bivariate 3D plane regression models as visualization tools. A multivariate regulation model of Yantai apple orchards was also summarized, identifying key independent variables.

The proposed optimized fertilization strategy focuses on promoting awareness of soil testing and formula fertilization techniques in Yantai advantageous apple production areas. This strategy involves increasing organic and calcium source fertilizers, reducing nitrogen, phosphorus, and potassium source fertilizer application, and prohibiting chloride and sodium source fertilizers. The goal of this strategy is to reduce variations in soil properties, improve soil quality, mitigate acidification, increase organic matter, optimize the effectiveness of macronutrients, enhance secondary nutrient supply, stabilize electrical conductivity, and minimize the risks of negative indicators.

This study focuses on the “Yantai Apple” geographical indication industry, providing insights into the current soil quality status in key production areas and its relationship with soil physicochemical properties. It provides a theoretical basis and practical blueprint for optimizing orchard fertilization strategies and promoting sustainable industry development, offering crucial guidance for the green, efficient, and sustainable growth of the apple industry in this region.