Evolution trend of soil fertility in tobacco-planting area of Chenzhou, Hunan Province, China

Abstract In this study, the data of fertility indicators of soil samples (0–20 cm) in 1980s, 2000 and 2015 in Chenzhou city were used, and the soil integrated fertility index (IFI) was calculated. The results showed that the soil pH was decreased, total nitrogen (TN), organic matter (OM), available phosphorus (AP) and potassium (AK), exchangeable calcium (Ca2+), magnesium (Mg2+) and available copper (Cu) contents were increased, total phosphorus (TP), available sulfur (S) and water-soluble chlorine (Cl−) contents were decreased, total potassium (TK), available boron (B), iron (Fe), manganese (Mn) and zinc (Zn) were decreased first and then increased. In 2015, most of the fields were higher in pH, OM, TN, AN, AK, Ca2+, Mg2+, S, Fe, Mn, Cu and Zn, suitable in B, but lower in TP, AP, TK, available molybdenum (Mo) and Cl−. Most of the fields were in the middle grade of IFI in 2000 and 2015, and the mean IFI increased from 0.492 to 0.556 from 2000 to 2015. Thus, for soil improvement, more attention should be paid to adjust soil pH, reduce the application of organic, nitrogen and calcium fertilizers, while increase the fertilizer application of other nutrients.


Introduction
Soil fertility influences the growth, yield and quality of tobacco [1][2][3][4][5], which is continuously concerned in China. So far, lots of studies have been conducted with many literature studies published on fertility evaluation of tobacco-planting soil; for example, more than 300 literature studies in Chinese could be retrieved by the title or keywords of "tobacco," "fertility" and "evaluation" in the China National Knowledge Infrastructure (www.dlib.cnki. net/), which almost covered all the tobacco-planting regions in China and at various scales of province, city and county.
Tobacco usually is planted in the drylands with sandy soil texture, but in many areas of southern China, such as Guangdong, Fujian, Jiangxi, Guangxi and south Hunan and Anhui, it is very common that tobacco is planted in rice fields in the high ridge form (no matter what kind of soil texture) and rotated with late rice. Chenzhou city, with a long history of tobacco-planting as early as in 1,593 and where most paddy fields are under tobacco-rice rotation [6], is the most important and typical planting region of flue-cured tobacco with burnt-pure sweet aroma in China [7]. The area of tobacco-planting in Chenzhou is about 2.67 × 10 4 hm 2 in recent years, which plays an important role in ensuring the high-quality raw materials supply of the tobacco industry and the local social and economic sustainable development.
Some literature studies were published about tobaccoplanting soil nutrient status in Guizhou [8][9][10][11], which played an important guiding role in improving the soil fertility and quality of tobacco-planting fields. However, there are limited fertility indicators (e.g., pH, organic matter (OM), total nitrogen (TN) and available boron) were involved in the above-mentioned studies, and so far, there are few reports reflecting the changes in soil fertility [9]. Two questions are still unclear and should be answered which are concerned with the influences of tobacco-planting on soil fertility: (1) does tobacco-planting really can improve soil fertility? if so, by how much? Also, a new round of tobacco-planting soil improvement is underway in China; it is urgent and helpful to know the status and evolution of soil fertility; therefore, this study was conducted in order to quantitatively analyze the soil fertility of tobacco-planting fields in Chenzhou in order to provide further scientific guidance for fertilization and soil improvement for tobacco-planting in Chenzhou.

Data sources of soil fertility indicators
The data of soil fertility indicators used in this study came from three periods, the 2nd national soil survey conducted in the 1980s [12], and tobacco-planting soil surveys conducted in 2000 and 2015, which included 350, 746 and 1,055 soil samples, respectively. The obtained data of the 1980s were the statistic information of all soil samples, and no data of each sample is available.
According to the historical records, the soil sample of the plough layer in each field was collected randomly at 5-8 points with stainless steel soil drill and then mixed completely. The measured soil properties (soil fertility indicators) included OM, TN, total phosphorous (TP), total potassium (TK), available nitrogen (AN), available phosphorous (AP), available potassium (AK), available boron (B), available iron (Fe), available manganese (Mn), available copper (Cu), available zinc (Zn) and available molybdenum (Mo) for all soil samples in the three periods, and pH (H 2 O), exchangeable calcium (Ca 2+ ), exchangeable magnesium (Mg 2+ ), available sulfur (S), and water-soluble chlorine (Cl − ) for the soil samples in 2000 and 2015. The detailed determination methods for soil fertility indicators could be found in related literature studies [13,14].

Quantitative assessment of soil fertility
There are various methods for the assessment of soil fertility; in this study, soil integrated fertility index (IFI) was used to evaluate soil fertility, and IFI was calculated with the fuzzy mathematics method [15]; first, the membership function types and inflection points of indicators were determined; second, the membership values of the indicators were calculated; third, the weights of the indicators were determined, and finally, IFI was calculated for soil samples according to the following formula: where W i stands for the weight of indicator i and N i for the membership value of indicator i. IFI is ranged from 0 to 1. The higher the IFI value, the higher the soil fertility. Generally, IFI is divided into five grades according to the equidistant method [16,17]: ≥0.80 (higher), 0.6-0.8 (high), 0.4-0.6 (Middle), 0.2-0.4 (low) and <0.2 (lower).

Grading standards of fertility indicators
There are many reports available in China on the grading standards of soil fertility indicators for tobacco-planting fields. In this study, the indicators were divided into 4 or 5 grades as in Table 1 based on the actual situation of tobacco-planting soils in Hunan Province [18,19] and the corresponding grading of tobacco-planting soils in neighboring areas of Hunan Province [20][21][22][23][24][25][26].

Calculation of membership value of fertility indicator
The membership function types of the fertility indicators were determined according to their effects on the growth, yield and quality of tobacco, among which, pH, OM, TN,  [15,27]. Combining with the actual situation of tobacco growing soils in Chenzhou [8][9][10][11]17], the turning points of membership function of each fertility indicator were determined according to expert experience and related literature studies published [5,[16][17][18][19]22,26,28] ( Table 2; x 1 , lower limiting value; x 2 , upper limiting value; x 3 , lower optimal value; and x 4 , upper optimal value). The membership functions of S-type and parabolic indicators were calculated as follows:

Determination of weights of indicators
The weights of indicators were determined by principal component analysis (PCA), which is commonly used in soil fertility and quality evaluation [21,29,30]. In order to avoid the appearance of a negative value in weight, the measured values of pH, OM, TN, TP, TK, AN, AP, AK, AS, Ca, Mg, S, B, Fe, Cu, Zn, Mo and Cl −1 were standardized using Z-score standardization method, while those of Mn were standardized using the negative range normalization method in SPSS software [31,32]. The KMO test coefficient obtained was 0.727, indicating that the data structure was good and the linear correlation between the data was satisfied, which could be used for principal component analysis, while the p value of Bartlett's test was less than 0.001, rejecting the null hypothesis, indicating that the data could be extracted by principal components. The obtained weights of indicators are shown in Table 3, and the detailed routine for the acquisition of indicator weights was not listed here.

Data processing and statistics
Microsoft Excel 2016 and IBM Statistics SPSS 22.0 software were used for the statistical analysis of the data, and Duncan test method (p < 0.05) was used for the analysis of variance and multiple comparisons [31,32]. In the 1980s, OM, TN, TK, AN and AK belonged to the low variation (C.V. < 10%,), while TP and AP belonged to the moderate variation (C.V. = 10-100%). In 2000, pH, OM, TN, TP, TK, AN, AP, AK, Ca, Mg, S, B, Fe and Cl − belonged to moderate middle variation, while Mn, Cu, Zn and Mo belonged to the strong variation (C.V. > 100%). In 2015, pH, OM, TN, TP, TK, AN, AP, AK, Ca, Mg, S, B, Fe and Mn remained in the moderate middle variation, Cu, Zn and Mo remained in the strong variation (C.V. > 100%), while Cl − changed from the moderate variation to the strong one. It also can be seen from Table 4

Statistics and comparison of soil IFIs
Tables 6 and 7 present the general and grade statistical results of soil IFIs of tobacco-planting fields, respectively. Tables 6 and 7 show that there was no sample with IFI <     [31,32]. It is shown in Table 7 Table 7 also shows that the proportion of samples in the higher grade of IFI was increased from

Discussion
Soil fertility assessment is one of the most basic works in soil science research, but it is very important for crop planting; therefore, the latest literature studies could be found even now [33][34][35][36][37]. Meanwhile, soil fertility evaluation of tobacco planting also has been reported more so far; however, the fertility indicators mainly involve pH, OM, TN, AN, TP, AP, TK, AK, Ca 2+ , Mg 2+ , trace elements of B, Fe, Mg, Cu and Zn, and Cl − ; in our study, besides the above indicators, S and Mo were also added; thus, it should be said that the fertility indicators are more complete in our study, which would enable the obtained results are closer to the reality and more feasible in guiding the scientific fertilization and soil improvement.
It should be pointed out that CEC is also one of the most important indicators of soil fertility; however, as in many studies of soil fertility assessment conducted in China, the determined indicators of soil fertility usually include soil pH and the contents of OM and main nutrients [1][2][3][4][5], which can not only understand the real state of in each index (whether suitable or not for high-quality tobacco planting) in order to guide the reasonable fertilization and soil improvement, but also can evaluate soil comprehensive fertility based on these indicators. The reason why CEC is rarely used is that CEC is a comprehensive indicator of soil fertility, it can only indicate the general level of soil fertility, but cannot provide the real information on soil pH and the contents of OM and nutrient contents, thus cannot guide the reasonable fertilizer application and soil improvement; therefore, CEC was also not used in our study. Meanwhile, soil pH and OM content are the factors that can influence CEC, so CEC can be omitted when pH and OM are used for soil fertility assessment. We add some content in the second revised manuscript. Meanwhile, the Ca/K and Ca/Mg ratios should be considered in the assessment of soil fertility for tobaccoplanting, which is important to instruct the scientific application of potassium by reflecting the nutrient antagonism in soils, but currently, there are no threshold values or grade classifications of Ca/K and Ca/Mg for tobacco, so these two ratios were also not used in this study. As shown in Table 5, 65.42 and 62.37% of the samples were higher in pH (≥7.0) in 2000 and 2015, respectively. The high value of soil pH of tobacco-planting fields in Chenzhou could be attributed to the application of superphosphate fertilizer and the habit of local farmers using fired soil to improve soil quality [38,39], and it may also be related to that tobacco-planting fields in Chenzhou are mostly located in the limestone hill and mountainous area [10], which also resulted in the increases of Ca 2+ and Mg 2+ from 2000 to 2015, increased significantly by 157.31 and 20.44%, respectively ( Table 4).
OM was increased from 38.8 g/kg in the 1980s to 45.87 g/kg in 2000 and to 48.00 g/kg in 2015, the increase was decided by tobacco-rice rotation, straw returning to the field and organic fertilizer application [11,40,41]. AN increased from 147.97 mg/kg in the 1980s to 224.76 mg/kg in 2000 and then decreased to 202.98 mg/kg in 2015, the decrease in AN from 2000 to 2015 is because the higher content of AN in 2000 is unsuitable (suitable grade is 100-180 mg/kg) for the high-quality tobacco [25], and thus, the applied amount of nitrogen fertilizer was reduced gradually [42]. AP increased from 8.90 mg/kg in the 1980s to 28.20 mg/kg in 2000 and to 36.48 mg/kg in 2015; the remarkable continuous increase could be attributed to the long-term excessive application of phosphatic fertilizer by farmers in China due to the cheap price and yieldincrease effect [43,44]. AK increased from 76.6 mg/kg in the 1980s to 120.75 mg/kg in 2000 and to 205.71 mg/kg in 2015, the continuous significant increase was contributed by the large amount application of potassium to guarantee the high-quality tobacco leaves usually with high K content [1][2][3][4][5]. S deceased from 51.92 mg/kg in 2000 to 39.42 mg/kg in 2015, which could be attributed to the reduced use of potassium sulfate for tobacco-planting because it caused soil acidification, and usually high content of S would worsen the quality of tobacco leaves [45,46]. Cl − decreased from 18.49 mg/kg in 2000 to 6.30 mg/kg in 2015, which could be attributed to the worry that high Cl − content could severely deteriorate the quality of tobacco leaves, so chlorine fertilizer is seldom used for tobacco-planting in many regions [47]. Mo decreased from 0.25 mg/kg in the 1980s to 0.21 mg/kg in 2000 and to 0.16 mg/kg in 2015, which may be related to little concern about Mo and little literature on Mo fertilizer application for tobacco-planting in Hunan [48]. Cu increased from 3.74 mg/kg in the 1980s to 4.36 mg/kg in 2000 and to 4.70 mg/kg in 2015, the increase, on the one hand, may be related to the application of livestock and poultry manure which usually containing Cu [49,50], and on the other hand, may be related to the higher Cu content in paddy soil itself [51,52], as for the changes of B, Zn, Mn and Fe, which decreased from the 1980s to 2010 and then increased from 2000 to 2015, which could be attributed to the gradual application of related trace fertilizers in the farmlands [53].
Li et al. [23] compared soil fertility indicators tobaccogrowing areas in Kunming in southwest China in 2010 and 2020 and found that from 2010 to 2020, soil pH decreased by 0.15 units. Soil SOM, N, P and K increased by 1.87 g/kg, 7.21 mg/kg, 5.17 mg/kg and 53.05 mg/kg, respectively; the overall soil fertility increased, and these findings are generally similar to the results in our study.
The change of climate may influence the changes in soil chemical properties, so we analyzed the correlation between the mean annual temperature (T) and precipitation (P) with years from 2000 and 2015, and the results showed that the ranges of the two parameters were 17.9-19.1°C and 1069-1854 mm with the means of 18.6°C and 1450 mm, respectively; but there was no significant correlation between T and P with year. The Pearson coefficients were 0.499 (p = 0.082) and 0.224 (p = 0.462), respectively, which means it is hard to clarify the influence of climate change on the change of soil fertility instructors.
Soil fertility affects or determines the growth, yield and quality of tobacco; it also determines the economic benefits of local tobacco-planting farmers, so more and great concerns have been paid continuously to soil improvement in the tobacco-planting regions with higher input and sufficient guarantee, almost all the tobacco-planting regions in China have formulated and implemented the technical regulation of flue-cured tobacco planting, which enable the same measures adopted for tobacco-planting in ploughing, ridging, fertilization, irrigation, film mulching, etc. It may homogenize the soil fertility of tobacco-planting fields in a large region, so no significant difference in IFI was found in this study between most tobacco-planting regions in Chenzhou, in which significant difference in IFI was only founded between Guiyang with Linwu (Sig. = 0.009) and Anren (Sig. = 0.014) in 2000 and between Anren and Suxian (Sig. = 0.047) in 2015. Our study also showed the increasing tendency of soil IFI of tobacco-planting fields in Chenzhou, it was not only proved by the increases in the contents of OM, AN, AP and AK from 1980 to 2000 (Table 1), meanwhile, but also proved by the proportion of fields with the lower grade of IFI was decreased from 12.87% in 2000 to 1.61% in 2015, the proportion of fields with the higher grade of IFI was increased from 9.79% in 2000 to 27.87% in 2015, and IFI meanly increased from 0.492 in 2000 to 0.556 in 2015, increased by 13.00%. However, it should be pointed out that the soil fertility of tobacco-planting fields in Chenzhou is still in the middle level of IFI (0.4-0.6), the proportion of tobacco-planting fields with the middle grade of IFI was the highest (77.35% in 2000 and 70.52% in 2015), and there was no tobacco-plating field with the highest grade of IFI (≥0.8); therefore, the soil fertility of tobaccoplanting fields in Chenzhou still needs to be promoted. These tobacco-planting fields should be paid attention to the modification of soil alkalinity or acidity; 90.33, 83.13 and 64.55% of the samples were higher in OM, TN and AN, respectively, and these fields need to reduce the use of organic and nitrogen fertilizers. In Ca 2+ and Mg 2+ , 85.12 and 44.45% of the samples were higher, respectively; these fields should be controlled by the application of alkaline substances or calcium and magnesium fertilizers. 27.61, 32.04, 12.99, 66.54 and 76.30% of the samples were lower in AK, Mg 2+ , Mn, Mo and Cl − , respectively; these fields should be applied with the corresponding fertilizers. Usually, tobacco is a chlorine-free crop, but chlorine is also one of the essential nutrients for tobacco growth [5], and some studies conducted in south China have shown that proper application of chlorine fertilizer to Cl-deficient soils could increase the elasticity, oiliness and yield of tobacco leaves while not reducing the quality of tobacco leaves [47,54]. Nevertheless, for the fields with higher contents of other various nutrients, such as S, Fe, Cu and Mn, there is no need to apply the corresponding fertilizers.
It should be pointed out that there is a certain relationship between soil fertility or IFI with the growth, yield and quality of tobacco [7,[55][56][57][58]; it is not evaluated in this study but would be conducted in our further research. , and the mean IFI was increased by 13.00% from 0.492 in 2000 to 0.556 in 2015, but both still belonged to the middle grade of IFI. Thus, soil fertility still needs to be promoted, and more attentions should be paid to modify of soil acidity and alkalinity, reduce the application of organic, nitrogen, calcium fertilizers, and increase the application of fertilizers of potassium, magnesium, molybdate and chloride according to the real situation of tobacco-planting fields.

Conclusion
Funding information: Authors state no funding involved.

Conflict of interest: Authors state no conflict of interest.
Data availability statement: Due to confidentiality agreements, supporting data can only be made available to bona fide researchers subject to a non-disclosure agreement. Details of the data and how to request access are available from Mrs. Yansong Xiao (Email: 35149517@qq.com, Mobile No: +86-18975717573) at Chenzhou Tobacco Company of Hunan Province.