Site Classification and Evaluation of Eucalyptus urophylla × Eucalyptus grandis Plantation in Southern Yunnan, China

Background and Objectives: The site types of Eucalyptus urophylla × Eucalyptus grandis clonal plantations in southern Yunnan were compared, aiming to provide basis for site selection and scientific plantations management. Materials and Methods: In this study, 80 standard plots were set up in the 6−9-year-old Eucalypts plantations in Pu'er City and Lincang City. Furthermore, the quantitative theory I model and canonical correlation analysis were used to analyze the relationship between dominant tree growth traits and site factors, and evaluate the growth potential of E. urophylla × E. grandis plantation. Results: The multiple correlation coefficient between 8 site factors (altitude, slope, slope level, soil thickness, slope direction, texture, soil bulk density, and litter thickness) and the quantitative growth of the dominant wood was 0.825 (P < 0.05). According to the correlation coefficient of the quantitative regression model, slope, altitude, and soil thickness were the main factors for the classification of E. urophylla × E. grandis plantations in southern Yunnan. In addition, E. urophylla × E. grandis plantations grew best downhill and mid uphill at relatively low altitude, where the soil layer was thick and composed of weathered red soil. Contrastingly, E. urophylla × E. grandis plantation growth was extremely poor in uphill sites at higher altitude, where the soil layer was thin and composed of semi-weathered purple soil. Furthermore, total N, and available B, Cu, and Zn content, as well as soil organic matter content in the soil had a great influence on the growth of E. urophylla × E. grandis. Conclusions: Nitrogen and phosphate fertilizer as well as trace elements such as B, Zn, and Cu can be properly applied in middleand low-yield forests to promote the growth and development of E. urophylla × E. grandis plantations.


Introduction
Eucalypts has become the most important industrial raw material forest tree species in southern China due to its advantages of fast-growing, high-yield, wide-ranging and short-rotation [1]. By 2018, the area of Eucalypts plantations in China had reached 5.647 million ha, accounting for 2.5% and 6.8% of the total and planted forest area of the country, respectively. Moreover, the annual average wood production was 30 million m 3 , i.e., 25% of the country's commercial forest products, providing an important guarantee for the production of domestic wood and fiber raw materials.
Since 2004, the National Development and Reform Commission has promoted the "Tenth Five Year Plan of national forestry paper integration project construction and 2010 special plan," and large domestic and foreign paper-making industries have built large-scale Eucalypts forests for fiber From November to December 2017, standard sites were set according to site factors such as altitude, slope, slope position, and soil layer thickness, among others. A total of 130 standard plots (20 × 20 m) were set. Among them, 80 were selected from standard plots 6−9-year-old to construct quantitative regression models, and the other 50 plots were used for model testing. Standard surveys were performed by GPS to record geographic coordinates, altitudes, and measure diameter at breast height (DBH), and tree height for each tree of E. urophylla × E. grandis. Soil samples were collected by digging three soil sections at equal distances in the diagonal direction. Among them, 0.5 kg of mixed soil along 0−60 cm was used to analyze chemical composition, and soil sampled by 100-cm 3 ring knife was used to measure soil bulk density and other physical properties.
The statistics of understory shrubs were recorded by setting five plots with an area of 2 × 2 m in the four corners and center of the standard plot. The highest single plant was selected as the dominant tree, one disk of 1.3 m was cut down at chest height, and a disk of 8−10 cm thick was cut at a length of 2 m to analyze the trunk. In addition, five cores with an average diameter of 5 mm were selected on the north and south side, with a core height of 1.3 m. The analysis of pulping test materials consisted on selecting a standard tree, removing the top 3 m and 1 m at the base, then slice the rest into 30−40-cm sections. The quantitative regression model formula was as follows [9]: where bik is an unknown coefficient and εt is a random error. The least square estimate bik for this is called the score of category k of item i, and the prediction equation was as follows: and the single timber volume of dominant trees was calculated as [10]: V = π × (DBH / 2) 2 × H / 30,000, where V represents the volume (m³) and H is the tree height (m).

Sample soil analysis
Soil analysis determined 11 indexes of soil, including pH, organic matter; total N, P, and K, as well as available N, P, K, B, Zn, and Cu. Soil pH was calculated with a soil:water mass ratio of 1:2.5 and measured by potentiometric method. Soil organic matter was measured by the potassium dichromate oxidation-capacity method. Total N was digested with H2SO4+ catalyst and measured by alkaline hydrolysis diffusion method, while P was determined by sodium hydroxide alkali melting-molybdenum antimony colorimetric method, and K was determined by sodium hydroxide alkali melting-atomic absorption spectrophotometer method. Available N, P and K were determined by the alkaline hydrolysis diffusion method, 0.05 mol L -1 HCl-0.025 mol L -1 H2SO4 extraction method, and 1 mol L -1 ammonium acetate extraction-atomic absorption spectrometry, respectively. Effective B was determined by curcumin colorimetry, and effective Cu and Zn were extracted by 0.1 mol L -1 hydrochloric acid and measured by atomic absorption spectrophotometry [11].

Data statistics and analysis methods
Quantitative theory I model was used to list each site factor by (0,1) and SPSS 13.0 for Windows (SPSS Inc., New York, NY, USA) was used for quantitative regression analysis. In addition, CCA of the dominant E. urophylla × E. grandis tree height and its single timber volume relative to soil chemical properties were performed in R Studio. Soil physical and chemical properties and the trends of productivity grade of E. urophylla × E. grandis were plotted in Excel 2016 (Microsoft, USA).

Classification of site factors
Standard site factors and high average annual growth of dominant wood were listed in Table 1. The standard plot classification was performed (Table 2) according to the changes in site factors and growth factors of forest trees, combined with the principles and methods of site elements for classification and evaluation of forest sites.

Building a quantitative regression model
By combining site factor classification with standard survey information (Table 2), each site factor of the plot was listed in (0,1). Quantitative theory I was used for site index modeling, and the site factor categories were obtained by fitting regression coefficients and related parameters of annual average height growth of the dominant species, i.e., E. urophylla × E. grandis (Table 3).
According to the scores of each site factor, slope position > altitude > soil layer thickness > soil texture > soil bulk density > slope > slope direction > litter thickness (Table 3), and the proportion of the first three site factor scores was 61.04%. Among site factors, the effect of slope position and altitude on the growth of E. urophylla × E. grandis plantation was stronger than that of slope and slope-exposure. The results of t-test showed that the partial correlation between the eight site factors and the annual average height increment of E. urophylla × E. grandis was significant (P < 0.05). Furthermore, the regression model complex correlation coefficient was R = 0.825 and statistically significant (F(8,71) = 24.137 > F0.05(8,71) = 2.072). Thus, the eight selected site factors were closely related to the average annual growth of dominant trees, and this equation is theoretically reliable to evaluate the site quality of E. urophylla × E. grandis plantations. Thick：>10.0 cm Note:** Statistical significance at P < 0.01,* Statistical significance at P <0.05.

Site Classification and evaluation
Eighty plots were divided into 13 different site types (Table 4). Among the four site types on upper slope, the average site index value of "upper slope, medium-low altitude, medium-thick soil layer" was 3.61 m, while that of the "upper slope, low elevation, thin soil layer" was 3.01 m. Among the four site types of the middle slope position, the average site index value of the "middle slope, medium-low altitude, medium-thick soil layer" was 3.41 m, while that of the "middle slope, medium-low altitude, thin soil layer" was 2.77 m. Among the four site types of the lower slope, the average site index value on the "lower slope, low elevation, medium-thick soil layer" was 4.02 m, while that of "lower slope, medium elevation, middle-thick soil layer" was 3.04 m. In the flat site type, the average site index of the "flat slope, medium-high altitude, medium soil layer" was 3.16 m.
Regarding the average site index of E. urophylla × E. grandis at different slopes, lower slope > medium slope > upper slope > flat ground, and in the same slope, the average site index of "medium-low altitude, medium-thick soil layer" was higher than that of "medium-high altitude, thin soil layer." E. urophylla × E. grandis enjoyed a warm climate, and site productivity level was the highest at the "lower slope, medium-low altitude, thick soil layer." Note: "*" indicates that the soil layer covers three levels of thickness: thin, medium and thick.
Following the site factor regression model (Table 5), and the principle that the model's predicted value was close to the measured value, sample plots were divided into three productivity levels (high-, medium-, and low-yield), and each level was divided into thirteen site types in conformity with the difference in slope, altitude, and soil thickness.
The high-yield group was mainly located downhill, where the soil layer was medium-thick and the elevation was 1,000−1,400 m. The average annual height growth, DBH, and accumulation growth of E. urophylla × E. grandis were 2.71 m, 2.11 cm, and 32.84 m³ hm -2 , respectively.
The middle-yield group was mainly located in the middle and upper slope, where the soil layer was not that thick, and its elevation was 900−1,800 m. The average annual height growth, DBH, and accumulation growth of E. urophylla × E. grandis were 2.32 m, 1.93 cm, and 23.5 m³ hm -2 , respectively.
The low-yield group was mainly located on the upper slope, where the soil layer was thin and the elevation was 1,000−1,600 m. The average annual height growth, DBH, and accumulation growth of E. urophylla × E. grandis were 1.98 m, 1.82 cm, and 17.94 m³ hm -2 , respectively.

Relationship between the growth of E. urophylla × E. grandis and soil chemical properties
To explore the effect of soil chemical properties on the growth of E. urophylla × E. grandis, the tree height and single timber volume of dominant trees , and soil chemical properties were used as two sets of variables for CCA (Table 6). From the scatter plot (Figure 1), it can be seen that the first and second canonical variables of E. urophylla × E. grandis were distributed near a straight line, and the statistically significant correlation coefficients were 0.8045 and 0.8263, respectively.
In the first canonical variable, owing to the analysis of soil chemical properties, the load of the total N content was 0.1362. The remaining five factors were available B content, organic matter content, available Zn, Cu, and P content in turn, and the loads were 0.0961, 0.0924, 0.0858, 0.0673, and 0.0582, respectively.Regarding growth parameters, the trait with the maximum load was the height of dominant trees followed by the single timber volume of dominant trees, with loads of 0.1393 and 0.0605, respectively.
In the second canonical variable, regarding soil chemical properties, the maximum load was N content (0.0902). The remaining five factors were effective B and Zn content, organic matter content, and effective Cu and P content in turn, with loads of 0.0752, 0.0671, 0.0528, 0.0503, and 0.0459, respectively. Regarding growth parameters, the trait with the maximum load was the height of dominant trees (0.1046), followed by the single timber volume of dominant trees (0.0857).
Results indicated that during the growth process of E. urophylla × E. grandis, total N content, trace elements B, Zn, and Cu content, and organic matter content had the greatest influence on the tree height and single timber volume of dominant trees of E. urophylla × E. grandis.

Relationship between soil chemical properties and productivity of E. urophylla × E. grandis
CCA results of E. urophylla × E. grandis growth and soil chemical properties, the effects of the first six factors (total N content, effective B content, soil organic matter content, and effective Zn, Cu, and P content) had a great impact on the growth of E. urophylla × E. grandis. As can be seen in Figure  2, through the analysis of all soil chemical factors and E. urophylla × E. grandis productivity, the total N content of soil in the high-yield group was 8.0% and 28.6% higher than that in the middle-and low-yield group, respectively. The content of soil alkalinized N in the high-yield group was 11.1% and 42.9% higher than that in the middle-and low-yield groups, respectively. Soil effective P content of the high-yield group was 50.2% and 94.4% higher than that of the middle-and low-yield group, respectively. Total K content of the soil in the high-and middle-yield groups was slightly different, but it was 20.2% and 23.5% higher than that in the low-yield group, respectively. The soil available K content in the high-and middle-yield groups were 14.8% and 13.7% higher than that of the low-yield group, respectively. The effective B content of the high-yield group was 29.4% and 57.1% higher than that of the middle-and low-yield groups, respectively. The effective Zn content of the high-yield group was 25.3% and 59.3% higher than that in the middle-and low-yield groups, respectively. The effective Cu content of the high-yield group was 23.4% and 61.2% higher than that of the middleand low-yield groups, respectively. The soil organic matter content of the high-yield group was 16.5% and 90.9% higher than that of the middle-and low-yield groups, and that of the middle-yield group was 64% higher than that of the low-yield group.

Differences between site types of E. urophylla × E. grandis in southern Yunnan and other regions
Since the 1950s, Leizhou Forestry Bureau has had more than 60 years of Eucalyptus cultivation history, with several varieties planted such as Eucalyptus exserta, Eucalyptus leizhou No.1, Eucalyptus urophylla, Eucalyptus ABL No.12, Eucalyptus urophylla U6 clone, and TH and LH series clones of Eucalyptus urophylla × Eucalyptus tereticornis , etc. Zhang et al. [12] conducted a site index model study on Eucalyptus ABL No. 12 (W5 clone) . Zhen et al. [13,14] adopted successive generations of continuous crop farming, and found that soil organic matter decreased significantly. The average single plant biomass decreased with the increase of continuous cropping [15], and there was a fertility decline phenomenon. At the same time, summer typhoons were frequent in the coastal areas of southern China, which was a major threat to the growth and cultivation of Eucalypts [16]. The survey period was relatively short, approximately 4 to 5 years, so it is difficult to classify site types in the Leizhou Peninsula, and Eucalypts cultivation in southern Yunnan has developed since the 1990s [17]. Since 2001, Yunnan Yunjing Forestry and Pulp Mill Co., Ltd and Leizhou Forestry Bureau have carried on the cooperative introduction and scale planting DH clones of E. urophylla × E. grandis. As this is relatively new, reforestation has not yet started. Therefore, the stand of the raw forest belongs to the first generation forest and first generation with 2−3 times sprouting forest, with no decline in soil fertility caused by continuous planting of multiple generations. In addition, there are no meteorological factors such as typhoons, and the survey period is mainly from 6 to 9 years, which can better reflect the site productivity of forest land. The excellent clones represented by E. urophylla × E. grandis DH clones were screened out, and their average annual accumulated growth was more than 30 m 3 hm -2 , which basically was in accordance with the site production potential of this study [18]. Hence, the reliability of site type division of E. urophylla × E. grandis plantations in southern Yunnan has been further improved. However, the paper-wood properties and pulp yield of E. urophylla × E. grandis plantations will be analyzed in a later stage.

Main impact factors on site types of E. urophylla × E. grandis in southern Yunnan
Site type division and evaluation are the basis of scientific afforestation, and particularly important to evaluate the potential of site production [19]. Slope position, elevation, and slope direction are the dominant factors for site type classification of birch plantations in the southwestern Daqing Mountains of Guangxi [20]. Meanwhile, soil nutrients, density, and texture are the dominant factors for the site types of teak plantations in Xishuangbanna, Yunnan [21], while slope position, altitude, and soil thickness are the dominant factors for the site types of E. urophylla × E. grandis plantations in southern Yunnan. Moreover, the growth of Eucalypts is greatly affected by the soil physical and chemical properties at different slopes. The soil layer downhill was more loose than that of the middle and upper slopes, and the ventilation, water permeability and conservation, as well as fertilizer retention ability downhill are better than those uphill, which is beneficial to promote root growth and normal development [22]. With decreasing slope, the soil layer thickness and soil moisture content have an increasing trend, which is related to the physical and chemical properties of the soil, and has a great impact on soil nutrition [23]. Furthermore, the conservation rate of Eucalypts plantations is little affected by different slope levels, but the annual average height and single timber volume of dominant trees increase with the decrease of slope position. In this study, the maximum value of annual average growth of the dominant trees at the lower slope was 4.02 m, while that at the top of the hill was 3.16 m. Such differences may be caused by the presence of certain shade between stands affecting photosynthesis or associated with drainage factors [24].
Additionally, some studies have shown that the biomass of Eucalypts plantation is significantly negatively correlated with the increase of elevation gradient, which changes the hydrothermal conditions, affecting the growth and development of stand, and reducing the biomass [25−28]. In this study, the elevation of E. urophylla × E. grandis in the high-yield group was generally below 1,400 m, while the altitude of the middle-and low-yield groups was more than 1,400 m. Therefore, at higher altitudes, sites were less suitable for the growth of E. urophylla × E. grandis.

Physical and chemical properties of soil and productivity
Nitrogen fertilizer can increase plant chlorophyll content and photosynthetic rate, and promote the growth of plant stems and leaves [29]. B, Zn, and Cu are essential trace elements for plant growth and development, and the lack of three trace elements is extremely detrimental to the growth of young Eucalypts forests and it will inhibit growth and cause physiological diseases [30−32]. When elements such as B, Zn, Cu, and Mn were applied alone, the production of Eucalyptus ABL No. 12 (W5 clone) and Eucalyptus urophylla exceeded that of unfertilized sites in three years [33]. B participates in the metabolism and transportation of carbohydrates in plants, and promotes the growth and development of the root system. B deficiency will lead to thin, curly, dark, and dull leaves, and the tip growth will be inhibited or withered shoots and branches will be produced. The lack of Zn results in shortened node spacing, smaller leaves, clusters of tender shoots, reddish leaves and their edges, or withered leaves, while Cu deficiency causes deformed leaves and young shoots. Furthermore, the texture of the leaves will become thicker, crisper, and dull or produce tip shoot [34].
Moreover, soil organic matter content represents soil fertility, and is an important source of various nutrients such as nitrogen, phosphorus, and trace elements required by plants. Increasing organic matter content plays an important role in high and stable yield of Eucalypts trees, and thus forest litter should be kept as much as possible [35]. Total N in soil has a great impact on the productivity level of Eucalyptus grandis, and ethanol deficiency can inhibit the growth of above-ground parts and roots of Eucalyptus, affect chlorophyll synthesis, cause leaf yellowing, and may cause spot blight [36].
According to our results and the soil nutrient grading standards of the second national soil census [37] (Table 7). the average content of soil organic matter and total N in the three productivity levels of E. urophylla × E. grandis plantations in southern Yunnan were 8.41−16.06 and 0.63−0.81 g kg -1 respectively, both belong to lower middle level; Available P was 1.12−2.13 mg kg -1 , belonging to extremely level; Available K was 117.3−138.4 mg kg -1 , belonging to upper middle level; Available B was 0.14−0.22 mg kg -1 , between extremely low and low level; Available Zn was 0.59−0.91 mg kg -1 , belonging to upper middle level, and available Cu was 0.49−0.79 mg kg -1 , belonging to upper middle level. Available Zn and available Cu were 0.59−0.91 and 0.49−0.79 mg kg -1 respectively, both belong to upper middle level. Therefore, for medium-and low-yield groups, N and P fertilizer should be appropriately added, and at the same time, trace elements such as B, Zn, and Cu should be added to promote the growth and development of forests. The logging operations should keep the dead branches, fallen leaves, and harvested residues in the forest land to increase soil organic matter, thus providing a guarantee for later nutrient recycling. Note: Ⅰ represents extremely high level; Ⅱ represents high level; Ⅲ represents upper middle level; Ⅳ represents lower middle level; Ⅴ represents low level; Ⅵ represents extremely low level.
Additionally, the soil of high-yield group was composed of red soil developed by the weathering of argillaceous rocks, which was intact and deep, thus ensuring that fertility lasted for a long time and Eucalypts grew vigorously. Therefore, proper tending and scientific management can give full play to the potential of site production, which is the best site for cultivation of E. urophylla × E. grandis with high pulp ratio and large diameter plywood. The soil of middle-yield group was mainly composed of lateritic red soil and natural red soil with deep development, i.e., good site conditions and production potential. Therefore, it is an ideal site to cultivate E. urophylla × E. grandis with high pulp ratio under proper management. The soil of low-yield group was mainly composed of purple sandy shale semi-weathered purple soil, where the biological accumulation effect was weak, soil development was not deep, water retention ability was low, and the site condition was poor. Thus, it is not suitable to cultivate fast-growing and high-yielding Eucalypts. However, it can be reformed to cultivate native tree species such as Pinus kesiya var. langbianensis, so as to avoid great loss.

Conclusion
Slope position, soil bulk density, and altitude were the dominant factors for the site types of E. urophylla × E. grandis plantations in southern Yunnan. E. urophylla × E. grandis grew best in the downhill, with relatively low elevation and thick soil layer composed of red soil developed by weathering of argillaceous rocks. E. urophylla × E. grandis grew relatively well on the middle and upper slopes, with slightly higher elevation and moderate soil bulk density, composed of red soil formed by the development of sandstone and shale and sandstone. E. urophylla × E. grandis grew comparatively less gently on the uphill and hilltops, with higher elevation and thinner soil layer than those of other sites, composed of purple semi-weathered soil of purple sandy shale rocks. Besides, total N, available B, soil organic matter, available Zn and Cu, and available P content had a significant effect on the growth of E. urophylla × E. grandis. However, the soils of this study were deficient in organic matter, total N, available P, B, Zn, K, and Cu. Hence, according to the above-mentioned site conditions and to achieve the full production potential of the forest stand, the selection of forest land should be combined with the actual situation, and a scientific and reasonable management mode should be adopted to create a broader prospect for the development of E. urophylla × E. grandis plantation in southern Yunnan.