Impacts of forest age on soil characteristics and fertility quality of Populus simonii shelter forest at the southern edge of the Horqin Sandy Land, China

The study area

The study area of this study is located in Naiman County, Tongliao City, in the Inner Mongolia Autonomous Region of China and physically located at the southern edge of the Horqin Sandy Land in China (120°19′40″–121°35′40″E, 42°14′40″–43°32′20″N). The area is 8,137.6 km. In the 1970s, 62% of the total study area was sandy soil. The government began large-scale afforestation measures in the Horqin Sandy Land. The area has now reached 30.45% forest cover. The climate of the study area is northern temperate and arid with a continental monsoon. Elevation ranges from 250 to 570 m. The average annual temperature is 6.5–7.7 C. The average annual wind speed is 3.6–4.1m/s. The average annual precipitation is 274–344 mm. The average annual evaporation is 1,817.4–2,032.6 mm and is about six times the average precipitation. According to the FAO (1988), soil texture in the sampling area is mainly Arenosols. The main herbaceous vegetation under the Populus simonii shelter forests is Lespedeza davurica, Setaria viridis, Salsola collina, and Bassia scoparia. The study area is shown in Fig. 1.

Study methods

Soil sample collection and determination

Soil samples were collected from August 23rd–27th, 2022. Three soil sampling points were randomly selected on the diagonal of the sample plot and a soil pit with a depth of 60 cm was dug out at the sampling point. Samples were taken from the bottom to the top of the soil profile in three layers (0–20, 20–40, 40–60 cm). The sampling method is shown in Fig. 2. Soil water content, bulk density, field water holding capacity, and porosity were collected from undisturbed soil with a 100 cm³ ring knife. Samples were dried at 105 °C for 12 h and analyzed using the gravimetric method. Samples to determine the soil chemical properties were brought back to the laboratory in sealed bags and were air-dried in the absence of sunlight, All soil was divided into two portions; one portion was passed through a 1 mm soil sieve in order to determine the available nutrients content, the other portion was passed through a 0.25 mm soil sieve in order to determine the total nutrients content. Soil pH was measured using the potentiometric method. Soil organic carbon was measured using the volumetric heating method with potassium dichromate. Available nitrogen was measured using the alkali-diffusion method. Available phosphorus was measured by 0.5 mol/L NaHCO₃-H₂SO₄ leaching using a molybdenum antimony anti-colourimetric assay. Available potassium was measured using the 1 mol/L CH₃COONH₄ leaching-flame photometric method. Total nitrogen was measured using the Kjeldahl method with a concentrated sulphuric acid-perchloric acid mixture. Total phosphorus was measured by concentrated sulphuric acid-perchloric acid mixed cooking-molybdenum antimony anti-colourimetric method. Total potassium was measured using the acid dissolution-flame photometric method. This experimental study covered five age classes. Each age class had three sample plots. Three layers were sampled in each sample plot. Three soil samples were collected in each soil layer. A total of 135 soil samples were collected.

Changes in soil bulk density and porosity

Figure 3 shows the changes in soil bulk density and porosity of Populus simonii shelter forests of different forest ages. The soil bulk density in general showed the pattern of change as NMF < MF < OMF < MAF < YF. Soil bulk density at different soil depths of the same age class increased at each soil depth so that 0–20 < 20–40 < 40–60 cm. There was a significant difference in soil bulk density among YF, MAF and NMF (p < 0.05) and there was no significant difference between MF, OMF and MAF, NMF (p < 0.05). This indicated that when the shelter forest grows beyond MF the forest age no longer has a significant effect on the soil bulk density. The changes in the soil non-capillary porosity, capillary porosity, and total porosity was opposite that of soil bulk density. Specifically, we determined that: NMF > MF > OMF > MAF > YF. The porosity of different soil depths of the same age class showed 0–20 > 20–40 > 40–60 cm. Soil total porosity was significantly different (p < 0.05) among the sample plots of all age classes, indicating that the growth and development of artificial vegetation as the forest aged had a significant effect on soil porosity.

Changes in soil water content and field water holding capacity

Changes in soil water content and field water holding capacity of Populus simonii shelter forests in different age classes are shown in Table 2. Soil water content and field water holding capacity in general both showed the pattern of change as NMF > MF > MAF > OMF > YF. Soil water content at different soil depths of the same age class varied as follows: 0–20 < 20–40 < 40–60 cm. Soil water content at the same age class was significantly different at all soil depths (p < 0.05). The variation of field water holding capacity in soil depth was 0–20 > 20–40 > 40–60 cm. The field water holding capacity of the 0–20 cm soil layer was significantly different from the water holding capacity at a soil layer depth of 20–40 and 40–60 cm (p < 0.05).

Changes in soil pH and organic carbon content

Table 3 shows the changes in soil pH and organic carbon in shelter forests of different forest ages. Soil pH gradually decreased as forest age increased and the soil pH of MF and OMF was significantly different from that of the YF, MAF, and NMF sample plots (P < 0.05). This indicates that the soil shifted from alkaline to neutral as the forest aged. The soil organic carbon content of different age classes varied as NMF > MF > OMF > MAF > YF and it decreased with increasing soil depth. In the YF and MAF classes, organic carbon was significantly different (P < 0.05) between the 0–20 cm soil depth and the 20–40 cm, whereas there was no significant difference (P > 0.05) between the organic carbon content of the 20–40 and 40–60 cm soil depths. There was a significant difference (P < 0.05) between the organic carbons in each soil depth of the NMF, MF and OMF. This indicates that the changes in soil organic carbon content in the early stage of forest growth and development were mainly concentrated in the surface soil layer and the differences in organic carbon content among soil depths were more obvious after NMF.

Changes in soil nutrient content

Soil nutrient content indicators of different age classes of Populus simonii shelter forests all peaked during NMF (Fig. 4). The soil total nitrogen, total phosphorus, and available phosphorus content varied as a whole with NMF > MF > OMF > MAF > YF. Soil total potassium, available nitrogen, and available potassium (0–20 cm soil depth) were NMF > MAF > MF > OMF > YF. Soil nutrient content all decreased with increasing soil depth. ANOVA showed that there was a significant difference in the distribution of soil nutrients among the three soil depths during the NMF (P < 0.05) whereas the soil nutrient content of the MF and the OMF gradually decreased and was closer to the nutrient level of the MAF. This may be related to the competition for water and nutrients caused by the increase in vegetation cover in the later stages of forest growth and development.

Evaluation of soil fertility quality

Constructing the minimum data set

A total of two principal components with eigenvalues ≥1 were extracted through PCA (Table 4), with eigenvalues of 9.983 and 2.079. Their cumulative variance contribution rate was 86.159%, indicating that the two extracted principal components are representative of the comprehensive properties of each indicator. The results of the Pearson correlation analysis of the indicators are shown in Fig. 5. In accordance with the requirement that the absolute value of the principal component loadings of the soil indicators must be greater than 0.5, the indicators used in the MDS were those with relatively high Norm values and relatively small correlation between the indicators. Soil total phosphorus with the highest Norm value was included in the MDS at PC-1. Available phosphorus (correlation coefficient <0.6 with total phosphorus), which had a relatively large Norm value and a relatively small correlation coefficient with total phosphorus, was included in the MDS. The correlation coefficient between pH and soil water content in PC-2 was <0.55 and both were included in the MDS. In summary, the final indicators selected for inclusion in the MDS were soil total phosphorus, available phosphorus, water content, and pH.

Table 4:

Results of principal component analysis of soil quality indicators.

Soil indicator	Symbol	Principal component		Norm value	Minimum date set
		PC-1	PC-2
Soil bulk density	SBD	−0.869	0.068	2.75
Soil noncapillary porosity	SNP	0.908	0.254	2.89
Soil capillary porosity	SCP	0.913	0.232	2.90
Soil total porosity	STP	0.945	0.251	3.01
Soil water content	SWC	−0.155	0.882	1.36	add in
Soil field capacity	SFC	0.898	0.163	2.85
Soil total nitrogen	TN	0.895	0.009	2.83
Soil total phosphorus	TP	0.954	0.074	3.02	add in
Soil total potassium	TK	0.886	−0.28	2.83
Soil organic carbon	SOC	0.936	0.246	2.98
Soil available nitrogen	AN	0.854	−0.183	2.71
Soil available phosphorus	AP	0.92	−0.306	2.94	add in
Soil available potassium	AK	0.913	−0.017	2.88
Soil pH	pH	0.26	−0.904	1.54	add in
Characteristic root		9.983	2.079
Variance contribution rates/%		71.308	14.850
Accumulated variance contribution rates/%		71.308	86.159

DOI: 10.7717/peerj.17512/table-4

Calculation of the SQI

The soil quality index was represented by TDS-SQI for the total data set and MDS-SQI for the minimum data set (Fig. 6A). The values of TDS-SQI ranged from 0.38 to 0.51 and the values of MDS-SQI ranged from 0.37 to 0.50. The pattern of change of TDS-SQI was NMF(0.51) > MAF(0.48) > MF(0.47) > OMF(0.44) > YF(0.38) and the pattern of change of MDS-SQI was NMF(0.50) > MAF (0.47) > MF(0.45) > OMF(0.44) > YF(0.37). The two evaluation results were consistent. By calculating the TDS-SQI and MDS-SQI at different soil depths (Fig. 6B), we see that the results of the soil fertility quality assessment at different soil depths are still NMF > MAF > MF > OMF > YF.

Reliability tests for minimum data sets

Regression analyses of the TDS-SQI versus the MDS-SQI were performed to verify the reliability of the MDS selection by means of scatter plots of the SQI for three different sites in the same age class. By assessing the fit of the linear regression analysis (Fig. 7) we see that TDS-SQI was significantly correlated with MDS-SQI (P < 0.05, R² = 0.8263). This indicates that the MDS established in this study, is a better alternative to TDS in responding to soil fertility quality in the study area.

Analysis of changes in soil characteristics indicators at different soil depths

In the 0–60 cm soil depth of the shelter forests studied in this article, soil bulk density and water content increased with soil depth. Soil porosity, field water holding capacity, organic carbon, pH, and nutrient content decreased with soil depth. Many scholars have observed similar patterns in their studies of forest age and soil properties in artificial and shelter forests (Jiang et al., 2013; Wu et al., 2010; Yesilonis et al., 2016). The low water content of the top soil layer of the Populus simonii shelter forests in all age classes is mainly due to the high wind speed and strong solar radiation in the Horqin Sandy Land; the top soil layer evaporates faster due to the wind, temperature, and humidity in the surface environment. On the one hand, elements and organic matter in the surface soil mainly come from the decomposition of litter, which increases as the forest ages. The surface humus and plant root growth have a compounding effect. As the porosity of the surface layer of the soil increases the soil bulk density decreases and the soil’s permeability, air permeability, and storage capacity increases. In this way the soil water-holding properties improve. On the other hand, the dense distribution of vegetation roots in the soil surface layer and the growth, extension, and aging of the root system have reduced the soil compactness and improved the soil porosity to a certain extent. The deeper soil is subjected to upper loading so the bulk of the soil is more compact in the deeper layer. This means the soil bulk density is greater in the deeper soil, meaning that soil nutrients are usually higher in the surface layer (Gao et al., 2021).

Analysis of changes in soil indicators and fertility quality at different forest ages

After afforestation in sandy soil, the growth of surface herbaceous vegetation, the accumulation of surface litter, and changes in the degree of cover over the years have led to corresponding changes in the microenvironment of the soil. With the increase in the number of years of afforestation, the distribution area of the vegetation root system in the soil has become wider and more active and the physico-chemical properties of the soil have changed accordingly. Litter accumulated on the soil leads to less direct sunlight on the soil and provides more thermal insulation. This reduces the evaporation of surface water and organic matter decomposition. The content of the organic matter and water in the soil, porosity, and bulk density is closely related to the soil organic matter itself, which has a loose porous structure. This not only promotes the formation of soil aggregates but also improves the soil aeration and water absorption capacity of the soil (Chen et al., 2021a). The physical properties of the sandy soil changed and major nutrient elements of the soil increased. This led to a replenishment of the accumulated nutrients in the topsoil in the original sandy soil (Gao & Huang, 2020; Jiang et al., 2013). Vegetation and soil play a complementary and positive role in the early stages of afforestation. However, when the age of the forest increases the canopy closure of the forest changes and results in an overly large canopy closure. This means that the underlying low vegetation does not receive a reasonable distribution of the ecological resources. Sunlight and rainfall are intercepted, which affects the photosynthesis and growth and development of the understory vegetation. In addition, the newborn seedlings that occur with natural succession do not get enough space to grow. Vegetation and individual trees compete for water and nutrients in the soil (Guo et al., 2022). Populus itself has a well-developed root system and a strong water-absorbing capacity. For the Horqin Sandy Land, which has an arid climate and infertile soil, the forest is too old which limits the understory biodiversity. Thus the soil quality begins to decline and the forest land begins to degrade.

Recommendations and outlook

As afforestation time increases, we see that the establishment of sand fixation forests strengthens the soil formation process of sandy soil and improves soil fertility. This makes the moving sand gradually become fixed or semi-fixed sand, alleviating the local dust storms. However, people often neglect the care and management of sand-fixation forests. The results of our study on all age groups of poplar shelterbelts show that sand shelterbelts also need to be nurtured and managed at the right time to be sustainable. It is recommended that the study area carries out the corresponding artificial care measures for Populus simonii shelter forests of more than 30 years of age and reasonably applies fertilizers to improve the quality of the forest land soil.

With the world’s increasing awareness of the importance of shelterbelts, more studies have been published on the subject. The top three countries for research on shelterbelts are China, the United States, and Canada (Mayrinck et al., 2019). The United States has extensive experience in reforesting sandy areas but their shelterbelts have long been subject to human interference (Ghimire et al., 2014). The type of shelterbelts in Canada are mainly farmland shelterbelts (Mayrinck et al., 2019) and according to the Government of Canada farmers in Western Canada have planted more than 600 million trees over the past century (Piwowar, Amichev & Van Rees, 2016). China is the country that has done the most research on shelterbelt forests in sandy areas but researchers have been more concerned with the effect of silvicultural density and the existing studies mainly focus on a few stand ages or a few soil indicators (Cao, 2008; Liu et al., 2023). We analyzed the soil properties of sandy shelterbelt forests in their natural state for all age groups. The fertility of soils at different growth stages of Populus simonii shelter forests when there was no continuous human interference or maintenance is clarified in this study. This is an important reference for the management of poplar shelterbelts in arid and semi-arid areas.