Response of soil pH to biochar application in farmland across China: a meta-analysis

Data sources

The Web of Science (https://www.webofscience.com/) and CNKI (https://www.cnki.net/) databases were systematically searched for relevant studies published from January 2010 to July 2022. The keywords used for the search were (biochar OR black carbon OR charcoal) AND (soil pH). Jia Yao and Shixiang Zhao evaluated the title and abstract of each study to determine whether it contained original data on the response variable used in the present study. In case of disagreement, seek the help of other co-authors to determine the primary referee by vote. The selection criteria were as follows: (1) the object of the study was farmland soil in China; (2) the design of the study was field experiment, plot experiment, or pot experiment; (3) the experiment consisted of at least one treatment with biochar and one treatment without biochar (control), and other field conditions were consistent; (4) the basic physicochemical properties of soil and biochar samples were reported; and (5) each treatment was replicated at least three times. Exclusion criteria: (1) The research object does not match: excluding the study of non-farmland soil (such as forest soil, grassland soil, etc.); (2) Research design does not match: excluding laboratory simulation tests, model simulation studies or the lack of control group studies; (3) Incomplete data: Excluding studies that did not report the basic physical and chemical properties of soil or biochar, or studies with incomplete data that could not be effectively analyzed; (4) Insufficient repetition: studies with less than three repetitions per treatment were excluded.

A total of 104 valid research articles were retrieved and 598 sets of data were collected. The flow chart of literature removal is shown in Fig. 1. The province and latitude and longitude of the soil sampling points for each test are listed in File S2. We recorded the observed data of the response variable—soil pH—in each article, including the mean and standard deviation (SD) or standard error (SE) of control and biochar treatments. In addition to directly extracting tabular data, we digitized figure data using GetData version 2.25 (Getdata Pty Ltd, Manhattan Beach, CA, USA). Moreover, the data of the original soil pH and biochar characteristics (feedstock, pyrolysis temperature, application rate, pH) were recorded. All data were standardized appropriately before further analysis. The biochar application rate was converted into a unified unit (%). If SEs were reported in a selected article, they was converted to SDs using Eq. (1). For articles that did not report SDs or SEs, SDs were estimated to be 10% of the means (Luo, Hui & Zhang, 2006).

(1) $$SD=SE\sqrt{n}.$$

The data were categorized by soil pH level and biochar characteristics as follows. Based on their original pH, soils were divided into three groups: acid (pH < 6.5), neutral (pH 6.5 = 7.5), and alkaline (pH > 7.5). Biochar feedstocks were also classified into three types: crop straw (e.g., wheat husk, rice straw, corn cob), woody matter (e.g., timbers, leaves, branches), and others (e.g., sewage sludge, livestock and poultry manure, municipal solid waste). Li & Chen (2018) found that the physical and chemical properties of high temperature biochar ≥600 °C and ≤400 °C biochar are different, so according the biomass pyrolysis temperatures were ranked into three levels: low (200–400 °C), medium (400–600 °C), and high (600–800 °C). Biochar application rates were grouped into four levels: low (<1%), medium (1–5%), high (5–10%), and extremely high (10–20%).

Data analysis

In the meta-analysis, the treatments with and without biochar addition were defined as the treatment and control groups, respectively. We computed effect sizes as natural log-transformed response ratios (lnR):

(2) $$lnR=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{X_t}{X_c}}\right)=\mathrm{l}\mathrm{n}{X}_t-\mathrm{l}\mathrm{n}{X}_c$$where X_t and X_c are the mean responses of soil pH in the treatment and control groups, respectively. The variance (v) of response ratio was estimated as follows:

(3) $$v={\displaystyle \frac{S_t^2}{n_t{X}_t^2}+{\displaystyle \frac{S_c^2}{n_c{X}_c^2}}}$$where n_t and n_c are the sample sizes for the treatment and control groups, respectively; S_t and S_c are the SD values for the treatment and control groups, respectively.

We conducted weighted meta-analysis, where the mean effect size of each categorical variable was calculated using a fixed effects model. Groups with less than two treatments were omitted from the analysis. Metawin version 2.0 (Rosenberg, Adams & Gurevitch, 2000) was adopted to calculate the overall mean effect size and 95% confidence interval (CI) for each categorical variable. If the 95% CI did not overlap with zero, the mean effect size was considered significant; X_t and X_c were considered significantly different if the 95% CI did not overlap from one another. One-way analysis of variance with Tukey’s test was used to determine whether there were significant differences in soil pH among different groups or levels of categorical variables. Publication bias is one of the most concerned and studied bias in Meta-analysis. In order to test the publication bias of the collected literature and ensure the accuracy of the evaluation results and conclusions of meta-analysis, this study uses Eggar’s method to quantify. When the bias test value P > 0.05, it is considered that there is no publication bias in the collected literature, and vice versa (Egger et al., 1997). Statistical analyses were conducted using OriginPro version 2022 (OriginLab Corp., Northampton, MA, USA).

Total effect of biochar on soil pH

Averaged across the entire dataset, biochar addition significantly increased soil pH in the samples, with an effect value of 0.0450 (Fig. 2). When separating the samples into acidic, neutral, and alkaline soils, biochar exhibited a positive effect on the pH of acidic soils (0.0662), and this effect was greater than that for neutral (0.0136) and alkaline soils (0.0054).

Effect of biochar from different feedstocks on soil pH

The addition of biochar derived from various feedstocks displayed a positive effect on soil pH (Fig. 3). The response of soil pH to straw biochar was greater in acid soils (0.0801) than in neutral and alkaline soils (0.0135 and 0.0033, respectively). Woody biochar exhibited a similar effect to straw biochar in both acid and alkaline soils. However, other biochar displayed a greater effect in neutral soils, with no significant difference between acid and alkaline soils. Among the three feedstock types, woody biochar and straw biochar had prominent effects on acid soils (0.0820 and 0.0801, respectively), which accounted for a significant increase in soil pH. On the contrary, other biochar had the greatest effect on neutral (0.0498) and alkaline (0.0441) soils.

Effect of biochar with different pyrolysis temperatures on soil pH

Upon addition of biochar prepared at 200–400 °C, the pH of both acid and alkaline soils responded positively, and the effect on acid soils was greater than that on alkaline soils (Fig. 4). The addition of biochar prepared at 400–600 °C exhibited a positive effect on soil pH in all three groups, and the greatest effect emerged in acid soils. Notably, adding the biochar prepared at 600–800 °C exerted a negative effect on the pH of alkaline soils (−0.0238), which contrasted with its positive effect on the pH of neutral and acid soils. Among the three temperature levels, the positive effect of high-temperature biochar was the most remarkable on acid (0.1008) and neutral (0.0517) soils. However, for alkaline soils, the greatest negative effect was attributed to high-temperature biochar (–0.014), whereas the most evident positive effect was posed by low-temperature biochar (0.0270).

Effect of biochar applied at different rates on soil pH

The effects of different biochar application rates on soil pH were divergent (Fig. 5). Biochar addition at low rates <1% mainly positively affected the pH of acid (0.0319) and neutral (0.0223) soils, with a minimal negative effect on alkaline soils (–0.0017). The addition of 1–5% and 5–10% biochar had an exclusively positive effect on soil pH over a range of acidity, neutrality, and alkalinity. The positive effect of biochar was generally enhanced with increasing application rate, and adding 10–20% biochar specifically affected the pH of acid and neutral soils. Among the three levels of application rate, 5–10% biochar exhibited the greatest positive effect on acid soils (0.1568), whereas 10–20% and 1–5% biochar respectively showed the most prominent effect on neutral (0.0643) and alkaline (0.0167) soils.

Interaction effects of biochar feedstock and pyrolysis temperature on soil pH

Biochar feedstock and pyrolysis temperature showed varied interaction effects on soil pH (Fig. 6). The responses of pH in acid and neutral soils to straw biochar with different pyrolysis temperatures (Fig. 6A) were consistent with the response patterns to all biochar (Fig. 4). In the case of alkaline soils, straw biochar prepared at 200–400 °C and 400–600 °C had a slight positive effect on soil pH (0.0003 and 0.0075, respectively), whereas a negative effect was observed for straw biochar prepared at 600–800 °C (–0.0250).

The positive effect of woody biochar on the pH of acid soils was heightened with increasing pyrolysis temperature. In alkaline soils, woody biochar showed contrasting effects on soil pH depending on pyrolysis temperature, as indicated by a positive effect value at 400–600 °C (0.0210) and a negative effect value at 600–800 °C (–0.0238; Fig. 6B). The effects of other biochar prepared at different temperatures were positive across all soil groups (Fig. 6C).

Interaction effect of biochar feedstock and application rate on soil pH

There were distinct interaction effects of biochar feedstock and application rate on soil pH (Fig. 7). Specifically, different straw biochar rates had a positive effect on the pH of acid and neutral soils (Fig. 7A), and the response of pH in acid soils was consistent with the pattern observed for all biochar (Fig. 5). In the case of alkaline soils, adding 1–5% straw biochar had a small positive effect (0.0105) on soil pH, whereas lower and higher straw biochar rates had a minor negative effect on soil pH (–0.0022 and –0.0019, respectively; Fig. 7A).

The response of pH in acid soils to different woody biochar rates was also consistent with the pattern observed for all biochar (Fig. 5). In alkaline soils, the effect of woody biochar rate on soil pH resembled that of straw biochar rate. When woody biochar was added at rates of 1–5%, there was a positive effect on soil pH (0.0263). Conversely, a negative effect emerged with woody biochar addition at lower and higher rates (–0.0001 and –0.0259, respectively; Fig. 7B).

The addition of other biochar at different rates had consistent positive effects on the pH of acid, neutral, and alkaline soils (Fig. 7C). Together, the results indicate that greatest positive effect on soil pH was observed in acid and neutral soils upon straw biochar addition at high (5–10%) and extremely high (10–20%) rates, respectively. The greatest negative effect observed in alkaline soils was attributed to woody biochar addition at high rates.

Interaction effect of biochar feedstock, pyrolysis temperature, and application rate on soil pH

Regardless of the application rate, the addition of straw biochar prepared at 200–400 °C exhibited a positive effect on the pH of acid soils (0.0767). However, when this biochar was applied between rates of 1–5%, a negative effect on soil pH emerged in alkaline soils (–0.0178). The effect of straw biochar prepared at 600–800 °C was positive across all soils, being the strongest in acid soils at the application rate of 5–10%. Woody biochar prepared at 400–600 °C also positively affected the pH of both acid and alkaline soils, and this effect was strengthened in acid soils with increasing application rate. The addition of other biochar with different pyrolysis temperatures had a positive effect on the pH of acid, neutral, and alkaline soils, and there was an upward trend in this effect with increasing application rate (Fig. 8C).

Among the different treatments, the pH of acid soils was positively affected the most by addition of 5–10% straw biochar with high temperatures (600–800 °C; 0.2062). The same application rate of medium-temperature straw biochar (400–600 °C) was most effective in increasing the pH of neutral soils (0.0713). In alkaline soils, however, the greatest positive response of pH was observed upon addition of 5–10% other biochar prepared at 200–400 °C (0.0742), whereas the addition of 1–5% straw biochar prepared at 200–400 °C had a negative effect on alkaline soils (−0.0178).

Limitations

SE to SD depends on the accuracy of the sample size n. If the original study does not report n or there is a selective report, it may lead to SD being overestimated or underestimated, thus distorting the calculation of effect size. Assume that SD = 10% mean will be divorced from reality. SD reflects the inherent dispersion of data and has no fixed proportional relationship with the mean (Fig. 9). Therefore, using these two methods to estimate SD may cause errors in the results. In the future research, the original data should be obtained first, and the advanced statistical methods such as maximum likelihood estimation should be used to deal with the missing SD, and the influence direction of different SD assumptions on the results should be tested in the sensitivity analysis.

The data of this study focused on short-term experiments (<3 years), which could not reflect the long-term effect of biochar aging (such as oxidation, microbial degradation) on pH (Lehmann et al., 2021); the regulation of soil microbial communities (acid-producing bacteria and nitrifying bacteria) on biochar effect was not quantified. The sample size in some areas (alkaline soil in the arid region of Northwest China) was insufficient (File S2), which may affect the universality of the conclusion. It is suggested that future research should combine long-term positioning experiments and metagenomic techniques to reveal the interaction mechanism of biochar-microorganism-soil chemistry.