Soil responses to manipulated precipitation changes: A synthesis of meta-analyses

In the face of ongoing and projected precipitation changes, precipitation manipulation experiments (PMEs) have produced a wealth of data about the effects of precipitation changes on soils. In response, researchers have undertaken a number of synthetic efforts. Several meta-analyses have been conducted, each revealing new aspects of soil responses to 15 precipitation changes. We synthesize the findings of 16 meta-analyses focused on the effects of decreased and increased precipitation on 42 soil response variables, covering a wide range of soil processes and examining responses of individual variables as well as more integrative responses of carbon and nitrogen cycles. We found a strong agreement among metaanalyses that decreased and increased precipitation inhibits and promotes belowground carbon and nitrogen cycling, respectively, while microbial communities are relatively resistant to precipitation changes. Much attention has been paid to 20 fluxes and pools in carbon, nitrogen, and phosphorus cycles, such as gas emissions, soil carbon, soil phosphorus, extractable nitrogen ions, and biomass, but the rates of processes underlying these variables are less frequently covered in meta-analytic studies (e.g., rates of mineralization, fixation, and de/nitrification). Shifting scientific attention to these “processes” would, therefore, deepen the current understanding of the effects of precipitation changes on soil and provide new insights. By comparing meta-analyses focused on different variables, we provide here a quantitative and holistic view of soil responses to 25 changes in precipitation.

meta-analyses focused on different but complementary soil properties [e.g., soil C (Zhou et al., 2016) or N (Yue et al., 2019)]. A combined analysis of these meta-analyses would provide a holistic view of the potential effects of projected precipitation changes on soil processes.
In this paper, we synthesize 16 meta-analyses that have examined soil responses to manipulated (increased and decreased) 70 precipitation in-situ, encompassing 42 response variables including greenhouse gas exchanges, carbon and nitrogen dynamics, phosphorus content, microbial community, and enzyme activities. By collating the results of the published metaanalyses, we aimed to (1) provide a more holistic view of the effects of precipitation changes on soil composition and functioning, (2) discuss the potential underlying mechanisms of each response, and (3) identify knowledge gaps and propose future research directions. This study covers an unusually wide range of soil processes and examines the responses of 75 individual variables as well as nutrient cycles.

Meta-analysis collection
We collected peer-reviewed meta-analyses focused on the effects of decreased and/or increased precipitation on soil variables. We collected meta-analyses that included only field studies where the magnitude of precipitation was manipulated. 80 We used Google Scholar and Web of Science with the search terms "meta-analysis" AND "soil" AND ("respiration" OR "CO2" OR "carbon" OR "nutrient" OR "nitro" OR "phosph" OR "N2O" OR "CH4" OR "microb" OR "enzyme" OR "bacteria" OR "fungi") AND ("altered precipitation" OR "drought" OR "decreased precipitation" OR "increased precipitation" OR "water addition" OR "water reduction"). We identified 16 meta-analyses (Table 1); four of them focused on decreased precipitation (DP), one of them on increased precipitation (IP), and 11 on both DP and IP. A total of 42 soil 85 variables were covered, encompassing a wide range of soil characteristics such as soil greenhouse gas exchanges, soil carbon, nitrogen, phosphorous, microbial and bacterial communities, enzymes, and physical characteristics of soil (Table 2).
Only meta-analyses written in English and published before April 2019 were included in our analysis.

Effect sizes
From each meta-analysis, we obtained the mean effect size of each soil variable. In this review, effect sizes are the natural 90 log of response ratios (lnRR) defined as: where Xt and Xc are the mean values of the treatment (DP or IP) and control, respectively, for each observation. Homyak et al. (2017) used Hedge's d instead of Eq. 1 for N2O emissions and N supply due to the negativity of RR. Hedge's d is defined as J(Xt-Xc)/S where S is the pooled standard deviation, and J is the correction of small sample bias (Homyak et al., 2017). 95 https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License.
Both lnRR and Hedge's d are negative for inhibitory effects, and positive for stimulatory effects (Brzostek et al., 2012;Homyak et al., 2017). All meta-analyses calculated mean effect sizes and 95% confidence intervals (CI) with sample size or the inverse of the variance as the weighting function. The effect is considered significant when 95% CI does not overlap zero. Some meta-analyses applied additional weighting functions or normalized the measurements under different manipulation levels (Liu et al., 2016;Wu et al., 2011). We used these sample size-or variance-weighted effect sizes when 100 available. We obtained the values from the main texts or supplementary materials of the articles. If necessary, we used the digitizing software Plot Digitizer (Huwaldt, 2015), to extract values from graphs. When only percent changes were reported, we converted to lnRR as in Ren et al. (2017Ren et al. ( , 2018: (2) Some 95% CI were unavailable because points were not visible on graphs or because values of percent change below -100% 105 were not convertible using Eq. 2 (e.g. He and Dijkstra, 2014). We also obtained the sample size, defined as the number of studies or observations included in the meta-analyses. The collected information is available in Abbasi et al. (2020).

Responses of soil respiration and belowground biomass
Meta-analyses on autotrophic (Ra), heterotrophic (Rh), and total soil (Rs = Ra + Rh) respiration provide strong agreement that 110 DP decreases, and IP increases, Rs, Ra, and Rh (Fig. 1a). Litter biomass (B) follows the same pattern (Fig. 1b). Although the response of Ra reaches significance in only one of two meta-analyses, the direction of the response is consistent. Responses of soil carbon variables [total carbon (C), soil organic C (SOC), and dissolved organic C (DOC)] to precipitation differ among meta-analyses, both in direction and significance (Fig. 1b). Interestingly, root B is strongly suppressed by both DP and IP. In contrast, IP stimulates belowground B and belowground net primary productivity (NPP), and DP increases root C 115 ( Fig. 1b).
To understand the effects of precipitation on Rs, we need to understand the responses of roots, microbes, and substrates to DP and IP. Commonly, Ra decreases under limited water supply due to (1) reduced plant growth and nutrient demand, (2) reduced root tissue activity due to limited soil water, and (3) reduced respiratory substrate production from photosynthetic 120 activity (Hasibeder et al., 2015). In contrast, increased water supply enhances plant growth and photosynthetic rate (Heisler-White et al., 2008;Maire et al., 2015), which results in increased Ra. The responses of belowground B to IP and belowground NPP to DP and IP (Fig. 1b) are also likely to result from these changes in plant and root growth. However, not all belowground responses are consistent with this storyline. Root C and total C increase with DP, and root B decreases with IP ( Fig. 1b). 125 https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License.
This contradictory evidence could be due to variability by biome and soil type. For example, the effects of DP on total C is negative in temperate forests, and positive in tropical forests and grassland (Yuan et al., 2017;Zhou et al., 2016). Also, in our study, both significant effects (Zhou et al., 2016) and no effects (Liu et al., 2016) of DP and IP on Ra were found. This difference in effects could be attributed to, for example, the IP effects on Ra depending on biome and Ra separation method. 130 Specifically, IP effects can be significant in temperate forest and grassland, but not in boreal forest (Zhou et al., 2016), and Ra separated from Rh by clipping methods responded more positively than when trenching methods were used (Liu et al., 2016). Sample sizes remain relatively small for Ra responses to changes in precipitation, suggesting that additional research could help to identify how this process response varies with biomes and methods.

135
Rh is the consequence of soil microbial activity decomposing soil organic matter (SOM) under aerobic conditions. SOM is frequently estimated by measuring its carbon component, SOC. Rh is mainly regulated by microbial access to substrate and the physiological condition of microbes (Schimel, 2018). In dry soil, substrate tends to be isolated from microbes as solute mobility is low (Manzoni et al., 2012;Schimel, 2018). Furthermore, a great number of empirical observations and synthetic studies have shown that microbial activity is lower during droughts (Hueso et al., 2012;Jensen et al., 2003;Manzoni et al., 140 2012). This is because dry conditions force microbes into dormancy or shift their efforts from growth to survival (Salazar et al., 2018;Schimel et al., 2007). Excess water, on the other hand, not only increases substrate availability to microbes (Skopp et al., 1990), it also makes microbes dispose of osmolytes from their body cells to regulate the osmotic pressure (Schimel et al., 2007) and can activate dormant microbes (Salazar et al., 2018). These responses are particularly rapid and drastic when dry soils rewet, yielding a large pulse of respiration, which could significantly affect the carbon exchanges in terrestrial 145 ecosystems (Placella et al., 2012).
As with Ra, Rh typically decreases under DP and increases under IP, with variations among biomes and Rh separation methods. DP effects on Rh are significant in boreal forest and wetland, but not in tropical and temperate forests (Zhou et al., 2016). Likewise, IP effects on Rh are significant in forest and grassland, but not in wetland (Liu et al., 2016;Zhou et al., 150 2016). Furthermore, DP and IP effects are significant if the clipping method was used, but not with trenching and root extraction (Liu et al., 2016).
Microbial activity in soils is strongly controlled by the actions of enzymes. Many of these enzymes, which are produced and released by microbes, depolymerize complex carbon compounds (Ren et al., 2017). While enzyme activity is relatively 155 unresponsive to IP (Fig. 2), DP increases hydrolytic enzyme activity (breakdown of labile carbon) and inhibits oxidative activity (de-polymerization of recalcitrant carbon) (Fig. 2). This indicates that under dry conditions, the relative contributions of substrates from labile carbon sources increase, while the respective relative contributions from recalcitrant sources decrease.
The summary diagrams (Fig. 1c, 1d) illustrates how DP generally slows the belowground carbon cycle, while IP promotes it.
Nearly all steps of the carbon cycle -carbon stock, substrates, microbial activity, and respirationare altered by both types of precipitation changes. However, enzyme activity tends to be relatively unresponsive, particularly to IP, and the observations of biomass and carbon variables vary both in direction and significance among meta-analyses. These variables also tend to vary across biomes, ecosystems, and soil types. 165

Responses of methane uptake
We found only one meta-analysis that addressed the effects of precipitation on soil CH4 (Yan et al., 2018). The results show a significant increase and decrease of soil CH4 uptake in response to DP and IP, respectively (Fig. 1a). Soil CH4 fluxes involve two groups of microbes: methanogens and methanotrophs. Methanogens produce CH4 and are predominantly active in anaerobic conditions, while methanotrophs oxidize CH4 and are active in aerobic environments (Conrad, 2007). CH4 170 oxidation seems to peak at 10-15% volumetric water content because these conditions favor methanotroph activity as well as CH4 and O2 diffusion (Adamsen and King, 1993;Del Grosso et al., 2000).
The results of Yan et al. (2018) were significant across a wide range of ecosystem types, treatment durations, and magnitudes of precipitation manipulation. The effects of DP were greater in farmlands than other land types, in shorter-term 175 (< 1 year) experiments than longer-term ones, and in more extreme experiments (> 50% rain reduction). The effects of IP were greatest in boreal forest and in longer-term experiments (1-5 years) with greater rain addition (> 50%). However, a few empirical studies have shown opposite responses to this meta-analysis (Billings et al., 2000;Christiansen et al., 2015); for instance, a precipitation removal experiment in a floodplain decreased CH4 uptake, possibly due to the acclimation of methanotrophs to high soil moisture conditions (Billings et al., 2000), or differences in the types of methanotrophs in 180 floodplain (low-affinity methanotrophs) versus upland soil, where most CH4 uptake occurs (Christiansen et al., 2015).

Responses of soil nitrogen dynamics
Several soil nitrogen variables, including root nitrogen (N), N2O emissions, total N, dissolved organic nitrogen (DON), and extractable NH4 + + NO3are significantly affected by precipitation changes (Fig. 3a). Specifically, DP decreases root N and N2O emissions and increases total N, DON, and extractable NH4 + + NO3 -. We also found that two meta-analyses (sample 185 sizes < 20) suggest no change in total N, while one (sample size = 156) suggests an increase with DP. Similarly, one metaanalysis suggests an increase of extractable NH4 + with DP while other two meta-analyses suggest no effects. In contrast, IP increases root N, N2O emissions, and extractable NH4 + (Fig. 3a). Two meta-analyses suggest that total N decreases with IP, while one meta-analysis suggests no effects.
The input (nitrification) and outputs (denitrification, plant uptake and microbial assimilation) of NO3both decline under DP, leaving extractable NO3unchanged (Fig. 3b). 200 Extracellular enzyme activity, here shown both as total proteolytic activity (pro-enzyme) and three particular N-acquisition enzyme activities (β-1,4-N-acetyl-glucosaminidase, leucine amino peptidase, and urease), does not change with DP or IP ( Fig. 2). This indicates that the production of N-enzymes is not sensitive to water stress. Important outputs of the soil nitrogen cycle (denitrification and plant uptake) decrease while inputs remain constant or decline (Fig. 3b). As a result, total 205 soil N increases or remains unchanged.

215
Soil nitrogen undergoes a wide range of chemical and biological transformations, some of which are difficult to quantify.
Despite the large number of empirical studies included in meta-analyses, some nitrogen variables, such as rates of mineralization (for IP), nitrification, denitrification, and N2 fixation, have not yet been examined in meta-analyses focused on PMEs.

Responses of soil phosphorus 220
We found four meta-analyses that examined how precipitation changes affect the soil phosphorus (P) cycle (He and Dijkstra, 2014;Yan et al., 2018;Yuan et al., 2017;Yue et al., 2018). The results differ among meta-analyses; for instance, according to these meta-analyses, IP can have a negative, positive, or non-significant effects on total P (Fig. 4). Yuan et al. (2017) assembled the largest dataset and found that IP decreases total P, while DP increases total P. As phosphorus is commonly a https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License.
limiting nutrient for vegetation, plant P uptake and concentration are frequently studied, but studies of soil phosphorus 225 storage are rarer (He and Dijkstra, 2014;Yue et al., 2018).
Phosphorus in soil originates from weathering rocks, deposition from the atmosphere, and decomposition of organic matter (Wang et al., 2010). Outputs from soil involve plant uptake and consumption by microbes. As is the case with carbon and nitrogen, microbial decomposition and consumption activities of P can be affected by precipitation changes (Van Meeteren 230 et al., 2007). Plant P uptake tracks in the same direction as changes in precipitation (He and Dijkstra, 2014). However, challenges lie in generalizing the effects of precipitation changes on weathering and deposition, as these processes involve complex chemical and physical reactions. For example, soil water content determines the rate of chemical weathering, and humidity affects P deposition from the atmosphere (Newman, 1995). The effects on total P are strongly linked to soil type (Yuan et al. 2017). Although Yuan et al. (2017) found significant effects of DP and IP on total P, the effects were small (-0.1 235 < effect sizes < 0.1), and other meta-analyses show that soil P, as well as P-acquisition enzyme activity, are relatively unresponsive to precipitation changes (Fig. 2, 4). Other global changes such as warming, elevated CO2, and anthropogenic P and N deposition tend to have higher impacts on the terrestrial P cycle than precipitation changes (Yue et al., 2018).

Responses of microbial biomass and community structure
Microbial biomass (MB) in soil either decreases or does not respond to DP (Fig. 5a). MB responses to DP vary with the 240 amount of precipitation that is removed (Zhou et al., 2016;Ren et al., 2017Ren et al., , 2018, the length of droughts (Ren et al., 2018), vegetation type (Zhou et al., 2016;Ren et al., 2017Ren et al., , 2018 and mean annual precipitation (MAP; Ren et al., 2017). MB is affected by DP only when reduced precipitation is larger than ~33% (Ren et al., 2017(Ren et al., , 2018, the drought period is ≤ 2 years (Ren et al., 2018), and in wet (MAP > 600mm) regions (Ren et al., 2017). Additionally, vegetation type affects MB responses to DP; DP consistently decreases MB in forests (tropical and temperate but not in boreal; Zhou et al., 2016;Ren et 245 al., 2017Ren et 245 al., , 2018 and heathlands (Blankinship et al., 2011), but not in shrublands (Ren et al., 2017(Ren et al., , 2018. A meta-analysis conducted by Zhou et al. (2016) found that DP decreases MB in grassland soils. However, more recent meta-analyses that included more studies (Ren et al., 2017(Ren et al., , 2018 suggest that MB in grasslands does not respond to DP. In contrast, except when added precipitation is very high (> 70%; Ren et al., 2017), IP stimulates microbial growth and thus 250 increases MB. Contrary to DP, IP affects MB in dry (MAP < 600 mm) but not in wet (MAP > 600 mm) sites (Ren et al., 2017). This is consistent with IP increasing MB in soils from ecosystems that are generally water-stressed, such as deserts, shrublands, and grasslands (Zhou et al., 2016;Ren et al., 2017). Zhou et al. (2016) found that IP increases MB in soils in temperate forests. Other meta-analyses that included more studies (also including tropical forests) suggest that MB in forest soils is generally not affected by IP (Blankinship et al., 2011;Canarini et al., 2017;Ren et al., 2017). 255 https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License.
In contrast to the responsiveness of MB to altered precipitation, the composition of bacterial and fungal communities is rather unresponsive (Fig. 5b). Although Blankinship et al. (2011) and Yan et al. (2018) estimated significant effects on the abundance of fungi and F:B ratio (n = 4), other studies with sample sizes one order of magnitude larger (e.g., Ren et al. 2018) estimated non-significant effects. The high resistance of bacteria and fungi to soil moisture changes has been 260 frequently highlighted (Evans and Wallenstein, 2012;Schimel et al., 2007;Yuste et al., 2011). Fungi in particular, due to their filamentous structure, are capable of accessing substrates even in very dry soils (Manzoni et al., 2012). Bacteria and fungi also have a wide breadth of soil moisture niches; diverse types of bacteria and fungi tolerate water stress (Lennon et al., 2012). Differences in resistance between bacteria, fungi, and other functional types can alter microbial structure under precipitation changes; DP could promote a more fungi-dominated community (Yuste et al., 2011). Although gram-positive 265 bacteria are more resistant to soil moisture changes than gram-negative bacteria due to their thicker and stronger cell walls (Schimel et al., 2007;Salazar et al., 2019), both gram-positive and negative bacteria have been unresponsive to DP (Fig. 5b).
The sample sizes for bacteria and fungi in meta-analyses are small compared to MB meta-analyses (Fig. 5). Although an increase in the number of bacterial and fungal studies would improve our understanding of community responses to precipitation changes in terms of significance and magnitude of effects, current available data already covers a significant 270 range of locations and conditions, and highlights the clear trend of low responsiveness of bacterial and fungal communities to DP and IP manipulations.

Responses of belowground C:N:P stoichiometry
Belowground stoichiometric relationships of carbon, nitrogen, and phosphorus can help researchers interpret and infer nutrient movements in soil organisms and their environments. Yet, few meta-analyses have synthesized belowground 275 stoichiometric responses to precipitation treatments; greater attention has been paid to stoichiometry of aquatic systems and plants (Cleveland and Liptzin, 2007;Redfield, 1958;Yuan and Chen, 2015). He and Dijkstra (2014) and Yan et al. (2018) found no changes in soil C:N and N:P with DP (Fig. 3), but MBC:MBN responded to both precipitation changes (Fig. 5).
Increased MBC:MBN with IP indicates that wetter conditions stimulated greater metabolic activity of microbes, which accumulated more carbon in their bodies. This suggests that the soil microbial biomass C:N:P ratio, which is well-280 constrained globally (60:7:1) (Cleveland and Liptzin, 2007), could be altered by precipitation changes. Soil N:P ratios can be heavily dependent on plant nutrient uptake; as discussed in Sect. 3.3, DP reduces plant nitrogen uptake, which could increase soil N:P. However, this effect could be mitigated by strong mycorrhizal symbioses (Mariotte et al., 2017) and depend on site aridity (Sardans et al., 2012).

Knowledge gaps
Meta-analyses have substantially advanced our understanding of the impacts of precipitation changes on soil processes and properties. Specifically, a great number of meta-analyses have investigated soil respiration, nitrogen ions, total phosphorus, and microbial biomass. Nevertheless, there are still variables receiving less attention; for example, the sample size of autotrophic respiration is smaller than heterotrophic respiration, and substrate availability has not been analyzed while soil C, 290 N and P content have. CH4 fluxes have received less attention than CO2 and N2O, and no meta-analyses have examined the processes of nitrification, denitrification, and nitrogen fixation.
Filling these knowledge gaps could help to reveal the underlying mechanisms of soil responses to precipitation changes. For example, there is robust agreement across studies that soil respiration slows under DP and accelerates under IP, and so does 295 heterotrophic respiration. However, the relative importance of different mechanism in the response of heterotrophic respiration is still unknownin other words, how much of this response comes from changes in the level of microbial activity (e.g., entering and exiting dormancy) vs. substrate availability? Similarly, what are the most important mechanisms behind changes in N2O emissions, and how quickly will total soil nitrogen respond? Interestingly, the variables receiving the greatest attention are largely the easier to measure "fluxes" (i.e., greenhouse gas emissions) and "pools" (i.e., soil carbon, 300 biomass, and bacterial abundance). Measuring process rates (i.e., rates of nitrification, denitrification and fixation) that cannot be simply measured from gas fluxes requires more resources (time and money).

Challenges in meta-analyses
PMEs are quite diverse, adopting a variety of approaches, treatment levels, and treatment types (Beier et al., 2012;Kreyling and Beier, 2013), and so are the data derived from them. Many PMEs use long-term rainout shelters, which unavoidably 305 modify the ambient environment in other ways (Kreyling et al., 2017). While synthesizing the results of PMEs around the globe in the context of these experimental issues could be challenging, meta-analyses provide an exhaustive statistical summary of empirical studies (Hedges et al., 1999). One of the limitations of meta-analysis, however, is that it can obscure the substantial influence of environmental characteristics and methodological differences on effect sizes. Categorization by environmental characteristics, such as climate, geography, ecosystem, soil, and soil biota, can provide a local-to regional-310 view of soil responses that is specific to the given environmental characteristic. Categorization by methodology, such as experimental duration, intensity of treatment, measurement method, and fertilizer use, can clarify the human-derived impacts on effect sizes. These categorization efforts can identify when and how soil responses depend on their environmental context. As more and more PMEs are implemented, sample sizes available for meta-analysis are increasing (Song et al., 2019). In this regard, the recent deployment of broad networks of PMEs with standardized methodology and sampling 315 procedures (Halbritter et al., 2020) could ultimately contribute to more powerful meta-analyses with more easily interpreted https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License.
outcomes (Hilton et al., 2019;Knapp et al., 2012Knapp et al., , 2017. Details of categorization by environmental characteristics and methodology can be found in the Supplement (S1).

Conclusions
This synthesis of meta-analyses provides a broad perspective on how precipitation changes affect soils and belowground 320 processes. Belowground carbon and nitrogen cycles speed up with increased precipitation and slow down with decreased precipitation, while microbial communities are relatively resistant to precipitation changes. While response of the fluxes and pools of each cyclegas emissions, soil carbon, nitrogen ions, and biomasshave been studied extensively, responses of the associated process rates remain less studied or unexamined by meta-analyses. We suggest that additional scientific attention to these "processes" is warranted, and would strengthen the current knowledge of soil responses to precipitation 325 changes.

Author contribution 330
AOA, AS, YO, MRU, IR, and JSD designed the research. AOA, AS, YO, SR, MRU, JL, and IR conducted the synthetic analysis and contributed to writing the original draft. AOA prepared the manuscript with contributions from all co-authors.

585
https://doi.org/10.5194/bg-2020-30 Preprint. Discussion started: 11 February 2020 c Author(s) 2020. CC BY 4.0 License. Figure 1: (a, b) The effect size of (a) soil respiration and methane uptake, and (b) carbon and belowground biomass variables with respect to decreased (red) and increased (blue) precipitation. Filled points represent a significant effect size (95% CI not overlapping 0), and open points represent a non-significant effect size. Variable names correspond to Table 2. No. is meta-analysis 590 number and it corresponds to Table 1 and Table 2. The sample size is indicated by n. Asterisks indicate missing 95% CIs. (c, d) The effects of (b) decreased precipitation and (c) increased precipitation on soil carbon cycle. Negative, positive, and no effects are represented by −, +, and =, respectively. Red and blue are the variables found in one or more meta-analyses. Brown symbols in parentheses represent the variables that no meta-analyses quantified; in these cases, we estimated the effects based on our review of empirical studies in Sect. 3.1. Variable names correspond to Table 2. No. is meta-analysis number and it corresponds to Table 1 and Table 2. The sample size is indicated by n.