4. Discussion
4.1. Factors affecting crop yield
Our data indicated that at low MAT (0-3 °C), crop yield under NT was lower compared to CT (Fig. 3a). In NEC, many studies have reported similar findings (Y. Chen et al., 2011; S. Liu, Zhang, Kravchenko, & Iqbal, 2015), and similar results were also reported in other cooler regions worldwide (Malhi, Mumey, Osullivan, & Harker, 1988; T. D. West, Griffith, Steinhardt, Kladivko, & Parsons, 1996). This yield decline is often attributed to the lower soil temperatures under NT compared with CT in spring (Chassot, Stamp, & Richner, 2001; Drury et al., 1999; S. Liu et al., 2015; Sarkar & Singh, 2007), due to the lack of soil disturbance, which can lead to delays in maize emergence and a shorter growing period, thus causing a decline in yield (Soane et al., 2012). The straw mulching employed in NT is also a possible reason for the decrease in yield (Y. Chen et al., 2011; S. Liu, Zhang, Yang, & Drury, 2013), as this contributes to lower soil temperature and higher soil water concentration (Linden, Clapp, & Dowdy, 2000; X. J. Lu, Li, Sun, & Bu, 2015). However, in our research, as MAT increased, maize yield did not rise linearly (Fig. 8a), possibly because in higher temperature regions (e.g., southern part of NEC) lower concentrations of SOC were also observed, particularly above 7 ℃ (Fig. 9a). Therefore, RT may be a better conservation tillage practice in the colder areas of NEC than NT given its reduced impact on yield (Fig. 3a), likely due to the greater soil temperature under RT, which is very important in NEC in spring (He, Li, Kuhn, Wang, & Zhang, 2010). However, RT could still prevent the yield loss that occurs under CT when strong wind and heavy rain leads to lower crop lodging (Liang et al., 2017), which is particularly important during extreme weather events, such as typhoons.
The MAP also affected crop yield under conservation tillage. We observed that when MAP was either < 500 or > 600 mm, NT had a positive impact on yield, while when MAP was 500 - 600mm NT had a negative effect (Fig. 3b). Page et al. (2019) reported that the increased water infiltration and stocks common under NT can increase the soil water available for crop growth, providing a yield advantage compared to CT in drier regions, which could explain the improved yield observed in lower rainfall regions in the current study. The area of the NEC that experiences a MAP of 500 – 600 mm, mainly occurs in Heilongjiang Province, where MAT is also lower. Based on a global meta-analysis, W. J. Sun et al. (2020) reported that temperature might have a greater influence on yield in cold, arid areas, and this may explain the yield decrease observed in the current study.
Crop rotation is now recommended by government in NEC, however, our results found that NT with a rotational crop pattern decreased crop yield compared to CT (Fig. 3c). The reason for this might be that the main areas where crop rotation is applied are relatively cold and cool, and cooler temperature decreases crop yield under NT. However, there is abundant evidence for studies worldwide that rotation plays an important role in restoring soil fertility (Dumanski et al., 1998; Havlin, Kissel, Maddux, Claassen, & Long, 1990; Lal, Follett, Kimble, & Cole, 1999), so a rotational cropping pattern with RT or ST are likely to be most appropriate for cooler regions in NEC.
The adopted duration of conservation tillage also had an impact on the crop yield effect for ST. When ST had been in place for < 8 yrs, the yield improvements were smaller than when tillage practices had been in place for > 8 yrs (Fig. 3d). For the NT and RT treatments, there was a also trend towards increasing yield when practices had been in place for >8 yrs, although this was not significant. Other studies have reported that crop yield can increase when conservation tillage has been in place for both short (4 yrs) and long (12 yrs) periods of time (You et al., 2017; Zhang et al., 2015), however, a recent meta-analysis found that negative crop yield was commonly observed up until ~5 yrs (Pittelkow, Linquist, et al., 2015). The rise of yield effect in long experiments is possibly caused by the decrease of yield under CT as the soil degrades, as well as the improvements in soil structure under conservation tillage measures.
Similar to our findings, many studies also reported that ST has a significant positive effect on crop yield in northeastern and other parts of China (Feng et al., 2018; X. F. Sun et al., 2017). By loosening the soil, ST can increase rooting depth (Rajkannan & Selvi, 2002), improve infiltration and water storage (Peeyush, Tripathi, Surendra, & Ravindra, 2004; Rajkannan & Selvi, 2002), and finally increase crop yield. Therefore, tillage mode like ST rotated with NT, e.g., ST-NT-NT in 3 yrs, which has already been popularized and applied in Heilongjiang Province (Gong, Qian, & Yu, 2009), could be considered for the whole NEC area. However, the high cost of undertaking ST is a significant barrier to the large-scale promotion of this practice.
4.2. Factors affecting SOC concentration
As the core practice of conservation tillage, NT has been demonstrated to improve SOC sequestration by reducing soil disturbance and enhancing physical protection of SOC within soil aggregates (Six, Elliott, & Paustian, 2000; Six et al., 2002). However, the ability of conservation tillage measures to sequester SOC is known to differ under different climatic conditions (Zhao et al., 2015). Our result showed that, as MAT rose, the SOC effect increased significantly under NT and ST (P < 0.05) (Fig. 5a), although the SOC concentration still declined under both these tillage practices as MAT increased (Fig. 9a). This would indicate that there was a slower downward trend in SOC under NT and ST than CT as temperature increased. Temperature is an important factor responsible for SOC stability and higher temperatures increase SOC mineralization (Dong et al., 2019; Kan, Liu, Wu, et al., 2020). The greater protection of SOC afforded under NT and ST practices may have thus reduced the impact of temperature on mineralization rates relative to CT. Greater biomass production at high temperatures (as indicated by the increased yield effect at higher temperatures for NT (Fig. 3a) would also have contributed to greater SOC concentrations under NT.
The SOC concentration effect under NT in drier areas (MAP < 500 mm) was also significantly higher than that in more humid areas (P < 0.05) (Fig. 5b). Previous meta-analyses have observed decreased C sequestration following the implementation of NT in dry (< 1000 mm) compared to moist (> 1000 mm) rainfall regions (Ogle, Breidt, & Paustian, 2005). However, the current study analysed papers conducted over a much narrower rainfall range (408 - 714 mm). Within this rainfall range, areas with lower precipitation (< 500 mm) led to greater SOC gain following the introduction of NT. However, the reason for this is not fully understood.
NT used in combination with continuous cropping led to higher SOC concentrations than CT, but when used with a rotational pattern, the increase in SOC relative to CT was smaller, and it is significantly lower than that under continuous cropping pattern (Fig. 5c), which is consistent with the results found by T. O. West and Post (2002). Previous studies had found that SOC concentration or stocks could be increased only when NT or other tillage measures combined with crop residue retention (Dolan, Clapp, Allmaras, Baker, & Molina, 2006; Villamil & Nafziger, 2015). Some other studies also reported that the amounts of residue retention played an important role in SOC accumulation, with the more residue returned to soil, the greater the impact on SOC (Kubar et al., 2018; Plaza-Bonilla, Alvaro-Fuentes, & Cantero-Martinez, 2014). The common mode of rotational cropping in NEC is maize-soybean alternately planted every year, which produces less residue than continuous maize pattern, and may be responsible for the lower SOC increase under rotation (X. W. Chen et al., 2011).
Experimental duration of conservation tillage should have an impact on SOC stocks, although the results are sometimes contradictory especially when duration is below 10 yrs (Liang, Chen, Zhang, & Chen, 2014; VandenBygaart et al., 2011; VandenBygaart, Gregorich, & Angers, 2003). Our results show no significant increase in the SOC effect over time for NT or ST treatments, although there was a small but significant increase (P < 0.05) in the SOC effect in RT treatments for experiments conducted for > 8 yrs (Fig. 5d).
The SOC sequestration rate under NT at the depth of 0-20cm in NEC is 0.953 Mg C ha-1 yr-1, which is much higher than China’s national average in the same layer of 0.157-0.390 Mg C ha-1 yr-1(F. Lu et al., 2009). NT also had a significantly greater rate of sequestration than RT, and was trending higher than ST. This indicates that NT is likely to be an effective management strategy to sequester SOC and restore soil fertility in NEC.
4.3. The effects of SOC on crop yield
SOC can positively affect crop growth due to its impact on soil properties and processes (D’Hose et al., 2014; Doran & Zeiss, 2000). Previous studies have reported that SOC has a positively linear relation with crop yield (Lal, 2006; S. K. Wang et al., 2016). Our study found that maize yield had a positive relation with SOC concentration and stocks under NT up until ~20 g kg-1 or 40 Mg ha-1, but then decreased as SOC concentration continued to rise (Fig. 9). Some studies reported that there is a threshold where the continuous increase of SOC will not always contribute to growth of crop yield (Hijbeek et al., 2017), with 20 g kg-1 was the threshold suggested in the temperate climate zone (Greenland, 1975; Lal, 2020; Loveland & Webb, 2003; Oldfield, Bradford, & Wood, 2019). Our results also found the linear relation between SOC and maize yield up to SOC concentration of 20 g kg-1, but when SOC concentration increased to 30 g kg-1, maize yield dropped greatly under NT (Fig. 9a). The reason for this might be the colder temperatures experienced in regions with higher SOC, which could override any benefits of higher SOC concentration on crop yield.
4.4. Outlook of future research
As the Chinese government published the project of conservation tillage application in NEC this year, it can be predicted that there will be more research on conservation tillage in NEC in the future. Based on our research, we hope that upcoming research can focus on the following issues: (i) the need for more study on the impact of RT on yield and SOC sequestration, due to the large use of RT by farmers in NEC; (ii) there has been relatively little research conducted in the Inner Mongolia Autonomous region, and more research is required to better understand the operation of conservation tillage in this region; (iii) most research has been conducted in flat regions, and more research should be conducted on sloping and mountainous areas in order to better understand the role of conservation tillage in preventing soil erosion,; and (iv) greater quantitative evaluation should be conducted of the combined impact of conservation tillage on the yield, environment and social aspects of crop production.