Discussion

Many studies have explored the responses of carbon and nitrogen processes to increasing plant diversity from both experiments and meta-analysis (Tilman et al. 2001; Roscher et al. 2004; Ma and Chen 2016; Chen et al. 2019; Chen et al. 2020). Three aspects of our study distinguish it from previous syntheses studies. First, our study is the first to compare across a large number of carbon and nitrogen processes. Second, our study attempts to reconcile the effects of experimental type by exploring the differences between field and greenhouse experiments. Third and most importantly, we show that there are important interactions between plant diversity and experimental age in field experiments, with the effects of plant diversity becoming more pronounced with time.

Carbon and nitrogen cycle changed significantly in plant mixtures

We generally found support for the hypothesis that more diverse plant mixtures positively affect the provision of many carbon and nitrogen ecosystem services in grasslands. Overyielding in aboveground biomass (AGB), belowground biomass (BGB), and total biomass (TB) in plant mixtures has been commonly observed due to complementary plant interactions (Tilman et al. 1997; Ma & Chen 2016; Wang et al. 2020); the strong general effects of diversity found here demonstrate that these mechanisms are broadly important in grassland ecosystems. Further, we observed positive relationships between the soil carbon pool (SCP) response and BGB, suggesting that the higher SCP frequently observed in plant mixtures can be caused by complementary in BGB (Fig. S1a). Higher microbial biomass (MB), fungal and bacterial biomass, and heterotrophic respiration (Rh) in plant mixtures are likely also attributable to the effects of higher productivity on the carbon and nutrient inputs to the soil ecosystem (Bartelt-Ryser et al.2005; Eisenhauer et al. 2010). Soil respiration (Rs) combines Rh and root metabolism (Luo & Zhou 2008), so there is likely both Rh and plant BGB impacts in mixtures on Rs.
Positive plant diversity effects were also observed for nitrogen attributes. These can be attributed to the positive relationship between aboveground nitrogen pool (ANP) and AGB (Fig. S1b), and again demonstrate the general importance of diversity complementary effects in grasslands. Positive relationships were found between soil nitrogen stock (SNP) and SCP (Fig. S1c), and thus the higher SNP in plant mixtures may be induced by overyielding in above- and belowground biomass. Larger carbon accumulations in soils could provide more exchange sites for ammonium (Mueller et al. 2013), which may cause the observed higher soil ammonium under plant mixtures. Soil nitrate and nitrogen leaching in plant mixtures was lower than monocultures probably due to greater N uptake capacity in a diverse mixtures (Mueller et al. 2013; Leimer et al. 2016). Stoichiometric theory shows that net N mineralization was positively related to root N concentration (Manzoni et al. 2008), and reduction in root N concentration (Mueller et al. 2013) may lead to a decrease in soil nitrogen mineralization in plant mixtures.

Divergent response between field and greenhouse experiments

Our comparison of field and greenhouse experiments demonstrates that diversity effects were generally stronger in field experiments than in the greenhouse. Duration and experimental size constraints in greenhouse experiments likely explain much of the difference. The maximum diversity for individual measures in the greenhouse was 3-16 species compared to 16-60 in the field. Similarly, greenhouse experiments older than 3 years were rare (Fig. S2) while many field experiments had run for a decade or more. Given the experimental age by diversity interactions identified in the field experiments (discussed below), it is not surprising that field experiments show stronger effects than greenhouse studies. Mechanisms important in longer experiments that would not be detected in the greenhouse could include the accumulation of dead plant material with time (Bartelt-Ryser et al. 2005; Eisenhauer et al. 2010; Chen et al. 2019). Further, we found no significant diversity – climate interactions, suggesting that the stronger responses in field experiments was not induced by the greater range of climate conditions in the field.

Species richness - experimental age interactions

We show that the effects of plant diversity on carbon and nitrogen processes increased with experimental age in the field experiments. Increases in the magnitude of complementary effects with time have been observed (Cardinale et al. 2007; Ravenek et al. 2014; Wang et al. 2020). The likely mechanism behind these time effects is the accumulation of plant biomass and soil carbon with time (Khlifa et al. 2017). Positive interactions between plant diversity and experimental age on ANP and SNP may also be caused by the enhancement of complementary effects with time. However, There is a time lag in microbial processes response to changes in plant communities due to the accumulation of dead plant materials was needed before their response (Bartelt-Ryser et al. 2005; Eisenhauer et al. 2010), and thus the effects of plant diversity on microbial biomass, respiration, soil nitrate, and soil nitrogen mineralization, which are likely caused by the accumulation of primary production (Janssens et al. 2010; Mueller et al. 2013), was more pronounced at longer experimental duration.

Long-term impacts of diversity loss in grasslands

Our model shows that a 10% (from 100 to 90%) decrease in plant species richness could cause a small decline in SCP and SNP over one year, but cumulatively much larger declines are likely. The diversity of many plant communities worldwide is thought to be declining due to factors including global warming (Tilman & Lehman 2001; Tylianakis et al. 2008). The negative effects of plant diversity loss on carbon storage may generate positive feedbacks that could accelerate global warming. Nitrogen limitation of primary production in grasslands (Fayet al. 2015; Wieder et al. 2015), the positive interactive effects of plant diversity loss and time on SNP could aggravate the deficiency of nitrogen in grasslands, which substantially exacerbate the shortage of forage production.