1 | Introduction

Human activities such as agricultural production and fossil fuel combustion add ~200 Tg/year of reactive nitrogen (N) to global ecosystems, which is approximately equal to that provided by natural N fixation (Galloway et al., 2004; Schlesinger, 2009; Fowler et al., 2013). Apart from N deposition increasing, biodiversity loss is another global change issues that human being confront, which is mainly induced by land use change, climate change, and N deposition increase (Sala et al., 2000; Gossner et al., 2016; Harpole et al., 2016). Recent study showed that the world’s seed-bearing plants have been disappearing at a rate of nearly 3 species a year since 1900, which is up to 500 times higher than would be expected as a result of natural forces alone (Ledford, 2019). It has been proved that the plant life cycles and ecosystem functions is significantly changed under N deposition increase and plant diversity loss (Hautier et al., 2014; Grace et al., 2016; Wolf et al., 2017; Pennekamp et al., 2018), yet how N addition, biodiversity loss, and its interactions influence on plant phenology remains uncertain.
Flowering phenology, including flowering date and duration, is important developmental stages in plant phenology as well as a sensitive indicator of global changes (Hovenden et al., 2008; Hulme 2011; Wang et al., 2014; Suonan et al., 2017; Piao et al., 2019). Many factors control the flowering phenology in grassland ecosystems, N deposition increase and biodiversity loss are two of the most important (Sanz et al., 2011; Smith et al., 2012; Xia & Wan, 2013; Xi et al., 2015; Wolf et al., 2017; Wang & Tang, 2019a). While numerous studies reveal that N addition delayed the flowering phenology of plant (Smith et al., 2012; Xia & Wan, 2013; Wang & Tang, 2019a), and plant diversity loss advanced flowering events (Wolf et al., 2017). Specifically, increasing N inputs lead to an increase in available soil N (Smith et al., 2012), and then could delay flowering phenology indirectly through extending the length of vegetative growth (Wang & Tang, 2019a). While plant diversity loss could influence flowering phenology indirectly, via effects on abiotic processes (soil moisture and temperature) and resource availability (available soil N), or directly, via biotic interactions (plant density) (Wolf et al., 2017; Du et al., 2019). The projected N addition and plant diversity loss could change abiotic and biotic processes in the community and thus impact plant flowering phenology.
Functional traits of a plant are measurable biotic properties related to adaptation to environment (Enquist et al., 2015; He et al., 2018; Gustafsson & Norkko, 2019). Compared with abiotic factors, the alteration in biotic factors (e.g. functional traits) may be the more directly and better indicator of ecosystem function changes under N addition and plant diversity loss (Dickson et al., 2014; Cadotte, 2017). Given that soil N availability often alters the switches from the growth to reproduction stage (Cleland et al., 2006), and thus increasing N could delay flowering phenology (Wang & Tang, 2019a); available soil N increase under plant diversity loss, and then plant species flowered earlier, with peak flowering date advancing an average of 0.6 days per species lost (Wolf et al., 2017). However, previous studies showed that the indirectly effects of abiotic factors on phenology through altering plant morphological and physiological traits (König et al., 2018; Pérez-Ramos et al., 2019). Plant light and nutrient acquisition traits could elucidate linkages between plant phenology and environmental changes (Jia et al., 2011; Dorji et al., 2013; König et al., 2018; Pérez-Ramos et al., 2019). For instance, traits associated with resource acquisition, such as rooting depth and life history, mediate plant reproductive phenology responses to changing climatic conditions in an alpine grassland (Dorji et al., 2013); traits related to competition and growth rate, like plant height, specific leaf area and leaf dry matter content, had substantial explanatory power in the effects of climate changes on the first flowering day at the global scale (König et al., 2018; Pérez-Ramos et al., 2019). Therefore, our knowledge regarding the linkages between changes in flowering phenology and functional traits is crucial for better understanding the regulating mechanisms of N enrichment and plant diversity loss on plant phenology and predicting grassland community dynamics under increasing N deposition and biodiversity loss scenarios.
Here, we report on a common garden experiment investigating the influence of N deposition increase and plant diversity loss on three flowering phenology of Medicago sativa in an assemblage grasslands (Supplementary Fig. 1). Specifically, we explored the effects of N addition and multiple plant diversity levels on the first flowering day, the last flowering day, flowering duration, flower numbers, environmental factors, and a series of plant functional traits. We used partial correlation, variation partitioning, and structural equation modelling analysis to quantify the direct and indirect effects of N addition and plant diversity loss on flowering phenology and flower numbers. We hypothesized that (a) increased N inputs and plant diversity would delay flowering phenology; (b) increased N inputs promote light acquisition traits, and decreased plant diversity promote light and nutrient acquisition traits; and (c) changes in abiotic factors and functional traits would together drive the response of plant flowering phenology to N addition and plant diversity loss.