Global meta-analysis reveals agro-grassland productivity varies based on species diversity over time

Ecological research suggests increased diversity may improve ecosystem services, as well as yield stability; however, such theories are sometimes disproven by agronomic research, particularly at higher diversity levels. We conducted a meta-analysis on 2,753 studies in 48 articles published over the last 53 years to test: if biological N2 fixation (BNF) supplies adequate nitrogen (N) for plant growth relative to synthetic fertilizers; how crop physiological traits affect legume-grass symbiosis; and, how cultural practices affect BNF over a range of soils and climates overtime (in polycultures versus sole grasslands). Globally, net primary productivity (NPP; total aboveground production response of grass and legume in higher-diversity treatments) increased 44% via legume associations relative to sole grass controls (including both with and without N fertilizer). Several moderating variables affected NPP including: (i) plant photosynthetic pathway (mixtures of C3 grasses resulted in a 57% increase in NPP, whereas mixtures of C4 grasses resulted in a 31% increase; similarly cool-season legumes increased NPP 52% compared to a 27% increase for warm-season legumes relative to grasslands without diversity); (ii) legume life cycle [NPP response for perennial legume mixtures was 50% greater than sole grass controls, followed by a 28% increase for biennial, and a 0% increase for annual legumes)]; and, (iii) species richness (one leguminous species in a grassland agroecosystem resulted in 52% increase in NPP, whereas >2 legumes resulted in only 6% increases). Temporal and spatial effect sizes also influenced facilitation, considering facilitation was greatest (114% change) in Mediterranean climates followed by oceanic (84%), and tropical savanna (65%) environments; conversely, semiarid and subarctic systems had lowest Rhizobium-induced changes (5 and 0% change, respectively). Facilitation of grass production by legumes was also affected by soil texture. For example, a 122% NPP increase was observed in silt clay soils compared to 14% for silt loam soils. Niche complementarity effects were greatest prior to 1971 (61% change), compared to recent studies (2011–2016; -7% change), likely owing to reduced global sulfur deposition and increased ambient temperatures overtime. These historical trends suggest potential for legume intercrops to displace inorganic-N fertilizer and sustainably intensify global NPP. Results herein provide a framework for ecologists and agronomists to improve crop diversification systems, refine research goals, and heighten BNF capacities in agro-grasslands.

生态学研究表明，生物多样性提升或可改善生态系统服务功能，同时提高产量稳定性；然而此类理论有时会被农艺学研究推翻，尤其是在较高多样性水平的场景下。本研究针对过去53年间发表的48篇文献中的2753项研究开展了荟萃分析，旨在验证三个核心问题：其一，生物固氮（biological N2 fixation, BNF）能否为植物生长提供与合成氮肥相当的氮素（N）供应；其二，作物生理性状如何影响豆科-禾本科共生关系；其三，在不同土壤与气候条件下，栽培措施随时间如何影响生物固氮（相较于单播草地，混播草地的情况）。
全球范围内，相较于单播草地对照组（涵盖施用与不施用氮肥的情况），豆科植物共生体系使净初级生产力（net primary productivity, NPP；即高多样性处理下禾本科与豆科植物的总地上生物量响应）提升44%。多项调节变量会对NPP产生影响，包括：（i）植物光合途径：C3禾本科混播组合的NPP提升57%，而C4禾本科混播组合的NPP仅提升31%；与之类似，冷季型豆科植物使NPP提升52%，而暖季型豆科植物相较无多样性的草地仅提升27%；（ii）豆科植物生活周期：多年生豆科混播组合的NPP响应较单播草地对照组高出50%，其次为两年生豆科（提升28%），一年生豆科则无显著提升（0%）；（iii）物种丰富度：草地农业生态系统中加入1种豆科植物可使NPP提升52%，而添加超过2种豆科植物时，NPP提升仅为6%。
时间与空间效应量同样会影响共生促进作用：在地中海气候环境中促进作用最为显著（变化幅度达114%），其次为海洋性气候（84%）与热带稀树草原环境（65%）；与之相反，半干旱与亚北极系统中根瘤菌诱导的变化幅度最低，分别仅为5%与0%。豆科植物对禾本科植物生产的促进作用亦受土壤质地影响。例如，粉砂黏壤土中观测到NPP提升122%，而粉砂壤土中仅提升14%。
生态位互补效应在1971年之前最为显著（变化幅度61%），而近期研究（2011–2016年）则呈现-7%的变化，这或与全球硫沉降减少与环境温度逐年升高有关。上述历史趋势表明，豆科间作体系具备替代无机氮肥、可持续提升全球NPP的潜力。本研究结果可为生态学家与农学家优化作物多样化种植体系、明确研究目标以及提升农业草地的生物固氮能力提供理论框架。