Aims Our objective was to explore mechanisms underlying the effects of transgressive overyielding and plant diversity effects associated with legume-grass mixtures in Zhaosu Basin, Yili Prefecture, Xinjiang Uygur Autonomous Region, China.
Methods Replicate plots (n = 3) were sown in 2013 as legume and grass monocultures (8 in total, including Vicia sativa, Avena sativa, Onobrychis viciaefolia, Bromus inermis, Trifolium pratense, Phleum pratense, Medicage sativa, Dactylis glomerata), a two-species mixture (AM2, annual legume and grass; PM2, perennial legume and grass), a four-species mixture (PM4, 2 perennial legumes and 2 grasses), a six-species mixture (PM6, 3 perennial legumes and 3 grasses), and three mixed-ratio combinations (legume : grass = 6:4, 5:5 and 4:6). Total plot productivity was measured as forage yield (dry matter) each summer between 2013 to 2015, which were used to calculate transgressive overyielding effect (OV), transgressive overyielding effect 1 (OV1), and transgressive overyielding effect 2 (OV2). Plant diversity effects were determined by calculating the complementary effect (CE), selection effect (SEF), and net biodiversity effects (∆Y).
Important findings The OV amplitudes of AM2 in 2013, 2014 and 3 years average were less than those of PM2 and PM6. The OV amplitude of AM2 in 2015 was greater than those of PM2, PM4 and PM6. The difference between the productivity of the mixed community and the yield of the most productive product species in the community components and the average yield of each species showed similar laws. CE of AM2 in 2013, 2014, and 3 years average was greater than PM2, PM4 and PM6 in the respective years. While SEF were much smaller than CE in AM2, changes in CE were relatively stable in PM2, PM4 and PM6. Fitted curves of species richness, species evenness, and forage yield (community productivity) predominantly showed a single maximum in PM4, yet productivity was highest in the 5:5 ratio plot. CE, SEF,and ∆ Y in perennial legume-grass mixtures decreased over time, which also led to declines in OV amplitude, OV1 and OV2, and their stability. Thus, in the initial study year, CE and SEF jointly dominated OV, OV1 and OV2 in perennial legume-grass mixtures. Over the next two growing years, SEF became the main factor that influenced OV, OV1 and OV2 in perennial legume-grass mixtures. OV, OV1 and OV2 of annual legume-grass mixtures were all influenced by CE from 2013 to 2015.