Meta-analyses of mowing and warming effects on grassland species richness and harvested biomass

Climate and management affect grassland plant diversity but studies vary regarding the magnitude of changes in plant species richness. Here we develop a comprehensive understanding of species richness modification due to management (mowing) and climate (warming) variation worldwide, and present the results of two meta-analyses from 999 and 1793 records (articles). Recorded articles had at least one experiment with a case-control design. The results show that both mowing (43 articles) and warming (34 articles) modify species richness, which on average increased by c. 32% with once-a-year mowing (against no mowing) and reduced by c. 13% with warming (against ambient temperature). Our meta-analysis on the mowing regime supports the humped-back model, with one or two cuts per year being the level of disturbance optimising species richness. We also observed that warming-induced reduction in species richness is lower in dry climates (< 300 mm yr-1) and at low-elevations (< 1000 m a.s.l.). We also took into account, where available, a concomitant variable, harvested biomass (determined experimentally in both mown and unmown plots), and found that it overall decreased by c. 21% (mowing) and increased by c. 11% (warming). The evidence provided of an opposite response of species richness and harvested biomass to disturbance is consistent with the competitive-exclusion hypothesis of negatively correlated patterns between the two outcomes (high taxonomic diversity with low biomass production, and vice versa). Study results thus help develop a more complete picture of the role of increasing temperature and cutting regime on grassland species richness and add insights to published meta-analyses examining the outcomes of terrestrial ecosystems. The reported difficulties to retrieve representative studies in previous and the present meta-analyses highlight the need to focus on dedicated research for robust inference about environmental and management constraints on grassland performances.