Abstract
Leaf Area Index has been proven to be a valid proxy for grassland biomass estimation, but little focus has been put so far on different vegetation-related variables possibly influencing this relationship. In this study we evaluated the effect of vegetation parameters on the LAI-yield relationship. During field campaigns in 2021 and 2022, 1112 samples of both yield and LAI were taken at eight different permanent meadows. Moreover, the phenological stage (PS) of grasses, the yield proportion of grasses (G), forbs (F) and legumes (L) and the resulting sward type, lodging (LO) and herbage moisture (HM) were estimated by visual/sensory assessment and the growth cycle (GC) recorded. A baseline linear mixed effects model including just a quadratic polynomial accounting for LAI was compared to a more complex model including further vegetation-related explanatory variables, accounting for which allowed improving the model fit (R 2 improved from 0.72 to 0.91 and RMSE from 0.94 to 0.53). Dry matter yield was strongly increased by the advanced phenology of grasses and high yield proportion of them. Moreover, the slope of the increase in yield with increasing LAI decreased from the first to the following regrowths.