Nutrient resorption based on different estimations of five perennial herbaceous species from the grassland in inner Mongolia, China

Nutrient resorption at leaf level (leaf mass and leaf area) and at plant level were investigated for five temperate grassland species in a 4-years N-fertilization experiment (8 levels: 0, 1, 2, 4, 8, 16, 32, 64 g N m−2 year−1). Along N gradients, N concentrations in senescing tissues for all species increased; P concentrations in senescing leaves of Cleistogenes squarrosa, Potentilla acaulis and Artemisia frigida also increased in low N-addition plots but decreased in high N-addition plots. By contrast, the changes in P concentrations displayed insignificant trends for other two species in most cases; leaf-based N-resorption efficiency (NRE) generally decreased, whereas NRE at plant level did not show a clear trend in most species; leaf-based P-resorption efficiency (PRE) in C. squarrosa and P. acaulis showed decreasing trends, whereas PRE at plant level generally showed an insignificant relationship in most species. The findings suggested that the response of N-resorption was more dependent on methods and less on species than that of P-resorption. Both NRE and PRE estimated from leaf mass- and leaf area- levels were higher than those from plant level, indicating that different calculations can yield various patterns of nutrient-resorption in relation to soil N availability.

► We investigate the differences in nutrient resorption based on varied estimates.
► P-resorption proficiency is less sensitive to soil N supply compared to N-resorption proficiency.
► The response of PRE is less dependent on methods than the response of NRE.
► Different estimates yield various patterns of nutrient resorption responding to N availability.
► Resorption estimation at plant level is appropriate for perennial herbaceous species.