Few studies have comprehensively assessed the forage nutrition quality and pools among different plant functional groups across various grassland types and management practices. Moreover, the relationship between forage yield and nutrition quality remains controversial, hindering the development of effective grassland management strategies. This study investigated 37 paired fencing-grazing sites along an 1800-km transect on the Tibetan Plateau to systematically evaluate the variations of forage nutrition quality and pool at both community and functional group levels (graminae, sedge, edible forb and unpalatable forb) across alpine meadows, steppes and desert steppes. Fencing increased spatial average crude protein (CP) content of community by 7.90 %, while reducing crude ash (ASH) and water-soluble carbohydrate (WSC) contents by 6.20 % and 7.27 %, respectively. The CP, ASH, ether extract (EE), acid detergent fiber (ADF) and neutral detergent fiber (NDF) pools were increased by 13.15-29.48 %. The effects of fencing varied with the 37 sites. Fencing heterogenized the spatial pattern of community nutrition quality, but homogenized the CP content of sedge. Graminaes provided the highest NDF. Sedges offered the highest CP and WSC, and moderate ADF, and EE. Edible forbs served the highest EE. Unpalatable forbs supplied the highest ASH. The relationships between forage yield and nutrition quality varied with functional groups and management types. Compared with the spatial heterogeneity method (based on the proportion of edible forage), the traditional mean value method may overestimate or underestimate the nutrition carrying capacity. Therefore, fencing generally improved forage crude protein content, at the expense of decreasing crude ash and water-soluble carbohydrate content. Contrary to expectations, sedges did not exhibit higher nutrition quality compared to other functional groups. The trade-offs between forage yield and nutrition quality did not always hold true. A fixed proportion of edible forage should not be used for the assessment of potential nutrition carrying capacity.