Total annual precipitation cannot adequately represent the effective precipitation necessary for plant growth and production. In this research, forage production in north-central Iran rangelands was modeled by analyzing the effective precipitation during 10 water years, from 20 05–20 06 to 2014–2015. The production of key species, including Artemisia sieberi , Ephedra intermedia , Zygophyllum eurypterum , Stipa barbata , Scariola orientalis , Anabasis setifera , and annual plants, was measured. The relationship between forage production and total precipitation amounts, total precipitation exceeding 1, 3, 5, 7, and 10 mm, and temperature was quantified, and an optimal multivariate regression model was introduced to predict regional forage production for each species. Results revealed an average annual forage production of 57.71 kg ·ha−1 . Notably, Z.eurypterum exhibited minimal production fluctuations, ranging from 16.1 to 18.4 kg ·ha−1 , whereas annual species displayed a maximum coefficient of variation (79%). The most accurate model for estimating annual species forage production was based on April precipitation and April–May temperatures [Relative Root Mean Squared Error (RRMSE) = 0.19], which accounted for 98% of the variation ( P value < 0.01). The results also showed that the most accurate linear model for estimating total forage production was based on March–June precipitation events exceeding 5 mm, with R = 0.92 in the calibration step and RRMSE = 0.07 in the validation step. The results can be used for estimating the annual forage production, determining grazing capacity, predicting future changes in forage production, and informing optimal rangeland management.
