Abstract — One of the important factors in proper rangeland management is determination of grazing capacity which is related to accurate and precise estimation of forage production. As the estimation of forage production is only possible through sampling, so that determination of adequate sample size and suitable sampling method are the main objective of our applied research. The sampling was conducted in five locations in northern grasslands of Iran. In each location, two microplots of 64 m 2 were established based on international protocols. By combining the data of five locations, a layout of 640 m 2 was simulated in R software. This layout is considered as our statistical population. Primary sample sizes of 20, 25, 30, 35, and 40 were selected and by using different allowable errors with the lowest coefficient of variation, adequate sample sizes of 60 and 120 were obtained. These plot sizes were used to compare the accuracy and precision of methods including random, systematic, random–systematic, two-stage, and stratified sampling. The results showed as the allowable errors are decreasing, the sample sizes are increasing. In general, the mean of stratified sampling is almost equal to the actual mean of the population, indicating its higher accuracy and precision in comparison to other methods. Sample size of 60 and stratified sampling method with 3 to 5 strata were our ultimate selections.

中文翻译：

---

通过模拟估算草地产量的充足样本量和适当的抽样方法

摘要 — 牧场管理的重要因素之一是放牧能力的确定，它关系到牧草产量的准确和精确估计。由于牧草产量的估计只能通过抽样进行，因此确定足够的样本量和合适的抽样方法是我们应用研究的主要目标。采样是在伊朗北部草原的五个地点进行的。在每个地点，根据国际协议建立了两个 64 m 2 的微型小区。通过结合五个位置的数据，在R 软件中模拟了640 m 2 的布局。这种布局被认为是我们的统计人口。选择了 20、25、30、35 和 40 的主要样本量，并通过使用具有最低变异系数的不同允许误差，获得了 60 和 120 的足够样本量。这些地块大小用于比较随机、系统、随机-系统、两阶段和分层抽样等方法的准确性和精密度。结果表明，随着允许误差的减小，样本量在增加。一般来说，分层抽样的均值几乎等于总体的实际均值，表明与其他方法相比具有更高的准确度和精密度。60 个样本大小和 3 到 5 个层的分层抽样方法是我们的​​最终选择。一般来说，分层抽样的均值几乎等于总体的实际均值，表明与其他方法相比具有更高的准确度和精密度。60 个样本大小和 3 到 5 个层的分层抽样方法是我们的​​最终选择。一般来说，分层抽样的均值几乎等于总体的实际均值，表明与其他方法相比具有更高的准确度和精密度。60 个样本大小和 3 到 5 个层的分层抽样方法是我们的​​最终选择。