Climate rather than overgrazing explains most rangeland primary productivity change in Mongolia

Rangelands are Earth’s dominant land type, supporting the livelihoods of over two billion people. Concerns about rangelands degradation typically focus on overgrazing. But climate change may be a greater culprit. Using spatially disaggregated, nationwide data from Mongolia, 1984-2024, we exploit seasonal variation in grazing locations to quasi-experimentally estimate the causal effects of livestock herd size, weather, and climate change on rangeland primary productivity. At interannual frequency, herd size modestly but significantly negatively affects primary productivity, with notable variation across agro-ecosystems. The effects of weather fluctuations are, however, an order of magnitude larger. At decadal scale, over which herders can adapt to climate change, herd size effects disappear, and temperature effects dominate. In Mongolia, climate change seems to drive most long-term change in rangeland primary productivity., Data for this paper has been collected from a variety of publicly available and official sources in Mongolia. 
Data processing was completed using a combination of Google Earth Engine and Stata., , REPLICATION GUIDE FOR

“Climate Rather than Overgrazing Explains Most Rangeland Primary Productivity Change in Mongolia”

Avralt-Od Purevjav , Tumenkhusel Avirmed
Steven W. Wilcox, Christopher B. Barrett

May 19, 2025

Contact for replication and code: Dr. Avralt-Od Purevjav ([[e-mail]](mailto:[e-mail]))

The present document provides guidance for replication of all tables and figures present in Purevjav et al. 2025. All steps necessary for obtaining, organizing, and cleaning all final data in the paper are clearly described in the Supplementary Materials.

All analysis, figures, and tables were completed using Stata 18. Older versions of Stata are expected to work as well, though updates to software may be required if packages are not compatible with older versions of Stata.

There are three folders included that provide all data and code necessary for replication.

* code: contains all .do files that implement analysis
* data: contains all data files needed for rep...,

草原牧场（rangelands）是地球最主要的土地类型，维系着超过20亿人口的生计。针对草原牧场退化的担忧通常聚焦于过度放牧，但气候变化或许才是更主要的元凶。本研究利用1984年至2024年蒙古国全国范围内的空间细分数据，借助放牧地点的季节差异，采用准实验方法估算牲畜存栏量、天气波动以及气候变化对草原牧场初级生产力的因果效应。

在年际尺度上，牲畜存栏量对初级生产力产生小幅但显著的负向影响，且不同农业生态系统间存在显著差异。然而，天气波动的影响幅度要高出一个数量级。在十年尺度上，牧民可适应气候变化，此时牲畜存栏量的影响消失，温度的影响占据主导地位。在蒙古国，气候变化似乎是驱动草原牧场初级生产力长期变化的主要因素。

本论文所用数据来源于蒙古国各类公开及官方渠道。数据处理结合了Google Earth Engine（谷歌地球引擎）与Stata软件完成。

《相较于过度放牧，气候变化更能解释蒙古国草原牧场初级生产力的变化》复制指南
作者：Avralt-Od Purevjav、Tumenkhusel Avirmed、Steven W. Wilcox、Christopher B. Barrett
2025年5月19日
复制与代码咨询联系人：Avralt-Od Purevjav博士（邮箱：[e-mail]）

本文档旨在指导复现Purevjav等人2025年论文中的全部表格与图表。论文中获取、整理及清洗所有最终数据所需的全部步骤，已在补充材料中详细说明。

所有分析、图表及表格均通过Stata 18完成。旧版本Stata理论上也可运行，但如果软件包与旧版本Stata不兼容，则可能需要更新软件。

本数据集包含三个用于复现研究的完整数据与代码文件夹：
* code：包含所有用于执行分析的.do格式文件
* data：包含复现所需的全部数据文件