青藏高原3种常见恶性杂草野燕麦、一年生早熟禾、狗尾草分布图

以广泛分布于青藏高原农田中的3种常见恶性杂草，即野燕麦（Avena fatua L.）、一年生早熟禾（Poa annua L.）和狗尾草[Setaria viridis L.P.Beauv.]为研究对象，利用广义增强模型（GBM）、广义线性模型（GLM）、人工神经网络（ANN）、最大熵（MaxEnt）、随机森林（RF）及多元自适应回归样条（MARS）算法集合预测上述3种杂草在青藏高原的潜在地理分布以及驱动其变化的关键因子，以评估其对耕地的入侵风险.物种分布数据主要来源于国家标本平台（http://www.nsii.org.cn/）、中国数字植物标本馆（http://www.cvh.ac.cn/）、第二次青藏高原综合科学考察野外实地调查（野燕麦23个、一年生早熟禾19个和狗尾草21个）和相关文献记载。本文中地图省级及以上边界来源于审图号为GS（2019）1823的标准地图.

This study took three common noxious weeds widely distributed in farmlands of the Qinghai-Tibet Plateau, namely *Avena fatua* L. (wild oat), *Poa annua* L. (annual bluegrass), and *Setaria viridis* (L.) P.Beauv. (green foxtail), as research subjects. An ensemble of six algorithms including Generalized Boosted Model (GBM), Generalized Linear Model (GLM), Artificial Neural Network (ANN), Maximum Entropy (MaxEnt), Random Forest (RF), and Multivariate Adaptive Regression Splines (MARS) was used to predict the potential geographic distributions of the three weeds on the Qinghai-Tibet Plateau and the key factors driving their distribution changes, so as to evaluate their invasion risk to farmlands. Species occurrence data were mainly sourced from the National Specimen Platform (http://www.nsii.org.cn/), the Chinese Virtual Herbarium (http://www.cvh.ac.cn/), field surveys of the Second Tibetan Plateau Scientific Expedition and Research (23 records for *Avena fatua* L., 19 for *Poa annua* L., and 21 for *Setaria viridis* (L.) P.Beauv.), and relevant literature records. The provincial and higher-level administrative boundary maps used in this study were derived from the standard map with the map examination and approval number GS(2019)1823.

搜集汇总

背景与挑战

背景概述

该数据集提供了2019-2022年青藏高原地区野燕麦、一年生早熟禾和狗尾草3种恶性杂草的潜在地理分布图，空间分辨率为1km-10km，数据大小为522.47 MB。研究采用了多种算法模型预测杂草分布，并评估其对耕地的入侵风险，数据来源于国家标本平台、中国数字植物标本馆及实地调查。