欢迎访问林业科学,今天是

林业科学 ›› 2019, Vol. 55 ›› Issue (9): 177-184.doi: 10.11707/j.1001-7488.20190919

• 研究简报 • 上一篇    下一篇

2001—2016年喀斯特断陷盆地植被变化及其驱动因素

庄义琳1,2,3, 周金星1,2,3, 吴秀芹1,2,3, 曹建华4, 章维鑫1,2,3   

  1. 1. 北京林业大学水土保持学院 北京 100083;
    2. 北京林业大学云南建水荒漠生态系统国家定位研究站 建水 654399;
    3. 北京林业大学水土保持国家林业和草原局重点实验室 北京 100083;
    4. 中国地质科学院岩溶地质研究所 桂林 541004
  • 收稿日期:2018-01-04 修回日期:2019-08-12 发布日期:2019-10-28
  • 基金资助:
    国家重点研发计划(2016YFC0502500/2016YFC0502506);国家自然科学基金项目(41671080)。

Vegetation Change and It's Driving Forces in Karst Faulted Basins between 2001 and 2016

Zhuang Yilin1,2,3, Zhou Jinxing1,2,3, Wu Xiuqin1,2,3, Cao Jianhua4, Zhang Weixin1,2,3   

  1. 1. School of Soil and Water Conservation, Beijing Forestry University Beijing 100083;
    2. Jianshui Station for Desert Ecological System Observation and Research, Beijing Forestry University Jianshui 654399;
    3. Key Laboratory of Soil and Water Conservation of National Forestry and Grassland Administration, Beijing Forestry University Beijing 100083;
    4. Institute of Karst Geology, Chinese Academy of Geological Sciences Guilin 541004
  • Received:2018-01-04 Revised:2019-08-12 Published:2019-10-28

摘要: [目的]探究喀斯特断陷盆地2001-2016年间植被变化的空间趋势特征,并量化关键驱动因子对植被恢复的影响程度和贡献概率,为石漠化地区生态建设提供理论依据。[方法]以Savitzky-Golay(SG)滤波重构后的2001-2016年MODIS NDVI时间序列为数据源,采用Theil-Sen Median(TS)趋势分析和CMK检验获取断陷盆地年均NDVI值的空间趋势特征,利用Netica软件构建贝叶斯网络(BBN)模型,量化植树造林面积、岩性、海拔等驱动因子对断陷盆地年均NDVI变化的影响程度和贡献概率。[结果]2001-2016年间,断陷盆地年均NDVI显著上升的区域主要分布于云贵高原中部宣威和六盘水及青藏高原东南缘盐源地区,显著下降的区域主要为横断山脉中段木里县、南段米易以及云贵高原中部昆明和玉溪地区;关键驱动因子对年均NDVI变化的影响程度表现为植树造林面积 > 海拔 > 温度 > 人口密度 > 坡度 > 土地利用 > 岩性;植树造林面积越大,植被改善概率越大,植被改善明显的区域主要集中于植树造林面积>150 km2的范围;海拔小于1 845 m区域的植被改善概率明显变大,而高于1 845 m的区域植被改善的概率明显变小;增温显著的区域植被改善概率更低;人口密度和NDVI变化具有一定相关性,105~210人·km-2人口密度范围是植被改善概率变大的重点区域;坡度越低,植被改善越明显,并集中分布于0°~6°坡度较缓地区;从岩性来看,纯喀斯特区和亚喀斯特区植被恢复效果较好。[结论]喀斯特断陷盆地大部分地区(78%)植被变化不明显,但有11.4%的区域植被显著改善,10.6%的区域植被显著退化;人类生态建设中的植树造林面积是引起植被变化的最主要因素,其次为海拔和温度等自然要素;因植树造林等生态项目主要实施在石漠化地区,因此喀斯特地区的植被改善要优于非喀斯特地区;喀斯特石漠化综合治理中,对植被改善概率较低的地区,如高海拔地区和坡度较陡的地区,建议以自然修复模式为主;对于植被恢复的集中分布区,如植树造林面积较大地区和坡度较缓地区,植被改善明显,但后期应加强植被保护力度;由于气候和人工干扰强烈,增温显著和人口较密集地区植被退化的概率大,需格外关注。

关键词: NDVI, 喀斯特, 断陷盆地, 石漠化, 驱动因子, 贝叶斯网络

Abstract: [Objective]This paper was intended to explore the spatial characteristics of vegetation changes in karst faulted basins during 2001-2016, and to quantify the influence and contribution of the key driving forces to the vegetation restoration, in order to provide a theoretical basis for ecological construction of rocky desertification areas.[Method]Based on the time-series MODIS-NDVI from 2001 to 2016 after Savitzky-Golay (SG) filtering, Theil-Sen Median (TS) analysis and CMK test were applied to obtain the spatial characteristics of the annual average NDVI value of faulted basins, and then, the Bayesian Belief Networks (BBN) model was constructed using Netica software to quantify the influences and contributions of the driving factors on the annual average NDVI, such as afforestation area, lithology and elevation.[Result]During the period of 2001-2016, regions with significantly increased annual average NDVI were mainly distributed in the middle Xuanwei and Liupanshui of Yunnan-Guizhou Plateau, and the southeast of Yanyuan of Qinghai-Tibet Plateau, while regions with decreased annual average NDVI were mainly distributed in central Muli of Hengduan Mountains, southern Miyi county, and central Kunming and Yuxi of Yunnan-Guizhou Plateau; the influences of the key driving forces on annual NDVI changes were in the following order:afforestation area, altitude, temperature, population density, slope, land use, lithology; the probability of vegetation improvement grew with the increasing area of afforestation, which mainly concentrated on the afforestation area over 150 km2; the probability of vegetation improvement increased significantly in areas with altitudes lower than 1 845 m,and the probability of vegetation improvement decreased obviously with altitudes higher than 1 845 m; the probability of vegetation improvement dropped in the areas with significantly increased temperature; the NDVI change was correlated with population density, and the area with population density from 105 to 210 person·km-2 was the key area with increased probability of vegetation improvement; the probability of vegetation improvement went up in the areas with gentle slopes, mainly between 0 and 6°; from the perspective of lithology, the areas with vegetation improvement were mainly distributed in the karst area and the semi-karst area.[Conclusion]The vegetation changes in most areas of karst faulted basins (78%) were not significant, but the vegetation in 11.4% of the areas was significantly improved, and the vegetation in 10.6% of the area was evidently degraded. The most important factor causing vegetation change was the afforestation area in human ecological construction, followed by natural factors such as altitude and temperature; ecological projects such as afforestation project were mainly implemented in rocky desertification areas, so the improvement of vegetation in karst areas was better than that in non-karst areas; For restoration of rocky desertification in faulted basins, the natural restoration was recommended for areas with low probability of vegetation improvement, such as areas with high altitude and steep slopes; while in areas with centralized distribution of vegetation restoration, we should pay more attention to vegetation protection in near future; the areas with significant increase of temperature and dense population, due to strong climate and artificial disturbance, the probability of vegetation degradation have increased, which would be a focus in future researches.

Key words: NDVI, Karst, faulted basins, rocky desertification, driving forces, Bayesian Belief Networks

中图分类号: