XIANG Yuyong, TAO Qin, YU Shijun, et al. Analysis of nutrients in larva feces of Heterolocha jinyinhuaphaga[J]. Journal of Zhejiang A&F University, 2020, 37(5): 971-977. DOI: 10.11833/j.issn.2095-0756.20190624
Citation: WANG Yan, YU Yun, LIU Yong, et al. Health assessment and influencing factors of Fraxinus pennsylvanica in Beijing core area[J]. Journal of Zhejiang A&F University, 2022, 39(6): 1340-1349. DOI: 10.11833/j.issn.2095-0756.20220109

Health assessment and influencing factors of Fraxinus pennsylvanica in Beijing core area

DOI: 10.11833/j.issn.2095-0756.20220109
  • Received Date: 2022-01-09
  • Accepted Date: 2022-06-05
  • Rev Recd Date: 2022-05-28
  • Available Online: 2022-11-21
  • Publish Date: 2022-12-20
  •   Objective  For the rapid, accurate, and productive assessment of urban tree health, this study aims to investigate and analyze the growth, health, and management status of Fraxinus pennsylvanica in Beijing core functional area, so as to provide a scientific basis for comprehensive health assessment and technical measures for management and protection.  Method  Taking street trees of 11 main streets in the core area of Beijing as the research object, the health assessment model of F. pennsylvanica was constructed by principal component analysis and K-means clustering. The results were verified by discriminant analysis, and the BP neural network health prediction model was established by combining the assessment results with the assessment indexes. Meanwhile, the effects of DBH, tree height, the number of pruning stubble, anti-trampling pavement etc. on the health status of F. pennsylvanica were analyzed.  Result  The results indicated that healthy, sub-healthy, unhealthy, and dying plants accounted for 39.20%, 41.26%, 16.78%, and 2.76%, respectively. In the BP neural network health prediction model, the correlation coefficients between expected value and predicted value of the training set, verification set, test set, and total set were 0.9997, 0.9720, 0.9976 and 0.9953 respectively, all greater than 0.9500, indicating that the model could accurately reflect the relationship between 12 evaluation indexes and health assessment categories of F. pennsylvanica and could be used to predict the health status of other F. pennsylvanica in the same area. Variance analysis and multiple comparisons showed that DBH, pruning stubble, anti-trampling pavement, plant spacing, tree pool area, number of lanes and lane direction all had significant effects on the health status of F. pennsylvanica (P<0.05), but tree height had little effect. The health condition of F. pennsylvanica was the best under the following conditions: 60−70 cm DBH, tree body without pruning stubble, tree pool pavement with good water permeability, 5−10 m tree spacing, 1−2 m2 tree pool area, suitable road width, and north-south direction.  Conclusion  F. pennsylvanica street trees in the study area are in a sub-health state on the whole, with good conservation and maintenance potential. However, some of them are seriously damaged and need to be dealt with urgently. In planting and management process, important measures should be taken to ensure the good status of F. pennsylvanica, such as standard pruning, strong water permeability of pavement materials, 5−10 m plant spacing, 1−2 m2 tree pool area, and suitable road conditions . [Ch, 3 fig. 13 tab. 26 ref.]
  • [1] CHENG Jiajie, CHEN Rong, YU Weiwu, CHEN Wenchao, HU Yuanyuan, ZHANG Zuying, SONG Lili.  Differences in free amino acids during post-ripening process of Torreya grandis ‘Xi Fei’ and ‘Xiangya Fei’ seeds . Journal of Zhejiang A&F University, 2025, 42(2): 239-248. doi: 10.11833/j.issn.2095-0756.20250169
    [2] CUI Chuandeng, SHI Qingdong, LI Xiang, SHI Haobo, WAN Yanbo, LI Dinghao.  Health evaluation of Populus euphratica forest in Dali Yabuyi Oasis in the hinterland of the desert . Journal of Zhejiang A&F University, 2024, 41(6): 1134-1141. doi: 10.11833/j.issn.2095-0756.20240148
    [3] WANG Shuang, DONG Bin, WANG Yiguang, ZHAO Hongbo.  Analysis and evaluation of flower and fruit characteristics of different Prunus mume cultivars . Journal of Zhejiang A&F University, 2024, 41(1): 113-123. doi: 10.11833/j.issn.2095-0756.20230213
    [4] LUAN Jingran, FENG Guohong, ZHU Yujie.  Visible/near infrared spectrum wood identification based on SPA-GA-BP neural network . Journal of Zhejiang A&F University, 2022, 39(3): 671-678. doi: 10.11833/j.issn.2095-0756.20210377
    [5] NIU Yuan, AO Yan, LI Yun, TIAN Xiuming, YANG Changwen, LIU Xiaotian, LI Zhihong.  Selection of pollination combinations and analysis of fruit and seed characters of excellent clones of Xanthoceras sorbifolium . Journal of Zhejiang A&F University, 2020, 37(2): 209-219. doi: 10.11833/j.issn.2095-0756.2020.02.003
    [6] HUO Yan, ZHAO Xueqing, HUANG Houyi, HUANG Xianbin, XU Yunfang, ZHU Zunling, YUAN Zhaohe.  Phenotypic genetic diversity of ornamental pomegranate cultivars . Journal of Zhejiang A&F University, 2020, 37(5): 939-949. doi: 10.11833/j.issn.2095-0756.20190619
    [7] MAO Guangxian, TAN Wei, CHAI Zongzheng, ZHAO Yang, YANG Shenjun.  Diameter-height model for Pinus massoniana plantations based on BP neural network . Journal of Zhejiang A&F University, 2020, 37(4): 752-760. doi: 10.11833/j.issn.2095-0756.20190486
    [8] JIANG Dongyue, SHEN Xin, CHEN Yajing, ZOU Yihan, WU Fan, LI Yingang, LIU Xinhong.  Morphological variation analysis of leaf and branch traits of wild Cerasus spp. in Zhejiang . Journal of Zhejiang A&F University, 2019, 36(4): 723-732. doi: 10.11833/j.issn.2095-0756.2019.04.012
    [9] QIAN Li, ZHANG Chao, QI Peng, YU Shuquan.  Sourcing and evaluating heavy metal pollution in the urban topsoil of Yongkang City . Journal of Zhejiang A&F University, 2016, 33(3): 427-433. doi: 10.11833/j.issn.2095-0756.2016.03.008
    [10] ZHANG Guanglai, LI Lu, LIAO Wenmei.  Principal component analysis of China's forestry industry competitiveness . Journal of Zhejiang A&F University, 2016, 33(6): 1078-1084. doi: 10.11833/j.issn.2095-0756.2016.06.022
    [11] ZHENG Rong.  Canonical correlation and principal components analysis of different production areas of Dendrocalamopsis oldhami with quality indicators of bamboo shoots and its soil nutrients . Journal of Zhejiang A&F University, 2012, 29(5): 710-714. doi: 10.11833/j.issn.2095-0756.2012.05.012
    [12] YANG Jian-hua, LI Shu-fang, FAN Zhi-yuan, XI Xue-liang, ZOU Wei-lie, LIU Jiao, PAN Li.  Principal component analysis for major economic characteristics of Carya illinoensis and selection of superior cultivars . Journal of Zhejiang A&F University, 2011, 28(6): 907-910. doi: 10.11833/j.issn.2095-0756.2011.06.011
    [13] WANG Rong.  Principal component analysis and evaluation of waterfowl habitats along the coastline of Fujian Province . Journal of Zhejiang A&F University, 2011, 28(3): 472-478. doi: 10.11833/j.issn.2095-0756.2011.03.020
    [14] SHEN Jun-ling, NI Hui-qun, CHEN Xiao-yang, HUANG Shao-wei.  Genetic diversity of Jatropha curcas with SRAP molecular markers . Journal of Zhejiang A&F University, 2010, 27(3): 347-353. doi: 10.11833/j.issn.2095-0756.2010.03.005
    [15] JIN Xiao-xiao, CHEN Fa-di, CHEN Su-mei, FANG Wei-min.  Nutrition in 20 cultivars of chrysanthemum . Journal of Zhejiang A&F University, 2010, 27(1): 22-29. doi: 10.11833/j.issn.2095-0756.2010.01.004
    [16] LIU Li, CHEN Shuang-lin, LI Yan-hong.  Stand structure and bamboo shoot number production based assessment of degradation degree of Phyllostachys praecox covered with organic materials . Journal of Zhejiang A&F University, 2010, 27(1): 15-21. doi: 10.11833/j.issn.2095-0756.2010.01.003
    [17] ZUO Ji-lin, GONG Chun, WANG Jian-ping, ZHOU Wen-cai, WEN Qiang, XU Lin-chu.  Evaluation on quality of twenty-five clones of Camellia oleifera Group Gan . Journal of Zhejiang A&F University, 2008, 25(5): 624-629.
    [18] SHI Yong-jun, XU Xiao-jun, DU Hua-qiang, ZHOU Guo-mo, JIN Wei, ZHOU Yu-feng.  Remote sensing image based bamboo forest monitoring with a back propagation(BP) neural network . Journal of Zhejiang A&F University, 2008, 25(4): 417-421.
    [19] LI Wei, ZHOU Wei, JI De, ZHANG Ren-gong.  Habitat use of Syrmaticus humiae in Nanhua Part of Ailaoshan National Nature Reserve in spring . Journal of Zhejiang A&F University, 2006, 23(2): 153-158.
    [20] HUANG Bi-heng.  Statistical analysis on national economy of every city in Zhejiang Province . Journal of Zhejiang A&F University, 1999, 16(4): 420-424.
  • [1]
    ZHOU Shuwen, GAO Fei, MA Keming. Research on species optimization of typical road trees in Beijing based on ENVI-met [J]. Chin Landscape Archit, 2020, 36(6): 141 − 144.
    [2]
    WANG Ying. The Structural Analysis and Health Assessment of Roadside Trees in Beijing [D]. Beijing: Chinese Academy of Forestry, 2011.
    [3]
    CHOW W T L, ROTH M. Temporal dynamics of the urban heat island of Singapore [J]. Int J Climatol, 2006, 26(15): 2243 − 2260.
    [4]
    PAN Hui, LIU Xiaohua, HUANG Shide, et al. Effects of urban forest trees on CO concentration in street space [J]. J Fujian Coll For, 2008, 28(4): 356 − 360.
    [5]
    FERNÁNDEZ-JURICIC E. Avifaunal use of wooded streets in an urban landscape [J]. Conserv Biol, 2000, 14(2): 513 − 521.
    [6]
    SHAO Peng. A Study on Health Assessment of Sophora japonica Street Trees in Beijing Urban Core Area [D]. Beijing: Beijing Forestry University, 2020.
    [7]
    WENG Shufei, LI Caimin, PANG Ruijun. Establishment of landscaping tree health assessment model using analytic hierarchy process [J]. J Northwest For Univ, 2009, 24(1): 177 − 181.
    [8]
    YOU Huiming, YOU Xiuhua, CHEN Xiaoling, et al. Integrated evaluation and selection of street trees species in Fuzhou City [J]. J Chin Urban For, 2009, 7(1): 15 − 17.
    [9]
    LUO Guibin. Comprehensive evaluation of the evergreen streettrees planted in the downtown of Hanzhong city [J]. J Northwest For Univ, 2016, 31(2): 302 − 308.
    [10]
    LI Jiahui, PENG Zuodeng, LIU Yong, et al. Health evaluation and influencing factors of Sophora japonica in the capital functional core area of Beijing [J]. J Northwest A&F Univ Nat Sci Ed, 2021, 49(1): 64 − 73.
    [11]
    DU Jiabao, PAN Pan, YANG Fangrong. Application of principal component analysis method in the comprehensive function assessment of the street trees in Zhengzhou [J]. J Northwest For Univ, 2009, 24(3): 190 − 193.
    [12]
    QIE Guangfa, PENG Zhenhua, WANG Cheng. Growth and health status of Ginkgo biloba in Beijing urban street area [J]. Res For, 2013, 26(4): 511 − 515.
    [13]
    MO Xunqiang, YAN Wei, CHEN Xiaokui, et al. Application of street trees’ vigorous degree in greening management: a case of Tianjin airport economic area [J]. Territ Nat Resour Study, 2012(1): 88 − 90.
    [14]
    MA Zhilin, CHEN Lihua, YU Xianwei, et al. Study on health assessment and comparison of different tree species in the west-mountains area, Beijing [J]. Ecol Environ Sci, 2010, 19(3): 646 − 651.
    [15]
    JI Wenyuan, XING Shaohua, GUO Ning, et al. Health evaluation on spruce and fir forests in Miyaluo of the western Sichuan [J]. Sci Silv Sin, 2009, 45(3): 13 − 18.
    [16]
    ZHU Yu, LIU Zhaogang, JIN Guangze. Health assessment of individual trees in natural Larix gmelinii forest in Great Xing’ an mountains of China [J]. Chin J Appl Ecol, 2013, 24(5): 1320 − 1328.
    [17]
    ZHANG Nan, DONG Li, HAO Peiyao, et al. Study on structure of street trees in central districts of Beijing [J]. J Cent South Univ For Technol, 2014, 34(5): 101 − 106.
    [18]
    YU Yun. A Study on Health Assessment of Roadside Trees Fraxinus ssp. in Core Area of Beijing [D]. Beijing: Beijing Forestry University, 2020.
    [19]
    Beijing Municipal Bureau of Market Supervision and Administration. Technical Regulations for Health Diagnosis of Urban Trees: DB11/T 1692−2019 [S]. Beijing: China Quality and Standards Publishing & Media Co. , Ltd.
    [20]
    HAN Xin, WEI Zheng, ZHANG Baozhong, et al. Crop evapotranspiration prediction by considering dynamic change of crop coefficient and the precipitation effect in back-propagation neural network model[J/OL]. J Hydrol, 2021, 596(3/4): 126104[2021-12-10]. doi: 10.1016/j.jhydrol.2021.126104.
    [21]
    LEI Ting, JIA Junyuan, TIAN Fujin, et al. Prediction of rock thermal conductivity based on BP neural network [J]. Glob Geol, 2021, 40(1): 131 − 139.
    [22]
    KUMAR M, RAGHUWANSHI N S, SINGH R. Artificial neural networks approach in evapotranspiration modeling: a review [J]. Irrig Sci, 2011, 29(1): 11 − 25.
    [23]
    HUANG Xiaoqiao, LI Qiong, TAI Yonghang, et al. Hybrid deep neural model for hourly solar irradiance forecasting [J]. Renewable Energy, 2021, 171: 1041 − 1060.
    [24]
    ZHAO Dan, LI Feng, WANG Rusong. Effects of ground surface hardening on plant eco-physiological progress inurban landscape [J]. Acta Ecol Sin, 2010, 30(14): 3923 − 3932.
    [25]
    LI Keke, LI Yanming, CONG Richen, et al. Simulating the influence of road planting environment on the growth of Ginkgo biloba [J]. Chin Landscape Archit, 2020, 36(7): 117 − 122.
    [26]
    WU Di. Configuration Models Research of Avenue Trees Based on Sunlight Analysis in Zhengzhou City[D]. Zhengzhou: Henan Agricultural University, 2018.
  • Created with Highcharts 5.0.7Amount of accessChart context menuAbstract Views, HTML Views, PDF Downloads StatisticsAbstract ViewsHTML ViewsPDF Downloads2024-052024-062024-072024-082024-092024-102024-112024-122025-012025-022025-032025-0400.250.50.7511.25Highcharts.com
    Created with Highcharts 5.0.7Chart context menuAccess Class DistributionFULLTEXT: 15.7 %FULLTEXT: 15.7 %META: 82.5 %META: 82.5 %PDF: 1.8 %PDF: 1.8 %FULLTEXTMETAPDFHighcharts.com
    Created with Highcharts 5.0.7Chart context menuAccess Area Distribution其他: 9.0 %其他: 9.0 %China: 0.1 %China: 0.1 %Seattle: 0.1 %Seattle: 0.1 %[]: 1.4 %[]: 1.4 %三明: 0.1 %三明: 0.1 %上海: 1.1 %上海: 1.1 %东莞: 0.4 %东莞: 0.4 %临汾: 0.3 %临汾: 0.3 %保定: 0.1 %保定: 0.1 %兰辛: 0.1 %兰辛: 0.1 %北京: 4.4 %北京: 4.4 %北伯根: 0.1 %北伯根: 0.1 %十堰: 0.4 %十堰: 0.4 %南京: 1.4 %南京: 1.4 %南昌: 0.1 %南昌: 0.1 %南通: 0.3 %南通: 0.3 %南阳: 0.1 %南阳: 0.1 %台州: 0.6 %台州: 0.6 %合肥: 0.3 %合肥: 0.3 %呼和浩特: 0.1 %呼和浩特: 0.1 %喀什: 0.1 %喀什: 0.1 %嘉兴: 0.3 %嘉兴: 0.3 %天津: 0.8 %天津: 0.8 %太原: 0.1 %太原: 0.1 %奥克兰: 0.1 %奥克兰: 0.1 %娄底: 0.1 %娄底: 0.1 %安德森: 0.1 %安德森: 0.1 %宣城: 0.6 %宣城: 0.6 %川崎: 0.1 %川崎: 0.1 %广州: 0.4 %广州: 0.4 %弗吉: 0.1 %弗吉: 0.1 %张家口: 3.3 %张家口: 3.3 %悉尼: 0.8 %悉尼: 0.8 %成都: 0.6 %成都: 0.6 %扬州: 0.8 %扬州: 0.8 %晋城: 0.1 %晋城: 0.1 %朝阳: 0.6 %朝阳: 0.6 %杭州: 2.5 %杭州: 2.5 %武汉: 0.1 %武汉: 0.1 %沈阳: 0.3 %沈阳: 0.3 %洛阳: 0.3 %洛阳: 0.3 %济南: 0.3 %济南: 0.3 %淄博: 0.1 %淄博: 0.1 %深圳: 0.7 %深圳: 0.7 %温州: 1.2 %温州: 1.2 %湖州: 0.8 %湖州: 0.8 %漯河: 4.0 %漯河: 4.0 %石家庄: 1.9 %石家庄: 1.9 %舟山: 0.1 %舟山: 0.1 %芒廷维尤: 23.9 %芒廷维尤: 23.9 %芝加哥: 2.2 %芝加哥: 2.2 %苏州: 0.8 %苏州: 0.8 %衡阳: 0.3 %衡阳: 0.3 %衢州: 0.1 %衢州: 0.1 %西宁: 18.5 %西宁: 18.5 %西安: 0.1 %西安: 0.1 %西雅图: 0.1 %西雅图: 0.1 %诺伊达: 0.3 %诺伊达: 0.3 %诺沃克: 3.6 %诺沃克: 3.6 %贵阳: 0.1 %贵阳: 0.1 %运城: 2.9 %运城: 2.9 %邯郸: 0.3 %邯郸: 0.3 %郑州: 1.2 %郑州: 1.2 %金华: 0.1 %金华: 0.1 %铁岭: 0.1 %铁岭: 0.1 %长沙: 1.8 %长沙: 1.8 %阜阳: 0.1 %阜阳: 0.1 %阳泉: 0.3 %阳泉: 0.3 %青岛: 0.4 %青岛: 0.4 %鞍山: 0.1 %鞍山: 0.1 %其他ChinaSeattle[]三明上海东莞临汾保定兰辛北京北伯根十堰南京南昌南通南阳台州合肥呼和浩特喀什嘉兴天津太原奥克兰娄底安德森宣城川崎广州弗吉张家口悉尼成都扬州晋城朝阳杭州武汉沈阳洛阳济南淄博深圳温州湖州漯河石家庄舟山芒廷维尤芝加哥苏州衡阳衢州西宁西安西雅图诺伊达诺沃克贵阳运城邯郸郑州金华铁岭长沙阜阳阳泉青岛鞍山Highcharts.com
  • Cited by

    Periodical cited type(6)

    1. 李莉,曾祥谓,周永红,王妍,杨开太,陈幸良. 虫茶营养成分及功能特性研究进展. 林产化学与工业. 2025(01): 162-172 .
    2. 金雅琴,李海玲,史莹,张谦,李冬林. 毛梾叶片营养成分与利用前景分析. 安徽农业大学学报. 2024(06): 914-919 .
    3. 袁冬寅,温立香,张芬,欧淑琼,陈家献,彭靖茹,赵媛,黄寿辉,檀业维. 不同干燥温度对酸枣叶虫茶品质成分的影响. 食品工业科技. 2023(08): 46-53 .
    4. 温立香,张芬,袁冬寅,陈家献,彭靖茹,黄寿辉. 广西虫茶毒理学安全性评价研究. 茶叶通讯. 2022(01): 102-107 .
    5. 温立香,袁冬寅,欧淑琼,张芬,陈家献,黄寿辉,冯春梅,彭靖茹. 广西三个产区虫茶主要茶特征成分、挥发性成分分析及氨基酸营养评价. 食品工业科技. 2022(20): 329-336 .
    6. 薄采颖,陈幸良,周永红,胡立红. 虫茶营养成分分析及评价. 林产化学与工业. 2022(05): 100-106 .

    Other cited types(0)

通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(3)  / Tables(13)

Article views(597) PDF downloads(13) Cited by(6)

Related
Proportional views

Health assessment and influencing factors of Fraxinus pennsylvanica in Beijing core area

doi: 10.11833/j.issn.2095-0756.20220109

Abstract:   Objective  For the rapid, accurate, and productive assessment of urban tree health, this study aims to investigate and analyze the growth, health, and management status of Fraxinus pennsylvanica in Beijing core functional area, so as to provide a scientific basis for comprehensive health assessment and technical measures for management and protection.  Method  Taking street trees of 11 main streets in the core area of Beijing as the research object, the health assessment model of F. pennsylvanica was constructed by principal component analysis and K-means clustering. The results were verified by discriminant analysis, and the BP neural network health prediction model was established by combining the assessment results with the assessment indexes. Meanwhile, the effects of DBH, tree height, the number of pruning stubble, anti-trampling pavement etc. on the health status of F. pennsylvanica were analyzed.  Result  The results indicated that healthy, sub-healthy, unhealthy, and dying plants accounted for 39.20%, 41.26%, 16.78%, and 2.76%, respectively. In the BP neural network health prediction model, the correlation coefficients between expected value and predicted value of the training set, verification set, test set, and total set were 0.9997, 0.9720, 0.9976 and 0.9953 respectively, all greater than 0.9500, indicating that the model could accurately reflect the relationship between 12 evaluation indexes and health assessment categories of F. pennsylvanica and could be used to predict the health status of other F. pennsylvanica in the same area. Variance analysis and multiple comparisons showed that DBH, pruning stubble, anti-trampling pavement, plant spacing, tree pool area, number of lanes and lane direction all had significant effects on the health status of F. pennsylvanica (P<0.05), but tree height had little effect. The health condition of F. pennsylvanica was the best under the following conditions: 60−70 cm DBH, tree body without pruning stubble, tree pool pavement with good water permeability, 5−10 m tree spacing, 1−2 m2 tree pool area, suitable road width, and north-south direction.  Conclusion  F. pennsylvanica street trees in the study area are in a sub-health state on the whole, with good conservation and maintenance potential. However, some of them are seriously damaged and need to be dealt with urgently. In planting and management process, important measures should be taken to ensure the good status of F. pennsylvanica, such as standard pruning, strong water permeability of pavement materials, 5−10 m plant spacing, 1−2 m2 tree pool area, and suitable road conditions . [Ch, 3 fig. 13 tab. 26 ref.]

XIANG Yuyong, TAO Qin, YU Shijun, et al. Analysis of nutrients in larva feces of Heterolocha jinyinhuaphaga[J]. Journal of Zhejiang A&F University, 2020, 37(5): 971-977. DOI: 10.11833/j.issn.2095-0756.20190624
Citation: WANG Yan, YU Yun, LIU Yong, et al. Health assessment and influencing factors of Fraxinus pennsylvanica in Beijing core area[J]. Journal of Zhejiang A&F University, 2022, 39(6): 1340-1349. DOI: 10.11833/j.issn.2095-0756.20220109
  • 洋白蜡Fraxinus pennsylvanica是中国重要的造林绿化树种,也是北京市使用频度最高的行道树种之一,其形体端正,枝叶繁茂而鲜绿,秋叶橙黄,不仅美观性强,还能捕获大量空气中的颗粒物,降低风速,改善城市空气质量[1]。行道树是城市绿地系统的骨干,以“线”的形式联系着城市中分散的“点”和“面”的绿化,构成完整的城市森林生态系统[2]。行道树在缓解城市热岛效应[3]、净化空气[4]、改善城市环境及维护城市内部生物多样性[5]等方面均能起到显著作用,其健康状况不仅反映了城市绿化质量高低,同时也是景观效果及生态效益得以发挥的重要基础。行道树易受人为活动的干扰,生长空间受限及后期疏于管护,易出现树势衰弱、腐烂空洞及病虫害频发等问题,存在较高的安全隐患,限制其景观和生态效益的发挥[6]

    近年来,层次分析法[7-9]、主成分分析[10-11]、综合打分法[12]、树木活力度分级[13]等方法广泛应用于树木健康评价中。不同评价方法各有优劣,评价结果精度也不尽相同,整体呈现出多方法相结合的发展趋势[14-16]。针对行道树开展健康评价研究,尚未形成实用性强的评价体系。基于此,本研究在北京市核心区11条街道洋白蜡行道树调查结果基础上,结合行道树生境、管护措施等多方面因素,运用主成分分析及聚类分析建立洋白蜡行道树健康评价体系,采用判别分析进行结果验证,并基于验证结果建立BP神经网络预测模型,以期了解北京市核心区洋白蜡行道树健康状况,为行道树快速、准确的健康评价提供技术支撑。

    • 北京市位于39°56′N,116°20′E,地处华北平原北部,面积为16 410.54 km2,下辖16个区。研究区位于新规划的首都功能核心区,地处北京城区中部,是全国政治、文化和国际交往中心的核心承载区,面积为92.5 km2,共32个街道183个街区[6, 17]。核心区年平均降水量为545.9 mm,年平均气温为11.4 ℃[17],气候为暖温带半湿润半干旱大陆性季风气候,冬季寒冷干燥,夏季高温多雨,春秋短促,土壤类型主要是棕壤土。

    • 调查于2018年夏季进行。选取分布于北京市核心区多个方位、包含多种道路类型的11条行道树为洋白蜡的街道。随机选取每条街道2侧各至少20株样树,对其生长状况、环境条件等指标进行每木调查,共计调查870株洋白蜡。道路概况见表1

      道路名称单向车道数道路走向道路类型行政区划道路名称单向车道数道路走向道路类型行政区划
      南礼士路二条1东西支路 西城青年沟路  1东西次干路东城
      南纬路   2东西次干路西城先农坛西路 1南北支路 西城
      燕京北街  1东西支路 西城二七剧场路 1南北支路 西城
      史家胡同  1东西支路 东城南花市大街 2南北次干路东城
      西兴隆街  1东西支路 东城德胜门内大街2南北次干路西城
      崇文门西大街3东西主干道东城

      Table 1.  Roads’ profile in study area

    • 将指标中定性数据量化后,与定量数据相结合,共同构建洋白蜡行道树健康评价指标体系。本研究参考汪瑛[2]、邵鹏[6]、余韵[18]以及DB 11/T1692—2019《城市树木健康诊断技术规程》[19]的指标选择和赋值原则,从行道树树冠、树干、根部的生长状况及病虫害、环境条件及管护等多个角度建立指标体系,共选取12个评价指标。所选指标及赋值原则如下:①长势与枯枝(x1):冠形饱满、长势良好无枯枝,赋值为4;树冠缺损和枯枝率≤10%,为轻度,赋值为3;树冠缺损和枯枝率为10%~≤30%,为中度,赋值为2;树冠缺损严重和枯枝率>30 %,为重度,赋值为1。②枝叶茂密程度(x2):非常茂密赋值为3;茂密赋值为2;不茂密赋值为1。③树干异常音(x3):木槌敲击树干,无异常音,赋值为3;轻微异常音,赋值为2;异常音较大,赋值为1。④树干完整度(x4):树干表皮无腐朽损伤,赋值为4;损伤度≤1/3,赋值为3;腐朽损伤度为1/3~≤1/2时,赋值为2;损伤度>1/2,赋值为1。其中,损伤度=树干表皮最大损伤宽度/树干周长。⑤基部空洞程度(x5):树干基部空洞深度占树干基部直径的比例,用钢钎测量。目测无空洞,赋值为4;存在空洞情况,空洞程度≤1/3,赋值为3;空洞程度为1/3~≤1/2时,赋值为2,损伤度>1/2,赋值为1。⑥根部腐朽程度(x6):用钢钎测量。如钢钎无法插入,则无腐朽,赋值为4;插入深度≤5 cm,为轻度,赋值为3;插入深度为5~≤20 cm时,则为中度,赋值为2;深度>20 cm,则为重度腐朽,赋值为1。⑦修枝愈合率(x7):伤口愈合良好的修枝数量占修枝总量的比例。⑧修枝留茬(x8):无修枝留茬,赋值为4;留茬数为1 个即为轻度,赋值为3;留茬数为2~3个即为中度,赋值为2;留茬数为3个以上即为重度,赋值为1。⑨病虫害情况(x9):该指标结合叶片病虫害受害率、树干部病虫害受害率、异色叶比例3个方面进行考量。叶片病虫害受害率=受害叶片数量/叶总量;树干部病虫害受害率=受害宽度/周长;异色叶比例=异常叶色数量/叶总量,其中受害宽度指寄生物危害、病害及虫害等致树干部位受到损伤的最大宽度,周长指受害部位所在位置的枝干周长。无病虫害、无叶色异常、无羽化孔和排粪孔,赋值为4;存在排粪孔或1个羽化孔,叶色异常≤30%,叶片或树干部受害率≤1/3,满足其中任意1项为轻度,赋值为3;存在2~5个羽化孔、叶色异常为30%~≤50%、叶片或树干部受害率为1/3~≤1/2时,满足其中任意1项为中度,赋值为2;存在5个及以上羽化孔,叶色异常比例>50%,叶片或树干部受害率>1/2,满足其中任意1项为重度,赋值为1。⑩环境影响(x10):生长空间充足无遮荫,有防踩铺装且透水性良好,无根系裸露,赋值为4;生长空间稍狭,有防踩铺装但透水性差或存在轻微根系裸露现象,赋值为3;阻碍交通、生长空间受限、无防踩铺装或存在1/3~≤1/2根系裸露,赋值为2;生长空间严重受限(距离建筑物≤1 m)、基部晃动或>1/2根系裸露,赋值为1。⑪冠穴比(x11):$\mathrm{冠}\mathrm{穴}\mathrm{比}={\mathrm{{\text{π}} }\;{\left(\dfrac{1}{4}\displaystyle \sum\limits _{i=1}^{4}{{W_{{\rm{C}}_i}}}\right)}^{2}}/{A_{{\rm{TP}}}}$。⑫冠高比(x12):$\mathrm{冠}\mathrm{高}\mathrm{比}= {\dfrac{1}{2}\displaystyle \sum\limits _{i=1}^{4}{W_{{\rm{C}}_i}}}/{H}$,其中:${W_{{\rm{C}}_i}} $表示东、西、南、北4个方向的冠幅之和,ATP表示树池面积,H表示树高。

    • 本研究所选取的评价指标均为正向指标。将12个指标数据进行归一化处理[20];采用主成分分析法计算相关系数矩阵、特征值和方差贡献率,得到因子载荷矩阵,提取特征值大于1且累计方差贡献率70%以上n个主成分F,得出主成分的表达式yi;再根据主成分的方差贡献率计算权重,从而得到主成分综合健康得分表达式Z;对健康综合得分进行K-means聚类分析,评定树木健康等级,结果共分为健康、亚健康、不健康及濒死4个层次;然后采用判别分析对评价结果进行验证。

    • BP神经网络是基于误差反向传播算法的多层前馈式神经网络(multiple-layer feedforward network),由输入层、隐藏层和输出层构成,其基本思想是学习过程由信号的正向传播与误差的反向传播2个过程组成,不断调整各层神经元的权值及阈值,进行误差校正,使训练的网络达到最优[21],具有较强的非线性适应性和多维函数映射能力[20,22]

      基于MATLAB R2019b工具箱nftool环境设计了1个3层BP神经网络,将归一化后的指标数据及评价结果类别分别作为输入层及输出层。根据经验公式[20],计算隐含层单元数,后通过试错法[21]对比,确定最终隐含层数量。采用均方误差(EMS)与相关系数(R)对模型性能进行评估[23]

    • 采用SPSS 26和MATLAB R2019b分别进行数据统计分析及构建神经网络模型,利用单因素方差分析(one-way ANOVA)和DUNCAN进行多重比较。

    • 样本数据通过KMO (K>0.5)和Bartlett (P<0.05)球形检验,对数据进行主成分分析,选择特征根值>1的前6个主成分作为洋白蜡行道树健康评价的综合指标。结果如表2

      主成分特征根方差贡献率(wi)/%累计方差贡献率(W)/%主成分特征根方差贡献率(wi)/%累计方差贡献率(W)/%
      F11.95216.26416.264F41.1379.47752.449
      F21.71214.26730.531F51.1029.17961.629
      F31.49312.44142.972F61.0088.39670.025
        说明:F1~F6为前6个主成分。i=1, 2, 3

      Table 2.  Principal component characteristic root and variance contribution rates

    • 根据表3,构建前6个主成分与洋白蜡行道树综合指标之间的线性关系yi

      指标主成分指标主成分
      F1F2F3F4F5F6F1F2F3F4F5F6
      x10.492−0.3120.211−0.1680.3510.291x70.781−0.156−0.3980.091−0.1800.019
      x20.4090.1090.047−0.3250.6360.148x80.722−0.280−0.3700.169−0.2600.045
      x30.3280.5330.369−0.206−0.3000.031x90.2720.3190.110−0.3200.222−0.401
      x40.2940.5340.234−0.174−0.4260.002x10−0.1160.0660.2630.062−0.1040.816
      x50.1610.0600.3370.6960.258−0.164x11−0.0410.675−0.4730.0430.2190.111
      x60.2900.3640.3570.5180.108−0.012x12−0.1360.482−0.6440.1770.1250.177

      Table 3.  Factor load matrix

      基于以上,按照公式$ Z=({w}_{1}{y}_{1}+{w}_{2}{y}_{2}+{w}_{3}{y}_{3}+\dots +{w}_{n}{y}_{n})/W $,可得最终模型为:$ Z=(16.264{y}_{1}+ 14.267{y}_{2}+ 12.441{y}_{3}+9.477{y}_{4}+9.179{y}_{5}+8.396{y}_{6})/70.025 $

    • 采用K-means聚类和判别分析分别进行分类和准确性检验,由表4可见:研究区域有洋白蜡健康树341株;亚健康树359株;不健康树146株;濒死树24株。线性判别结果可知:K-means分类结果不存在任何分类争议,正确率达100%。

      健康等级线性判别分析
      健康/株亚健康/株不健康/株濒死/株总计/株
      健康 341000341
      亚健康035900359
      不健康001460146
      濒死 0002424
      总计 34135914624870

      Table 4.  Health discrimination results based on K-means clustering analysis    

    • 结果表明:健康的洋白蜡有341株,占比39.20%,平均得分为1.0222;亚健康的洋白蜡有359株,占比41.26%,平均得分为0.9374;不健康的洋白蜡有146株,占比16.78%,平均得分为0.8377;濒死的洋白蜡有24株,占比2.76%,平均得分为0.6064。每条道路的洋白蜡健康状况见表5。洋白蜡行道树总体平均得分为0.9448,在该评价系统中属于亚健康等级,核心区洋白蜡行道树健康状况整体呈中上等水平,约19%的树木健康状况较差,亟需进一步加强管理。

      道路名称健康树亚健康树不健康树濒死树总数/株
      数量/株比例/%数量/株比例/%数量/株比例/%数量/株比例/%
      南礼士路二条35.402239.301933.901221.4056
      南纬路   2147.701840.90511.4000.0044
      燕京北街  5163.002632.1044.9000.0081
      史家胡同  1016.703456.701321.7035.0060
      西兴隆街  3852.802940.3045.6011.4072
      崇文门西大街2030.304162.1046.1011.5066
      青年沟路  6932.407937.106028.2052.30213
      先农坛西路 1841.902251.2037.0000.0043
      二七剧场路 2440.002236.701321.7011.7060
      南花市大街 4669.701725.8034.5000.0066
      德胜门内大街4137.604945.001816.5010.90109
      总数    34139.2035941.2614616.78242.76870

      Table 5.  Health condition of F. pennsylvanica in 11 roads

    • 将归一化后的12个指标数据作为输入层;将4个评价结果,即健康、亚健康、不健康和濒死,分别与(1,0,0,0) (0,1,0,0) (0,0,1,0) (0,0,0,1)一一对应,作为期望输出。隐含层通过试错法最终确定为10个。确定的网络拓扑结构为“12-10-4”(图1)。

      Figure 1.  Topological structure of BP neural network model

      训练及验证样本从样本数据中随机选取,且无交集。用于训练、验证及测试的样本数分别为608、131及131个。图2为神经网络EMS的曲线图。由图2可知:当训练步数增加时,模型误差曲线逐渐逼近误差的最优值。当模型迭代18次时,验证误差达到0.0104,网络训练稳定且迅速收敛,说明该模型能够满足需求。

      Figure 2.  Error decline curve of BP neural network model

      图3表示4个数据集的期望值及预测值之间的比较情况。当验证误差为0.0104时,建立的神经网络模型中的训练集、验证集、测试集与总体数据的相关系数分别为0.9997、0.9720、0.9976及0.9953,均大于0.9500,表明建立的BP神经网络模型对于训练集、验证集、测试集和总体样本都有很好的逼近能力,能较好地反映洋白蜡行道树12个评价指标与健康评价等级之间关系,故BP神经网络预测模型在洋白蜡健康状况评价中有较好的适用性。

      Figure 3.  Correlation coefficient between measured and predicted values

    • 研究区域洋白蜡行道树胸径平均值为31.88 cm,约68.20%的洋白蜡树胸径集中分布在20~40 cm,总体呈现左偏正态分布的趋势。单因素方差分析(表6)可知:胸径对洋白蜡行道树的健康状况影响达显著水平(P<0.05)。当胸径≥10~<60 cm时,呈现随着胸径增加,综合健康评分随之下降的趋势,但在胸径≥60~<70 cm时,健康综合得分达最大值,且显著高于胸径≥30~<60 cm区间内的3个水平(P<0.05)。

      胸径/cm样本量健康综合得分
      平均值±标准差极小值极大值标准误
      ≥10~<20820.96±0.07 ab0.711.150.01
      ≥20~<303540.96±0.08 ab0.391.150.00
      ≥30~<402410.93±0.11 a0.451.100.01
      ≥40~<501320.93±0.09 a0.451.070.01
      ≥50~<60540.92±0.10 a0.471.080.01
      ≥60~<7070.99±0.06 b0.881.060.02
        说明:不同字母表示不同胸径间差异显著(P<0.05)

      Table 6.  Effects of DBH on the health status of F. pennsylvanica

      调查区域内树高平均值为11.78 m。由表7可知:洋白蜡树高大多数≥6~<18 m,占比95.75%,其他范围分布较少;洋白蜡健康综合得分与树高间的关系整体呈正相关,但并未达到显著水平。

      树高/m样本量健康综合得分
      平均值±标准差极小值极大值标准误
      ≥0~<680.89±0.12 a0.641.000.04
      ≥6~<124940.94±0.10 a0.451.150.00
      ≥12~<183390.95±0.09 a0.391.150.00
      ≥18~<24290.95±0.05 a0.841.050.01
        说明:不同字母表示不同树高间差异显著(P<0.05)

      Table 7.  Effects of tree height on the health status of F. pennsylvanica

    • 不规范的修枝方法会导致树体留茬,而留茬处往往易腐烂进而导致分枝处更大面积腐烂等,如不及时清除,腐烂可蔓延至主干木质部,威胁行道树的健康。由表8可知:存在修枝留茬的洋白蜡行道树占比56.90%,留茬数多为1~3个,3个以上的相对较少。修枝留茬现象对于健康综合得分影响显著(P<0.05),无留茬的洋白蜡健康得分显著高于存在留茬的得分,但留茬数量对行道树健康综合得分无显著影响。

      修枝留茬/个
      样本量健康综合得分
      平均值±标准差极小值极大值标准误
      03750.97±0.08 b0.581.150.00
      12150.94±0.10 a0.391.080.01
      2~31800.92±0.11 a0.451.100.01
      >31000.92±0.08 a0.561.060.01
        说明:不同字母表示不同修枝留茬数间差异显著(P<0.05)

      Table 8.  Effects of pruning stubble on the health status of F. pennsylvanica

      调查区域内,88.16%的树池有防踩实措施,主要包括毛毡子、梅花砖、塑料篦子+塑料+石子等模式。不同铺装材料的透水性差异较大,影响土壤的水分、养分循环及树体的吸收,从而影响树木健康状况。由表9可知:防踩铺装的铺设及材料透水性显著影响洋白蜡健康状况(P<0.05),采用透水性强的铺装材料的洋白蜡行道树的健康综合得分最高,显著高于无铺装和透水性差的铺装。

      防踩铺装样本量健康综合得分
      平均值±标准差极小值极大值标准误
      无铺装   1030.88±0.12 a0.451.050.01
      透水性强铺装7480.95±0.09 b0.391.150.00
      透水性差铺装190.91±0.10 a0.621.030.02
        说明:不同字母表示不同防踩铺装间差异显著(P<0.05)

      Table 9.  Effects of anti-stepping paving condition on the health status of F. pennsylvanica

    • 株距限制着行道树生长范围,过小会导致树木地上和地下生长空间压缩,限制树木树冠及根系伸展,加剧土壤营养资源的争夺,进而导致生长所需矿质养分不足。调查区域内,株距平均值为5.38 m,最大值25.00 m,最小值1.50 m。株距对洋白蜡的健康状况存在显著影响(P<0.05),株距≥0~<5 m时,洋白蜡健康得分为最低,显著低于其他株距范围;当株距≥5 m时,洋白蜡健康得分并不会随着株距增加而产生显著的变化(表10)。

      株距/m样本量健康综合得分
      平均值±标准差极小值极大值标准误
      ≥0~<5680.90±0.11 a0.451.090.01
      ≥5~<107500.95±0.09 b0.391.150.00
      ≥10520.94±0.09 b0.661.050.01
        说明:不同字母表示不同株距间差异显著(P<0.05)

      Table 10.  Effects of plant spacing on the health status of F. pennsylvanica    

      行道树树池是城市道路绿化景观的节点。树池面积过小会导致土壤透水透气不良,树木根系伸展和生长状况不佳;过大的树池又会占用过多城市道路硬质地面,对行人行车造成妨碍。表11可知:调查区域内树池面积大多为≥1~<2 m2,占比88.74%,且在此树池面积下生长的洋白蜡行道树具有最高的健康综合得分,与其余树池面积下的洋白蜡健康得分差异显著(P<0.05)。

      树池面积/m2样本量健康综合得分
      平均值±标准差极小值极大值标准误
      ≥0~<190.90±0.10 a0.640.990.03
      ≥1~<27720.95±0.09 b0.391.150.00
      ≥2890.90±0.11 a0.451.040.01
        说明:不同字母表示不同树池面积间差异显著(P<0.05)

      Table 11.  Effects of tree pool area on the health status of F. pennsylvanica    

      车道数以及车道走向对于洋白蜡健康状况均有显著的影响(P<0.05)。表12可知:洋白蜡健康得分随车道数的增加呈先上升后下降的趋势,当车道数为2条时,洋白蜡的健康状况最佳,显著高于车道数为1条时的健康得分,而与车道数为3条时差异不大;车道走向决定了行道树接受日照的情况,车道为南北走向时,洋白蜡健康状况显著优于东西走向(P<0.05)(表13)。

      单向车
      道数/条
      样本量健康综合得分
      平均值±标准差极小值极大值标准误
      15850.93±0.10 a0.391.120.00
      22190.97±0.07 b0.641.150.00
      3660.95±0.06 ab0.701.070.01
        说明:不同字母表示不同单向车道数间差异显著(P<0.05)

      Table 12.  Effects of the number of unidirectional traffic lanes on the health status of F. pennsylvanica

      车道走向样本量健康综合得分
      平均值±标准差极小值极大值标准误
      东西5920.94±0.10 a0.391.150.00
      南北2780.96±0.07 b0.641.150.00
        说明:不同字母表示不同车道走向间差异显著(P<0.05)

      Table 13.  Effects of directions of traffic lanes on the health status of F. pennsylvanica

    • 本研究采用主成分分析计算北京市核心区洋白蜡行道树的健康综合得分,采用聚类分析进行结果分类,采用判别分析法进行结果验证,分类结果与实际生长状况基本一致。将行道树健康状况评价指标和分类结果构建的拓扑结构为“12-10-4”的BP神经网络预测模型,结果表明:训练集和测试集均方误差均小于0.0104,且训练集、测试集及总集相关系数均>0.990 0,说明BP神经网络模型能准确地反映洋白蜡行道树12个评价指标与健康评价类别的关系。与传统树木健康评价方法相比,该方法避免了庞杂的计算,从而实现对同一地区洋白蜡行道树健康状况的快速、准确评价。

      本研究所选指标覆盖了行道树的树冠、树干及根部的健康状况以及行道树的生长环境条件及管护情况等因素,并引入“冠穴比”及“冠高比”2个指标,将各定性指标进行分级、量化分析,研究结果表明:在所调查的870株洋白蜡中,健康、亚健康、不健康及濒死单株分别占39.20%、41.26%、16.78%及2.76%。可以看出:多数洋白蜡生长状况良好,极少部分洋白蜡生长状况堪忧,处于濒死状态,亟需处理。这一结果与邵鹏[6]、余韵[18]的研究结果基本一致。同时,本研究得到的健康评价结果与洋白蜡实际生长状况一致,表明本研究建立的健康评价模型在北京市核心区洋白蜡行道树健康评价中有很好的适用性。

    • 树木的健康状况与年龄紧密相关[10]。本研究结果表明:洋白蜡胸径为60~<70 cm时,健康状况最好。这可能是由于胸径在此范围的洋白蜡树龄较大,栽植时间长,生长状况稳定的缘故;当胸径为10~<60 cm,呈现随胸径增加,洋白蜡健康得分随之下降的趋势,这与李佳慧等[10]对北京市核心区国槐Sophora japonica行道树的评价结果基本一致。城市行道树受人为活动影响较大。调查发现:约87%的洋白蜡存在不同程度的枯枝状况,约27%的洋白蜡存在由车辆剐蹭、修枝等造成的树皮损伤,约57%的洋白蜡存在由修枝不当造成的不同程度的修枝留茬。建议规范行道树修枝技术,及时去除枯枝并合理处理受损部位。行道树树池防踩铺装材料的不同影响着植物吸收水分及养分的过程。北京核心区树池铺装材料多采用透水砖、石子、毛毡子、铁篦子等。研究结果表明:在铺装材料透水性强树池中生长的洋白蜡,健康状况显著优于不铺装及铺装透水性差条件下的洋白蜡。这一结果与赵丹等[24]、李佳慧等[10]的研究结果基本一致。

      行道树位于道路两侧。城市下垫面性质的改变,压缩了行道树生长的环境空间,阻碍了土壤水分的移动与循环,限制了行道树的生长[25]。株距、树池面积一定程度上决定了行道树的生长及管护空间,而车道数及车道走向决定着道路宽度及树体生长位置,影响着树体接受日照情况[6, 18]。本研究发现:株距、树池面积、车道数及走向都显著影响着行道树的健康状况,当保持洋白蜡行道树株距≥5 m、树池面积≥1~<2 m2、车道数为2条且车道为南北走向时,洋白蜡健康状况较为理想。当株距≥5 m时,随着距离继续增大,洋白蜡的健康状况不再发生显著的变化。随着树池面积和车道数增加,洋白蜡健康得分均呈现先上升后下降的趋势。这可能是随着生长范围及道路宽度的增加,生长空间及光照情况已不再是限制行道树健康的主要因素,而人为踩踏、车辆剐蹭等其他因素影响逐渐占据主导地位。当树池面积<2 m2时,树池的防踩铺装率为91.29%,树池面积≥2 m2时,防踩铺装率仅为60.67%。树池的防踩铺装率下降往往会导致人为踩踏的增加,导致土壤板结,影响行道树生长。这可能是导致树池面积增大但洋白蜡健康状况却显著下降的原因之一。本研究发现:车道走向对行道树健康状况影响显著,南北走向车道上的洋白蜡健康状况显著优于东西走向的洋白蜡。这是由于车道走向决定着树体的生长方位和接受日照情况。车道为东西走向时,南侧及北侧行道树受日照时间差异较大,北侧行道树接受日照时间相比南侧更长,而车道为南北走向时,东西两侧行道树接受日照角度及日照时间基本相同[18, 26],因此整体生长状况更好。道路状况往往还受到周边建筑物密度及高度、道路硬质铺装、车流量等多方面影响,因此仍需进一步探究。

Reference (26)

Catalog

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return