生境概率预测值转换为二元值过程中4个阈值选择方法的比较评估——以珙桐和杉木生境预估为例

doi:10.17521/cjpe.2016.0184

植物生态学报 ›› 2017, Vol. 41 ›› Issue (4): 387-395.DOI: 10.17521/cjpe.2016.0184

• • 下一篇

生境概率预测值转换为二元值过程中4个阈值选择方法的比较评估——以珙桐和杉木生境预估为例

张雷¹, 王琳琳², 刘世荣^3,^*(), 孙鹏森³, 余振⁴, 黄书涛⁵, 张旭东¹

¹中国林业科学研究院林业研究所,国家林业局林木培育重点实验室,北京 100091
²北京农学院,北京 102206
³中国林业科学研究院森林生态环境与保护研究所, 国家林业局森林生态环境重点实验室, 北京 100091
⁴School of Natural Resources, West Virginia University, Morgantown, WV 26506, USA
⁵山东省枣庄市市中区林业局,山东枣庄 277100

收稿日期:2016-05-31 接受日期:2017-01-03 出版日期:2017-04-10 发布日期:2017-05-19
通讯作者: 刘世荣
基金资助:
国家自然科学基金(41301056)、中央公益性院所基本科研业务专项(CAFYBB2014QB006和RIF2012-04)和林业软科学项目(2016-R21)

An evaluation of four threshold selection methods in species occurrence modelling with random forest: Case studies with Davidia involucrata and Cunninghamia lanceolata

Lei ZHANG¹, Lin-lin WANG², Shi-Rong LIU^3,^*(), Peng-Sen SUN³, Zhen YU⁴, Shu-Tao HUANG⁵, Xu- Dong ZHANG¹

¹Key Laboratory of Forest Silviculture of the State Forestry Administration, Research Institute of Forestry, Chinese Academy of Forestry, Beijing 100091, China

²Beijing University of Agriculture, Beijing 102206, China

³Key Laboratory of Forest Ecology and Environment of State Forestry Administration, Institute of Forest Ecology, Environment and Protection, Chinese Academy of Forestry, Beijing 100091, China

⁴School of Natural Resources, West Virginia University, Morgantown, WV 26506, USA
and
⁵Shizhong District Forestry Bureau of Zaozhuang City, Zaozhuang, Shandong 277100, China

Received:2016-05-31 Accepted:2017-01-03 Online:2017-04-10 Published:2017-05-19
Contact: Shi-Rong LIU

摘要/Abstract

摘要：

物种生境模型预测结果通常是概率性的, 然而在具体的保护管理等实践应用过程中通常需要基于二元值(存在/不存在)的分布图, 此时就需要把概率性的预测结果转化为二元值,在此转化过程中就涉及阈值选择问题。此外, 在评估模型预测准确度的时候, 多数评估指标也需要选择一个阈值用于转化概率预测结果, 这个阈值选择对于模型预测准确度也会有极大的影响。然而阈值选择却是物种生境模拟不确定性研究中较少涉及的领域。“随机森林”既可以生成物种生境概率分布图(回归算法)也可以生成二元分布图(分类算法), 然而还未见对两种预测方式的比较研究。该文以珙桐(Davidia involucrata)和杉木(Cunninghamia lanceolata)为例, 分别采用“随机森林”的分类算法和回归算法预测其生境二元分布图和概率分布图, 通过4个不同阈值选择方法(默认值0.5、MaxKappa、MaxTSS和MaxACC)把概率预测图转换为二元分布图, 进而比较分析转换结果对模型预估的影响。珙桐不同阈值选择方法所确立的阈值之间存在显著差异,而杉木没有显著差异; 两物种模型准确度之间没有显著差异; 在预测两物种未来气候条件下的生境面积变化、生境分布区迁移方向和距离以及最适宜海拔分布高度变化时, 二元值转换后的回归算法与分类算法之间存在显著差异,但回归算法中各阈值选择方法之间没有显著差异。空间生境分布图的相似性分析表明MaxKappa和MaxTSS法具有最大相似性, 分类算法与4种阈值选择方法之间具有最大差异。

关键词: 阈值, 概率生境图, 二元生境图, 随机森林, 珙桐, 杉木

Abstract:

Aims Predictive species distribution models (SDMs) are increasingly applied in resource assessment, environmental conservation and biodiversity management. However, most SDM models often yield a predicted probability (suitability) surface map. In conservation and environmental management practices, the information presented as species presence/absence (binary) may be more practical than presented as probability or suitability. Therefore, a threshold is needed to transform the probability or suitability data to presence/absence data. However, little is known about the effects of different threshold-selection methods on model performance and species range changes induced by future climate. Of the numerous SDM models, random forest (RF) can produce probabilistic and binary species distribution maps based on its regression and classification algorisms, respectively. Studies dealing with the comparative test of the performances of RF regression and classification algorisms have not been reported.
Methods Here, the RF was used to simulate the current and project the future potential distributions of Davidia involucrata and Cunninghamia lanceolata. Then, four threshold-setting methods (Default 0.5, MaxKappa, MaxTSS and MaxACC) were selected and used to transform modelled probabilities of occurrence into binary predictions of species presence and absence. Lastly, we investigated the difference in model performance among the threshold selection methods by using five model accuracy measures (Kappa, TSS, Overall accuracy, Sensitivity and Specificity). We also used the map similarity measure, Kappa, for a cell-by-cell comparison of similarities and differences of distribution map under current and future climates.
Important findings We found that the choice of threshold method altered estimates of model performance, species habitat suitable area and species range shifts under future climate. The difference in selected threshold cut-offs among the four threshold methods was significant for D. involucrata, but was not significant for C. lanceolata. Species’ geographic ranges changed (area change and shifting distance) in response to climate change, but the projections of the four threshold methods did not differ significantly with respect to how much or in which direction, but they did differ against RF classification predictions. The pairwise similarity analysis of binary maps indicated that spatial correspondence among prediction maps was the highest between the MaxKappa and the MaxTSS, and lowest between RF classification algorism and the four threshold-setting methods. We argue that the MaxTSS and the MaxKappa are promising methods for threshold selection when RF regression algorism is used for the distribution modeling of species. This study also provides promising insights to our understanding of the uncertainty of threshold selection in species distribution modeling.

Key words: threshold, probability habitat map, binary habitat map, random forest, Davidia involucrata, Cunninghamia lanceolata

张雷, 王琳琳, 刘世荣, 孙鹏森, 余振, 黄书涛, 张旭东. 生境概率预测值转换为二元值过程中4个阈值选择方法的比较评估——以珙桐和杉木生境预估为例. 植物生态学报, 2017, 41(4): 387-395. DOI: 10.17521/cjpe.2016.0184

Lei ZHANG, Lin-lin WANG, Shi-Rong LIU, Peng-Sen SUN, Zhen YU, Shu-Tao HUANG, Xu- Dong ZHANG. An evaluation of four threshold selection methods in species occurrence modelling with random forest: Case studies with Davidia involucrata and Cunninghamia lanceolata. Chinese Journal of Plant Ecology, 2017, 41(4): 387-395. DOI: 10.17521/cjpe.2016.0184

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.plant-ecology.com/CN/10.17521/cjpe.2016.0184

https://www.plant-ecology.com/CN/Y2017/V41/I4/387

图/表 5

参考文献 30

[1]	Bailey SA, Haines-Young RH, Watkins C (2002). Species presence in fragmented landscapes: Modelling of species requirements at the national level.Biological Conservation, 108, 307-316.
[2]	Barbet-Massin M, Jiguet F, Alber CH, Thuiller W (2012). Selecting pseudo-absences for species distribution models: How, where and how many?Methods in Ecology and Evolution, 3, 327-338.
[3]	Breiman L, Cutler A (. Cited: 2016-02-18.
[4]	Carpenter G, Gillison A, Winter J (1993). DOMAIN: A flexible modelling procedure for mapping potential distributions of plants and animals.Biodiversity and Conservation, 2, 667-680.
[5]	Editorial Committee of Vegetation Map of China, Chinese Academy of Sciences (2001). 1:1,000,000 Vegetation Distribution Map of China. Science Press, Beijing. (in Chinese)[中国科学院植被图编辑委员会 (2001). 1:1,000,000中国植被图. 科学出版社, 北京.]
[6]	Engler R, Guisan A, Rechsteiner L (2004). An improved approach for predicting the distribution of rare and endangered species from occurrence and pseudo-absence data.Journal of Applied Ecology, 41, 263-274.
[7]	Fielding AH, Haworth PF (1995). Testing the generality of bird-habitat models.Conservation Biology, 9, 1466-1481.
[8]	Fielding AH, Bell JF (1997). A review of methods for the assessment of prediction errors in conservation presence/ absence models.Environmental Conservation, 24, 38-49.
[9]	Freeman EA, Moisen GG (2008). A comparison of the performance of threshold criteria for binary classification in terms of predicted prevalence and Kappa.Ecological Modelling, 217, 48-58.
[10]	Guisan A, Zimmermann NE (2000). Predictive habitat distribution models in ecology.Ecological Modeling, 135, 147-186.
[11]	Guo Q, Liu Y (2010). ModEco: An integrated software package for ecological niche modeling.Ecography, 33, 637-642.
[12]	Hirzel AH, Hausser J, Perrin N (. Cited: 2016-02-18.
[13]	Jimenez-Valverde A, Lobo J (2007). Threshold criteria for conversion of probability of species presence to either-or presence-absence.Acta Oecologica, 31, 361-369.
[14]	Jin JX, Jiang H, Peng W, Zhang LJ, Lu XH, Xu JH, Zhang XY, Wang Y (2013). Evaluating the impact of soil factors on the potential distribution ofPhyllostachys edulis (bamboo) in China based on the species distribution model. Chinese Journal of Plant Ecology, 37, 631-640. (in Chinese with English abstract)[金佳鑫, 江洪, 彭威, 张林静, 卢学鹤, 徐建辉, 张秀英, 王颖 (2013). 基于物种分布模型评价土壤因子对我国毛竹潜在分布的影响. 植物生态学报, 37, 631-640.]
[15]	Liaw A, Wiener M (2002). Classification and regression by randomForest.R News, 2(3), 18-22.
[16]	Liu C, White M, Newell G (2013). Selecting thresholds for the prediction of species occurrence with presence-only data.Journal of Biogeography, 40, 778-789.
[17]	Liu C, Berry PM, Dawson TP, Pearson RG (2005). Selecting thresholds of occurrence in the prediction of species distributions.Ecography, 28, 385-393.
[18]	Nenzén H, Araújo M (2011). Choice of threshold alters projections of species range shifts under climate change.Ecological Modelling, 222, 3346-3354.
[19]	Norris D (2014). Model thresholds are more important than presence location type: Understanding the distribution of lowland tapir (Tapirus terrestris) in a continuous Atlantic forest of southeast Brazil. Tropical Conservation Science, 7, 529-547.
[20]	O’Hanley JR (2009). NeuralEnsembles: A neural network based ensemble forecasting program for habitat and bioclimatic suitability analysis.Ecography, 32, 89-93.
[21]	R Core Team (. Cited: 2016-02-18.
[22]	Robertson M, Caithness N, Villet M (2001). A PCA-based modelling technique for predicting environmental suitability for organisms from presence records.Diversity and Distributions, 7, 15-27.
[23]	Shao H, Tian JQ, Guo K, Sun OJ (2009). Effects of sample size and species traits on performance of bioclim in predicting geographical distribution of tree species—A case study with 12 deciduous Quercus species indigenous to china. Chinese Journal of Plant Ecology, 33, 870-877. (in Chinese with English abstract)[邵慧, 田佳倩, 郭柯, 孙建新 (2009). 样本容量和物种特征对BIOCLIM模型模拟物种分布准确度的影响——以12个中国特有落叶栎树种为例. 植物生态学报, 33, 870-877.]
[24]	terBraak CJF, Looman CWN (1986). Weighted averaging, logistic regression and the Gaussian response model.Plant Ecology, 65, 3-11.
[25]	Thuiller W (2004). Patterns and uncertainties of species’ range shifts under climate change.Global Change Biology, 10, 2020-2027.
[26]	Wang J, Ni J (2006). Review of modelling the distribution of plant species.Chinese Journal of Plant Ecology, 30, 1040-1053. (in Chinese with English abstract)[王娟, 倪健 (2006). 植物种分布的模拟研究进展. 植物生态学报, 30, 1040-1053.]
[27]	Zhang L, Liu S, Sun P, Wang T, Wang G, Zhang X, Wang L (2015). Consensus forecasting of species distributions: The effects of niche model performance and niche properties.PLOS ONE, 10, e0120056. doi:10.1371/journal.pone.0120056.
[28]	Zhang L, Liu SR, Sun PS, Wang TL (2011a). Comparative evaluation of multiple models of the effects of climate change on the potential distribution of Pinus massoniana. Chinese Journal of Plant Ecology, 35, 1091-1105. (in Chinese with English abstract)[张雷, 刘世荣, 孙鹏森, 王同立 (2011a). 气候变化对马尾松潜在分布影响预估的多模型比较. 植物生态学报, 35, 1091-1105.]
[29]	Zhang L, Liu SR, Sun PS, Wang TL (2011b). Predicting the potential distribution of Phyllostachys edulis with DOMAIN and NeuralEnsembles models. Scientia Silvae Sinicae, 47(7), 20-26. (in Chinese with English abstract)[张雷, 刘世荣, 孙鹏森, 王同立 (2011b). 基于DOMAIN和NeuralEnsembles模型预估中国毛竹潜在分布. 林业科学, 47(7), 20-26.]
[30]	Zhang L, Wang LL, Zhang XD, Liu SR, Sun PS, Wang TL (2014). The basic principle of random forest and its applications in ecology: A case study of Pinus yunnanensis.Acta Ecologica Sinica, 34, 650-659. (in Chinese with English abstract)[张雷, 王琳琳, 张旭东, 刘世荣, 孙鹏森, 王同立 (2014). 随机森林算法基本思想及其在生态学中的应用——以云南松分布模拟为例. 生态学报, 34, 650-659.]

精度指标 Accuracy measure	公式 Formula
总准确度 Overall accuracy	(a +d)/n
敏感度 Sensitivity	a/(a + c)
特异度 Specificity	d/(b + d)
Kappa	$\frac{\left( a\text{+}d \right)-\text{ }\!\![\!\!\text{ }\left( a\text{+}c \right)\left( a\text{+}b \right)\text{+}\left( b\text{+}d \right)\left( c\text{+}d \right)\text{ }\!\!]\!\!\text{ /}n}{n-\text{ }\!\![\!\!\text{ }\left( a\text{+}c \right)\left( a\text{+}b \right)\text{+(}b\text{+}d\text{)(}c\text{+}d\text{) }\!\!]\!\!\text{ /}n}$
真实技巧统计法 True skill statistic (TSS)	Sensitivity + Specificit -1

精度指标 Accuracy measure	公式 Formula
总准确度 Overall accuracy	(a +d)/n
敏感度 Sensitivity	a/(a + c)
特异度 Specificity	d/(b + d)
Kappa	$\frac{\left( a\text{+}d \right)-\text{ }\!\![\!\!\text{ }\left( a\text{+}c \right)\left( a\text{+}b \right)\text{+}\left( b\text{+}d \right)\left( c\text{+}d \right)\text{ }\!\!]\!\!\text{ /}n}{n-\text{ }\!\![\!\!\text{ }\left( a\text{+}c \right)\left( a\text{+}b \right)\text{+(}b\text{+}d\text{)(}c\text{+}d\text{) }\!\!]\!\!\text{ /}n}$
真实技巧统计法 True skill statistic (TSS)	Sensitivity + Specificit -1

	阈值选择方法 Threshold method	阈值 Threshold	Kappa	真实技巧统计法 TSS	总准确度 Overall accuracy	敏感度 Sensitivity	特异度 Specificity
珙桐 Davidia involucrata	默认值0.5 Default 0.5	0.500 (0.000)^a	0.871 (0.024)^a	0.871 (0.024)^a	0.935 (0.012)^a	0.976 (0.019)^a	0.894 (0.025)^a
珙桐 Davidia involucrata	最大总准确度 Maximizing overall accuracy (MaxAcc)	0.476 (0.187)^ab	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.975 (0.021)^a	0.897 (0.027)^a
	最大Kappa Maximizing Kappa (MaxKappa)	0.364 (0.185)^b	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.976 (0.020)^a	0.895 (0.027)^a
	最大真实技巧统计法 Maximizing true skill statistic (MaxTSS)	0.364 (0.185)^b	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.976 (0.020)^a	0.895 (0.027)^a
	随机森林分类 Random forest classification tree (RFCT)	-	0.869 (0.030)^a	0.869 (0.030)^a	0.935 (0.015)^a	0.982 (0.022)^a	0.888 (0.031)^a
杉木 Cunninghamia lanceolata	默认值0.5 Default 0.5	0.500 (0.000)^a	0.903 (0.010)^a	0.903 (0.010)^a	0.951 (0.005)^a	0.962 (0.010)^a	0.941 (0.009)^a
杉木 Cunninghamia lanceolata	最大总准确度 Maximizing overall accuracy (MaxAcc)	0.540 (0.078)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	最大Kappa Maximizing Kappa (MaxKappa)	0.540 (0.078)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	最大TSS Maximizing true skill statistic (MaxTSS)	0.541 (0.076)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	随机森林分类 Random forest classification tree (RFCT)	-	0.905 (0.010)^a	0.905 (0.010)^a	0.952 (0.005)^a	0.961 (0.010)^a	0.943 (0.007)^a

	阈值选择方法 Threshold method	阈值 Threshold	Kappa	真实技巧统计法 TSS	总准确度 Overall accuracy	敏感度 Sensitivity	特异度 Specificity
珙桐 Davidia involucrata	默认值0.5 Default 0.5	0.500 (0.000)^a	0.871 (0.024)^a	0.871 (0.024)^a	0.935 (0.012)^a	0.976 (0.019)^a	0.894 (0.025)^a
珙桐 Davidia involucrata	最大总准确度 Maximizing overall accuracy (MaxAcc)	0.476 (0.187)^ab	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.975 (0.021)^a	0.897 (0.027)^a
	最大Kappa Maximizing Kappa (MaxKappa)	0.364 (0.185)^b	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.976 (0.020)^a	0.895 (0.027)^a
	最大真实技巧统计法 Maximizing true skill statistic (MaxTSS)	0.364 (0.185)^b	0.872 (0.025)^a	0.872 (0.025)^a	0.936 (0.012)^a	0.976 (0.020)^a	0.895 (0.027)^a
	随机森林分类 Random forest classification tree (RFCT)	-	0.869 (0.030)^a	0.869 (0.030)^a	0.935 (0.015)^a	0.982 (0.022)^a	0.888 (0.031)^a
杉木 Cunninghamia lanceolata	默认值0.5 Default 0.5	0.500 (0.000)^a	0.903 (0.010)^a	0.903 (0.010)^a	0.951 (0.005)^a	0.962 (0.010)^a	0.941 (0.009)^a
杉木 Cunninghamia lanceolata	最大总准确度 Maximizing overall accuracy (MaxAcc)	0.540 (0.078)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	最大Kappa Maximizing Kappa (MaxKappa)	0.540 (0.078)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	最大TSS Maximizing true skill statistic (MaxTSS)	0.541 (0.076)^a	0.908 (0.011)^a	0.908 (0.011)^a	0.954 (0.006)^a	0.958 (0.013)^a	0.950 (0.009)^a
	随机森林分类 Random forest classification tree (RFCT)	-	0.905 (0.010)^a	0.905 (0.010)^a	0.952 (0.005)^a	0.961 (0.010)^a	0.943 (0.007)^a

	阈值方法 Threshold	当前适生区 Total habitat area (×10³ km²)	总生境变化比例 Total range change (%)	新生境比例 Habitat gained (%)	生境消失比例 Habitat lost (%)	东向迁移距离 Eastward shift (km)	北向迁移距离 Northward shift (km)	高程迁移距离 Uphill shift (m)
珙桐 Davidia involucrata	Default 0.5	762.8 (34.6)^a	-95.9 (3.8)^a	0.6 (0.9)^a	96.6 (3.0)^a	70.7 (133.2)^a	252.3 (43.5)^a	-341 (211)^a
	MaxAcc	761.1 (69.1)^a	-94.8 (6.4)^a	1.0 (1.3)^a	95.8 (5.1)^a	69.3 (164.5)^a	228.4 (80.5)^a	-336 (244)^a
	MaxKappa	780.1 (69.5)^ab	-94.3 (6.6)^ab	1.1 (1.4)^ab	95.4 (5.3)^ab	50.9 (164.4)^a	241.7 (41.7)^a	-341 (255)^a
	MaxTSS	780.1 (69.5)^ab	-94.3 (6.6)^ab	1.1 (1.4)^ab	95.4 (5.3)^ab	50.9 (164.4)^a	241.7 (41.7)^a	-341 (255)^a
	RFCT	804.3 (27.9)^b	-60.1 (1.9)^b	7.9 (1.1)^b	68.0 (1.8)^b	-236.0 (33.9)^b	134.5 (9.0)^b	242 (63)^b
杉木 Cunninghamia lanceolata	Default 0.5	1β401.5 (14.4)^a	-0.3 (0.1)^ab	0.1 (0.0)^ab	0.4 (0.1)^ab	-129.1 (22.1)^a	68.5 (14.9)^ab	243.3 (37.5)^a
	MaxAcc	1β367.4 (67.9)^a	-0.4 (0.2)^b	0.1 (0.1)^ab	0.5 (0.1)^b	-107.6 (48.2)^ab	57.6 (32.6)^b	238.7 (33.4)^ab
	MaxKappa	1β367.4 (67.9)^a	-0.4 (0.2)^b	0.1 (0.1)^ab	0.5 (0.1)^b	-107.6 (48.2)^ab	57.6 (32.6)^b	238.7 (33.4)^ab
	MaxTSS	1β365.7 (65.8)^a	-0.4 (0.2)^b	0.1 (0.1)^b	0.5 (0.1)^b	-108.0 (48.4)^ab	57.3 (32.5)^b	238.9 (33.5)^ab
	RFCT	1β391.2 (11.0)^a	-0.3 (0.1)^a	0.1 (0.0)^a	0.4 (0.1)^a	-82.0 (26.8)^b	81.5 (12.2)^a	183.0 (38.8)^b

生境概率预测值转换为二元值过程中4个阈值选择方法的比较评估——以珙桐和杉木生境预估为例

An evaluation of four threshold selection methods in species occurrence modelling with random forest: Case studies with Davidia involucrata and Cunninghamia lanceolata

全文

PDF (PC)

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 30

相关文章 15

编辑推荐

Metrics

本文评价

[1]	吴帆, 吴晨, 张宇辉, 余恒, 魏智华, 郑蔚, 刘小飞, 陈仕东, 杨智杰, 熊德成. 增温对成熟杉木人工林不同季节细根生长、形态及生理代谢特征的影响[J]. 植物生态学报, 2023, 47(6): 856-866.
[2]	曹珍, 刘永英, 宋世凯, 张莉娜, 高德. 陆地生境岛屿藓类植物小岛屿效应驱动因素分析——以太行山脉中段山顶为例[J]. 植物生态学报, 2023, 47(1): 65-76.
[3]	朱明阳, 林琳, 佘雨龙, 肖城材, 赵通兴, 胡春相, 赵昌佑, 王文礼. 云南轿子山不同海拔急尖长苞冷杉径向生长动态及其低温阈值[J]. 植物生态学报, 2022, 46(9): 1038-1049.
[4]	谢欢, 张秋芳, 陈廷廷, 曾泉鑫, 周嘉聪, 吴玥, 林惠瑛, 刘苑苑, 尹云锋, 陈岳民. 氮添加促进丛枝菌根真菌和根系协作维持土壤磷有效性[J]. 植物生态学报, 2022, 46(7): 811-822.
[5]	谢欢, 张秋芳, 曾泉鑫, 周嘉聪, 马亚培, 吴玥, 刘苑苑, 林惠瑛, 尹云锋, 陈岳民. 氮添加对杉木苗期磷转化和分解类真菌的影响[J]. 植物生态学报, 2022, 46(2): 220-231.
[6]	赵晏平, 王忠武, 温都日根, 赵玉金, 白永飞. 基于Sentinel-2数据的草地植物功能多样性遥感反演及其与生产力的关系[J]. 植物生态学报, 2022, 46(10): 1234-1250.
[7]	刘超, 李平, 武运涛, 潘胜难, 贾舟, 刘玲莉. 一种基于数码相机图像和群落冠层结构调查的草地地上生物量估算方法[J]. 植物生态学报, 2022, 46(10): 1280-1288.
[8]	周楷玲, 赵玉金, 白永飞. 基于Sentinel-2A数据的东北森林植物多样性监测方法研究[J]. 植物生态学报, 2022, 46(10): 1251-1267.
[9]	薛鹏飞, 李文龙, 朱高峰, 周华坤, 刘陈立, 晏和飘. 黄河首曲玛曲县高寒湿地景观格局演变[J]. 植物生态学报, 2021, 45(5): 467-475.
[10]	丁凯, 张毓婷, 张俊红, 柴雄, 周世水, 童再康. 不同密度杉木林对林下植被和土壤微生物群落结构的影响[J]. 植物生态学报, 2021, 45(1): 62-73.
[11]	张文强, 罗格平, 郑宏伟, 王浩, HAMDI Rafiq, 何惠丽, 蔡鹏, 陈春波. 基于随机森林模型的内陆干旱区植被指数变化与驱动力分析: 以北天山北坡中段为例[J]. 植物生态学报, 2020, 44(11): 1113-1126.
[12]	张雪皎, 高贤明, 吉成均, 康慕谊, 王仁卿, 岳明, 张峰, 唐志尧. 中国北方5种栎属树木多度分布及其对未来气候变化的响应[J]. 植物生态学报, 2019, 43(9): 774-782.
[13]	吕中诚, 康文星, 黄志宏, 赵仲辉, 邓湘雯. 不同林龄杉木组织迁移养分的再利用[J]. 植物生态学报, 2019, 43(5): 458-470.
[14]	邹显花, 胡亚楠, 韦丹, 陈思同, 吴鹏飞, 马祥庆. 磷高效利用杉木对低磷胁迫的适应性与内源激素的相关性[J]. 植物生态学报, 2019, 43(2): 139-151.
[15]	沈芳芳, 李燕燕, 刘文飞, 段洪浪, 樊后保, 胡良, 孟庆银. 长期氮沉降对杉木人工林叶、枝氮磷养分再吸收的影响[J]. 植物生态学报, 2018, 42(9): 926-937.