桃花心木木材定量解剖特征变异性及其识别

doi:10.11707/j.1001-7488.LYKX20220456

摘要/Abstract

摘要：

目的: 基于定量解剖方法探究桃花心木属3种树种木材的构造特征，揭示其不同构造特征的种间和种内变异规律，为木材“种”水平的准确识别提供科学依据。方法: 针对大叶桃花心木、桃花心木和矮叶桃花心木木材标本显微切片，通过光学显微镜采集其横、径、弦三切面微观构造图像，利用图像分析软件分别测量导管分子长度、管孔弦向直径、管孔频率、木纤维长度、木射线宽度、木射线高度和木射线频率7个定量解剖特征，分析其不同构造特征的种间和种内变异规律；采用随机森林算法对木材树种进行识别，比较木材不同定量解剖特征对识别准确率的贡献度。结果: 3种桃花心木木材显微切片图像中管孔、轴向薄壁组织和木射线等构造特征相似，通过人工方式难以区分；除木射线宽度外，其他6组定量解剖特征在种间均存在显著差异；随机森林算法对桃花心木的识别准确率达86.67%，7个定量解剖特征对木材识别结果的贡献度存在差异，其中导管分子长度的平均准确率降低度值（3.956）和平均基尼指数降低度值（6.311）最大，管孔弦向直径次之，木射线宽度的平均准确率降低度值（0.797）和平均基尼指数降低度值（2.175）最小，表明导管分子长度的贡献度最大，木射线宽度的贡献度最小。结论: 采用木材定量解剖方法，揭示3种桃花心木构造特征的种间和种内变异规律以及在木材识别中起关键作用的定量解剖特征，可为实现木材“种”水平的准确识别提供科学依据。

关键词: 桃花心木, 定量解剖, 变异规律, 随机森林, 木材识别

Abstract:

Objective: Based on quantitative wood anatomy (QWA) method, the wood anatomical features of three Swietenia species were analyzed to reveal the patterns of their interspecific and intraspecific variation, and provide a scientific basis for accurate wood identification at the level of “species”. Method: The microscopic images of S. macrophylla, S. mahagoni and S. humilis were collected from transverse, radial and tangential section respectively using a microscope, and the quantitative data of seven anatomical features of wood, i.e. the vessel element length (VEL), tangential diameter of vessel lumina (TVD), vessels per square milimeter (FOV), fiber length (FL), ray width (RW), ray height (RH) and rays per millimeter (RPMM) were measured by image analysis software to investigate the interspecific and intraspecific variation of wood anatomical features. Additionally, the random forest algorithm was used to discriminate three Swietenia species at the species level, and the contributions of different quantitative anatomical features to wood identification were comparatively analyzed. Result: The wood anatomical features of the three Swietenia species were highly closed in vessel, axial parenchyma and wood rays under microscope, therefore it is difficult to distinguish them artificially. However there were significant difference among three Swietenia species in six quantitative wood anatomical features except for RW. The random forests algorithm has a discrimination accuracy of 86.67% for Swietenia mahagoni. VEL showed the highest value of the mean decrease accuracy (3.956) and the mean decrease Gini (6.311), followed by the TVD. RW exhibited the lowlest value of the mean decrease accuracy (0.797) and the mean decrease Gini (2.175). Among the seven wood anatomical features, VEL exhibit the most contribution to the wood identification accuracy, while the RW had the least contribution. Conclusion: This study reveals the interspecific and intraspecific variation of three selected Swietenia species and the key quantitative anatomical features in wood identification based on quantitative wood anatomy, which provided a scientific basis of the accurate wood identification at “species” level.

Key words: Swietenia, quantitative wood anatomy, variation pattern, random forest, wood identification

中图分类号:

S781

刘守佳,何拓,陆杨,焦立超,郭娟,Alex CWiedenhoeft,殷亚方. 桃花心木木材定量解剖特征变异性及其识别[J]. 林业科学, 2024, 60(5): 169-176.

Shoujia Liu,Tuo He,Yang Lu,Lichao Jiao,Juan Guo,Wiedenhoeft Alex C,Yafang Yin. Quantitative Anatomy Analysis on Wood Feature Variability and Wood Identification of Swietenia Species[J]. Scientia Silvae Sinicae, 2024, 60(5): 169-176.

图/表 6

表1

图1

表2

图2

图3

图4

参考文献 0

	何　拓. 2019. 基于机器学习的黄檀属与紫檀属木材识别方法研究. 北京: 中国林业科学研究院.
	He T. 2019. Study on wood identification methods for Dalbergia and Pterocarpus species in combination with machine learning. Beijing: Chinese Academy of Forestry. ［in Chinese］
	何　拓, 刘守佳, 陆　杨, 等. iWood: 基于卷积神经网络的濒危珍贵树种木材自动识别系统. 林业科学, 2021a, 57 (9): 152- 159.
	He T, Liu S J, Lu Y, et al. iWood: an automated wood identification system for endangered and precious tree species using convolutional neural networks. Scientia Silvae Sinicae, 2021a, 57 (9): 152- 159.
	何　拓, 焦立超, 郭　娟, 等. 木材信息学: 发展、应用与展望. 木材科学与技术, 2021b, 35 (4): 15- 24.
	He T, Jiao L C, Guo J, et al. Wood informatics: history of development, application, and prospective trend. Chinese Journal of Wood Science and Technology, 2021b, 35 (4): 15- 24.
	姜笑梅, 殷亚方, 刘　波. 木材树种识别技术现状、发展与展望. 木材工业, 2010, 24 (4): 36- 39.
	Jiang X M, Yin Y F, Liu B. Current status, development and prospect of wood identification technology. China Wood Industry, 2010, 24 (4): 36- 39.
	吕红燕, 冯　倩. 随机森林算法研究综述. 河北省科学院学报, 2019, 36 (3): 37- 41.
	Lü H Y, Feng Q. A review of random forests algorithm. Journal of the Hebei Academy of Sciences, 2019, 36 (3): 37- 41.
	刘晓丽, 王喜明, 姜笑梅, 等. 沙棘材解剖及物理力学性质的研究. 北京林业大学学报, 2004, 26 (2): 84- 89,117.
	Liu X L, Wang X M, Jiang X M, et al. Anatomical and physico-mechanical properties of Hippophae rhamnoides L. Journal of Beijing Forestry University, 2004, 26 (2): 84- 89,117.
	任　宁, 刘一星, 巩翠芝. 木材微观构造与拉伸断裂的关系. 东北林业大学学报, 2008, 36 (2): 33- 35. doi: 10.3969/j.issn.1000-5382.2008.02.012
	Ren N, Liu Y X, Gong C Z. Relationship between wood microstructure and tensile fracture. Journal of Northeast Forestry University, 2008, 36 (2): 33- 35. doi: 10.3969/j.issn.1000-5382.2008.02.012
	汪师孟, 夏美君. 十二种落叶栎木的木材分类. 北京林业大学学报, 1986, 8 (3): 17- 23.
	Wang S M, Xia M J. Classification of twelve oak species by means of wood structure. Journal of Beijing Forestry University, 1986, 8 (3): 17- 23.
	王善武, 沈熙环, 汪师孟. 油松种内材质变异的研究. 北京林业大学学报, 1992, 14 (1): 87- 92.
	Wang S W, Shen X H, Wang S M. Study on variation of wood propertiesin Pinus tabulaeformis. Journal of Beijing Forestry University, 1992, 14 (1): 87- 92.
	卫广扬. 东南亚热带木材识别的若干解剖特征研究. 安徽农学院学报, 1990, 17 (1): 19- 25.
	Wei G Y. Observation on some anatomical features used in identification of tropical wood. Journal of Anhui Agricultural University, 1990, 17 (1): 19- 25.
	Beeckman H. Wood anatomy and trait-based ecology. IAWA Journal, 2016, 37 (2): 127- 151. doi: 10.1163/22941932-20160127
	Beery W H, Ifju G, McLain T E. Quantitative wood anatomy-relating anatomy to transverse tensile strength. Wood and Fiber Science, 1983, 15 (4): 395- 407.
	Bik H M. Let’s rise up to unite taxonomy and technology. PLoS Biology, 2017, 15 (8): e2002231. doi: 10.1371/journal.pbio.2002231
	Cornelius J P, Wightman K E, Grogan J E, et al. 2004. Tropical ecosystems. Swietenia (American Mahogany). Encyclopedia of Forest Sciences, Amsterdam: Elsevier, 1720-1726.
	Hellberg E, Carcaillet C. Wood anatomy of west European Betula: quantitative descriptions and applications for routine identification in paleoecological studies. É coscience, 2003, 10 (3): 370- 379.
	Gasson P, Miller R, Stekel D J, et al. Wood identification of Dalbergia nigra (CITES Appendix I) using quantitative wood anatomy, principal components analysis and naive Bayes classification. Annals of Botany, 2010, 105 (1): 45- 56. doi: 10.1093/aob/mcp270
	Gasson P E, Lancaster C A, Young R, et al. 2021. WorldForestID: addressing the need for standardized wood reference collections to support authentication analysis technologies; a way forward for checking the origin and identity of traded timber. Plants, People, Planet, 3(2): 130−141.
	Godfray H C J. Challenges for taxonomy. Nature, 2002, 417, 17- 19. doi: 10.1038/417017a
	Han H, Guo X L, Yu H. Variable selection using mean decrease accuracy and mean decrease gini based on random forest. 7th IEEE International Conference on Software Engineering and Service Science (ICSESS), 2016, Beijing, China, 219- 224.
	He T, Marco J, Soares R, et al. Machine learning models with quantitative wood anatomy data can discriminate between Swietenia macrophylla and Swietenia mahagoni. Forests, 2019, 11 (1): 36- 49. doi: 10.3390/f11010036
	Ifju G. Quantitative wood anatomy certain geometrical-statistical relationships. Wood and Fiber Science, 1983, 15 (4): 326- 337.
	Li S, Li X, Link R, et al. Influence of cambial age and axial height on the spatial patterns of xylem traits in Catalpa bungei, a ring-porous tree species native to China. Forests, 2019, 10 (8): 662. doi: 10.3390/f10080662
	Liu S J, He T, Wang J J, et al. 2022. Can quantitative wood anatomy data coupled with machine learning analysis discriminate CITES species from their look-alikes? Wood Science and Technology, 56(5): 1567−1583.
	Ovalle-Magallanes B, Madariaga-Mazón A, Navarrete A, et al. Mechanisms of action of antihyperglycemic mexicanolides isolated from Swietenia humillis: in vivo and in silico approaches. Planta Medica, 2016, 81 (S01): S1- S381.
	Panshin A J. Comparative anatomy of the woods of the Meliaceae, sub-family swietenioideae. American Journal of Botany, 1933, 20 (10): 638- 668.
	Ravindran P, Costa A, Soares R, et al. Classification of CITES-listed and other neotropical Meliaceae wood images using convolutional neural networks. Plant Methods, 2018, 14 (1): 1- 10. doi: 10.1186/s13007-017-0271-6
	von Arx G, Crivellaro A, Prendin A L, et al. Quantitative wood anatomy-practical guidelines. Frontiers in Plant Science, 2016, 7, 781.
	von Arx G, Carrer M, Crivellaro A, et al. Q-NET–a new scholarly network on quantitative wood anatomy. Dendrochronologia, 2021, 70, 125890. doi: 10.1016/j.dendro.2021.125890
	Wiedenhoeft A C, Simeone J, Smith A, et al. Fraud and misrepresentation in retail forest products exceeds U. S. forensic wood science capacity. PLoS One, 2019, 14 (7): e0219917. doi: 10.1371/journal.pone.0219917
	Yin Y, Jiang X, Yuan L. 2016. Identification manual of endangered and precious timber species common in trades. Beijing: Science Press.

大叶桃花心木Swietenia macrophylla		桃花心木Swietenia mahagoni		矮叶桃花心木Swietenia humilis
标本号 Specimen code	产地 Place of origin	标本号 Specimen code	产地 Place of origin	标本号 Specimen code	产地 Place of origin
RBw3805	阿根廷Argentina	MADw1203	古巴Cuba	MADw725	哥斯达黎加Costa Rica
RBw3806	美国America	MADw2671	美国America	MADw2414	美国America
RBw3809	阿根廷Argentina	MADw2676	美国America	MADw5109	萨尔瓦多El Salvador
RBw3813	阿根廷Argentina	MADw3905	波多黎各Puerto Rico	MADw8678	危地马拉Guatemala
RBw3815	阿根廷Argentina	MADw3939	美国America	MADw11879	墨西哥Mexico
RBw3821	阿根廷Argentina	MADw4780	美国America	MADw33827	危地马拉Guatemala
RBw3848	波多黎各Puerto Rico	MADw5382	菲律宾Philippines	SJRw3060	美国America
RBw4039	阿根廷Argentina	MADw7264	印尼Indonesia	SJRw4765	利比里亚Liberia
RBw4570	圭亚那Guyana	MADw8567	多米尼加Dominica	SJRw5359	菲律宾Philippines
RBw7214	菲律宾Philippines	MADw8676	美国America	SJRw5894	美国America
MADw2395	哥伦比亚Colombia	MADw8677	美国America	SJRw7484	韩国Korea
MADw3056	委内瑞拉Venezuela	MADw10139	古巴Cuba	SJRw8902	美国America
MADw3575	哥斯达黎加Costa Rica	MADw11039	海地Haiti	SJRw9630	墨西哥Mexico
MADw3576	哥斯达黎加Costa Rica	MADw13405	古巴Cuba	SJRw10364	厄瓜多尔Ecuador
MADw4039	委内瑞拉Venezuela	MADw13753	古巴Cuba	SJRw38312	印尼Indonesia

定量解剖特征 Quantitative wood anatomical features		大叶桃花心木 Swietenia macrophylla	桃花心木 Swietenia mahagoni	矮叶桃花心木 Swietenia humilis
导管分子长度 Vessel element length/μm	平均值±标准差Mean±SD	534±57a	422±53b	439±62b
	最小值Min.	426	347	324
	最大值Max.	618	524	583
管孔弦向直径 Tangential diameter of vessel lumina/μm	平均值±标准差Mean±SD	150±17a	117±15b	144±17a
	最小值Min.	121	86	106
	最大值Max.	182	141	170
管孔频率 Vessels per square milimeter/mm^?2	平均值±标准差Mean±SD	7.2±4.2b	11.3±3.5a	9.7±3.1a
	最小值Min.	5.2	5.5	5.9
	最大值Max.	9.7	19.5	16.6
木纤维长度 Fiber length/μm	平均值±标准差Mean±SD	1 295±99a	1 148±93b	1 217±119ab
	最小值Min.	1 028	993	1 023
	最大值Max.	1423	1317	1423
木射线宽度 Ray width/μm	平均值±标准差Mean±SD	50±11a	55±12a	52±10a
	最小值Min.	33	38	35
	最大值Max.	81	78	65
木射线高度 Ray height/μm	平均值±标准差Mean±SD	389±54.7a	318±42b	331±59b
	最小值Min.	274	262	244
	最大值Max.	522	397	412
木射线频率 Rays per millimeter/mm^?2	平均值±标准差Mean±SD	4.9±0.4b	5.6±0.5b	5.1±0.7a
	最小值Min.	4.0	4.8	4.1
	最大值Max.	5.6	6.5	6.5

[1]	朱诗豪,吴志伟,李政杰,李顺. 赣南马尾松林地表细小死可燃物含水率动态及模型[J]. 林业科学, 2024, 60(5): 158-168.
[2]	周小成, 黄婷婷, 李媛, 肖祥希, 朱洪如, 陈芸芝, 冯芝淸. 结合遥感林龄因子的亚热带森林蓄积量估算方法[J]. 林业科学, 2023, 59(4): 88-99.
[3]	田惠玲,朱建华,何潇,陈新云,简尊吉,李宸宇,郭学媛,黄国胜,肖文发. 基于随机森林模型的东北三省乔木林生物质碳储量预测[J]. 林业科学, 2022, 58(4): 40-50.
[4]	刘鑫源,杨光,宁吉彬,耿道通,于宏洲,邸雪颖. 红松人工林地表可燃物燃烧释放颗粒物质量及影响因素[J]. 林业科学, 2022, 58(3): 97-106.
[5]	郭旭展,陈巧,张晓芳,洪亮,尤媛媛,唐守正,符利勇. 基于无人机高分辨率影像的油松新造林健康树冠提取[J]. 林业科学, 2022, 58(10): 111-120.
[6]	孙忠秋,高金萍,吴发云,高显连,胡杨,高剑新. 基于机载激光雷达点云和随机森林算法的森林蓄积量估测[J]. 林业科学, 2021, 57(8): 68-81.
[7]	由珈齐,李明泽,范文义,全迎,王斌,莫祝坤,祝子枭. 基于高光谱和激光雷达数据的林分类型识别[J]. 林业科学, 2021, 57(5): 119-129.
[8]	潘颖,丁鸣鸣,林杰,代侨,郭赓,崔琳琳. 基于PROSAIL模型和多角度遥感数据的森林叶面积指数反演[J]. 林业科学, 2021, 57(4): 90-106.
[9]	李文博,吕振刚,黄选瑞,张志东. 河北省北部华北落叶松人工林立地指数空间分布预测[J]. 林业科学, 2021, 57(3): 79-89.
[10]	刘波,付跃进,马星霞,卢芸,王麟. 木材生物病害严重区域古建筑木构件树种选择及与生物病害的关系[J]. 林业科学, 2021, 57(12): 108-121.
[11]	袁霄,陈勇平,唐启恒,郭文静. 基于微损探测的古建筑木构件常用树种识别[J]. 林业科学, 2021, 57(12): 122-131.
[12]	宿恒硕,吕军,丁志平,唐彦杰,陈旭东,周强,张哲宇,姚青. 基于改进残差神经网络的木材识别算法[J]. 林业科学, 2021, 57(12): 147-154.
[13]	丁家祺,黄文丽,刘迎春,胡杨. 基于机器学习和多源数据的湘西北森林地上生物量估测[J]. 林业科学, 2021, 57(10): 36-48.
[14]	张雷, 孙鹏森, 刘世荣. 川西亚高山森林不同恢复阶段生长季蒸腾特征[J]. 林业科学, 2020, 56(1): 1-9.
[15]	侯晓静, 明金科, 秦荣水, 朱霁平. 基于随机森林模型的交界域火灾风险分析[J]. 林业科学, 2019, 55(8): 194-200.