基于林层划分的湖南栎类天然次生林断面积生长模型

doi:10.11707/j.1001-7488.20200920

摘要/Abstract

摘要：

目的: 研究复层林林分层次对湖南栎类天然次生林断面积生长的影响，从林层角度构建复层混交异龄林断面积生长模型，反映不同林层的断面积生长规律，为湖南栎类天然次生林的林分生长收获和经营管理提供参考。方法: 以湖南省51块栎类天然次生林样地为研究对象，采用全树高聚类法、国际林联标准和光竞争高度法划分林层，以Schumacher式分别对全林分及上层林、下层林断面积进行拟合，选择最优林层划分方法；考虑总断面积与各林层断面积的相容性问题，通过比例平差法构建相容性林层断面积生长模型。结果: 全树高聚类法、国际林联标准和光竞争高度法的林层划分结果均满足《森林资源规划设计调查技术规程》要求，且不同林层间统计量的F检验结果均显著；Schumacher式能较好模拟全林分及不同林层的断面积生长，其确定系数（R²）均在0.92以上，相比不分层的全林分断面积生长模型，分层后不同林层断面积生长模型R²从0.925 9提高到0.945 5~0.984 6；选择国际林联标准的优势高划分作为栎类天然次生林最优林层划分方法，其断面积生长模型精度最高（R²=0.970 4，MAE=1.458 8，RMSE=2.178 6），与不分层的全林分断面积生长模型相比，R²提高5.0%，平均绝对误差（MAE）、均方根误差（RMSE）分别降低39.8%、26.4%；以总断面积为基础控制的比例平差法能较好解决总断面积与各林层断面积的相容性问题，相比各分项的独立预估模型，上层林R²由0.974 6提高到0.986 8，下层林R²由0.980 2提高到0.988 8。结论: 复层林林分层次对湖南栎类天然次生林断面积生长影响显著，考虑林层效应对研究林层断面积特征、提高断面积建模精度具有重要作用。从林层角度揭示亚热带栎类天然次生林林分断面积生长过程，对栎类天然次生林的林分生长收获和经营管理具有重要意义。

关键词: 林层划分, 断面积生长模型, 比例平差法, 相容性, 栎类

Abstract:

Objective: This paper was implemented to study the effects of forest storey of multi-storied stand on Oaks natural forest basal area growth model in Hunan Province. It would reflect the basal area growth rules of different storeys by building basal area growth models from the perspective of forest storey in order to provide references for the growth harvesting and management of the Oaks natural secondary forest in Hunan Province. Method: After the storey identification by using whole tree height clustering,IUFRO standard and TRSRAT,the model of Schumacher was used to fit the basal area of whole stand,upper layer and lower layer based on the 51 plots of Oaks natural secondary forest in Hunan Province,and the optimal method of storey identification was selected. Considering the compatibility of the total basal area and each forest layer basal area,the compatible storey basal area growth model was built by adjustment in proportion. Result: The results of storey identification by whole tree height clustering,IUFRO standard and TRSRAT were all satisfied with the requirements of technical regulations for forest resources planning and design investigation,and the F testing of different forest layer statistics were significant. The model of Schumacher could better simulate the growth of whole stand and different forest layers,the coefficients of determination (R²) were all above 0.92. Compared with the whole stand basal area growth model without storey identification,the coefficient of determination (R²) of different forest layers with storey identification was increased from 0.925 9 to 0.945 5-0.984 6.The IUFRO standard was selected as the optimal method of storey identification,and the accuracy of the basal area growth model was the highest (R²=0.970 4,MAE=1.458 8,RMSE=2.178 6).Compared with the whole stand basal area growth model without storey identification,the determination coefficient (R²) was increased by 5.0%,the mean absolute error (MAE) and the root mean square error (RMSE) was respectively decreased by 39.8% and 26.4%. The proportional adjustment method based on total basal area could better solve the problem of the compatibility between total basal area and each forest layer basal area. Compared with the independent prediction model of each forest layer,the determination coefficient (R²) of upper layer was increased from 0.974 6 to 0.986 8,and the determination coefficient (R²) of lower layer was increased from 0.980 2 to 0.988 8. Conclusion: The effects of forest storey of multi-storied stand on Oaks natural forest basal area growth model in Hunan Province was significant. It was important to study the characteristics of different forest storeys and improve the modeling accuracy by considering storey effection. The research results were expected to reveal the growth process of forest basal area in subtropical Oaks forest,which was of great significance to growth harvesting and management of the Oaks natural secondary forest.

Key words: storey identification, basal area growth model, adjustment in proportion, compatible, Oaks

中图分类号:

S757

胡松,朱光玉,陈振雄,卢侃,黄朗,刘卓. 基于林层划分的湖南栎类天然次生林断面积生长模型[J]. 林业科学, 2020, 56(9): 184-192.

Song Hu,Guangyu Zhu,Zhenxiong Chen,Kan Lu,Lang Huang,Zhuo Liu. Basal Area Growth Model for Oaks Natural Secondary Forest in Hunan Province Based on Storey Identification[J]. Scientia Silvae Sinicae, 2020, 56(9): 184-192.

图/表 8

表1

表2

表3

表4

表5

图1

表6

图2

参考文献 0

	陈冀南. 2009.马尾松阔叶树混交林空间结构量化分析.福州:福建农林大学硕士学位论文.
	Chen J N. 2009. Studies on spatial structure quantification of the mixed forest of masson pine and broad-leaved trees. Fuzhou: MS thesis of Fujian Agriculture and Forestry University.[in Chinese]
	陈灵芝, 陈清朗, 刘文华. 中国森林多样性及其地理分布. 1997.
	Chen L Z , Chen Q L , Liu W H . Forest diversity and its geographical distribution in China. Beijing: Science Press. 1997.
	陈科屹, 张会儒, 雷相东, 等. 云冷杉过伐林垂直结构特征分析. 林业科学研究, 2017. 30 (3): 450- 459.
	Chen K Y , Zhang H R , Lei X D , et al. Analysis of vertical structure characteristics for spruce-fir over-cutting forest. Forest Research, 2017. 30 (3): 450- 459.
	董利虎, 李凤日, 贾炜玮, 等. 含度量误差的黑龙江省主要树种生物量相容性模型. 应用生态学报, 2011. 22 (10): 2653- 2661.
	Dong L H , Li F R , Jia W W , et al. Compatible biomass models for main tree species with measurement error in Heilongjiang Province of northeast China. Chinese Journal of Applied Ecology, 2011. 22 (10): 2653- 2661.
	杜纪山, 唐守正. 林分断面积生长模型研究评述. 林业科学研究, 1997. 10 (6): 599- 606.
	Du J S , Tang S Z . The review of studies on stand basal area growth model. Forest Research, 1997. 10 (6): 599- 606.
	杜纪山. 林木生长和收获预估模型的研究动态. 世界林业研究, 1999. 12 (4): 19- 22.
	Du J S . The research trends of forest growth and harvest prediction model. World Forestry Research, 1999. 12 (4): 19- 22.
	段爱国, 张建国, 孙洪刚, 等. 林分断面积生长模拟理论与技术研究. 世界林业研究, 2013. 26 (2): 43- 47.
	Duan A G , Zhang J G , Sun H G , et al. Research progress of growth simulation theories and technologies for stand basal area. World Forestry Research, 2013. 26 (2): 43- 47.
	符利勇, 雷渊才, 孙伟, 等. 不同林分起源的相容性生物量模型构建. 生态学报, 2014a. 34 (6): 1461- 1470.
	Fu L Y , Lei Y C , Sun W , et al. Development of compatible biomass models for trees from different stand origin. Acta Ecologica Sinica, 2014a. 34 (6): 1461- 1470.
	符利勇, 雷渊才, 曾伟生. 几种相容性生物量模型及估计方法的比较. 林业科学, 2014b. 50 (6): 42- 54.
	Fu L Y , Lei Y C , Zeng W S . Comparison of several compatible biomass models and estimation approaches. Scientia Silvae Sinicae, 2014b. 50 (6): 42- 54.
	李春明. 利用非线性混合模型进行杉木林分断面积生长模拟研究. 北京林业大学学报, 2009. 31 (1): 44- 49.
	Li C M . Simulating basal area growth of fir plantations using a nonlinear mixed modeling approach. Journal of Beijing Forestrty University, 2009. 31 (1): 44- 49.
	李俊清. 森林生态学. 北京:高等教育出版社, 2006. 291
	Li J Q . Forest ecology. Beijing:Higher Education Press, 2006. 291
	孟宪宇. 测树学. 北京:中国林业出版社, 2006. 44
	Meng X Y . Forest measurement. Beijing:China Forestry Publishing House, 2006. 44
	戎建涛. 2013.中亚热带天然阔叶用材林目标树经营研究.北京:中国林业科学研究院博士学位论文.
	Rong J T. 2013. Study on target tree management of natural broad-leavead timber forest in mid-subtropical zone. Beijing: PhD thesis of Chinese Academy of Forestry.[in Chinese]
	唐守正, 张会儒, 胥辉. 相容性生物量模型的建立及其估计方法研究. 林业科学, 2000. 36 (增): 19- 27.
	Tang S Z , Zhang H R , Xu H . Study on establish and estimate method of compatible biomass model. Scientia Silvae Sinicae, 2000. 36 (sup.): 19- 27.
	唐守正, 杜纪山. 利用树冠竞争因子确定同龄间伐林分的断面积生长过程. 林业科学, 1999. 35 (6): 35- 41.
	Tang S Z , Du J S . Determining basal area growth process of thinnined even-aged stands by crown competition factor. Scientia Silvae Sinicae, 1999. 35 (6): 35- 41.
	唐守正. 广西大青山马尾松全林整体生长模型及其应用. 林业科学研究, 1994. 4 (增): 8- 13.
	Tang S Z . Integrated stand growth model and its application of masson pine in Guangxi Daqingshan. Forest Research, 1994. 4 (sup.): 8- 13.
	吴征镒. 中国植被. 北京: 科学出版社. 1995.
	Wu Z Y . Vegetation of China. Beijing: Science Press. 1995.
	郑景明, 张春雨, 周金星, 等. 云蒙山典型森林群落垂直结构研究. 林业科学研究, 2007. (6): 768- 774.
	Zheng J M , Zhang C Y , Zhou J X , et al. Study on vertical structure of forest communities in Yunmengshan. Forest Research, 2007. (6): 768- 774.
	朱光玉, 胡松, 符利勇. 基于哑变量的湖南栎类天然林林分断面积生长模型. 南京林业大学学报, 2018. 42 (2): 155- 162.
	Zhu G Y , Hu S , Fu L Y . Basal area growth model for Quercus spp natural forest in Hunan province based on dummy variable. Journal of Nanjing Forestrty University, 2018. 42 (2): 155- 162.
	庄崇洋, 黄清麟, 马志波, 等. 林层划分方法综述. 世界林业研究, 2014. 27 (6): 34- 40.
	Zhuang C Y , Huang Q L , Ma Z B , et al. Review on defining methods for canopy stratification. World Forestry Research, 2014. 27 (6): 34- 40.
	庄崇洋, 黄清麟, 马志波, 等. 中亚热带天然阔叶林林层划分新方法——最大受光面法. 林业科学, 2017a. 53 (3): 1- 11.
	Zhuang C Y , Huang Q L , Ma Z B , et al. A new method of storey identification of natural broad-leaved forests in mid-subtropical zone-maximum light receiving plane. Scientia Silvae Sinicae, 2017a. 53 (3): 1- 11.
	庄崇洋, 黄清麟, 马志波, 等. 典型中亚热带天然阔叶林各林层直径分布及其变化规律. 林业科学, 2017b. 53 (4): 18- 27.
	Zhuang C Y , Huang Q L , Ma Z B , et al. Diameter distribution in each storey and law of typical natural broad-leaved forest in mid-subtropical zone. Scientia Silvae Sinicae, 2017b. 53 (4): 18- 27.
	庄崇洋, 黄清麟, 马志波, 等. 典型中亚热带天然阔叶林各林层树高胸径关系研究. 林业科学研究, 2017c. 30 (3): 479- 485.
	Zhuang C Y , Huang Q L , Ma Z B , et al. Study on height-diameter relationship in each stratum of typical natural broadleaved forests in mid-subtropics. Forest Research, 2017c. 30 (3): 479- 485.
	庄崇洋, 黄清麟, 马志波, 等. 典型中亚热带天然阔叶林林层特征及蓄积估计研究. 林业科学研究, 2017d. 30 (4): 559- 565.
	Zhuang C Y , Huang Q L , Ma Z B , et al. Study on characteristics and volume estimation in each stratum of typical natural broadleaved forest in mid-subtropics. Forest Research, 2017d. 30 (4): 559- 565.
	庄崇洋. 2016.中亚热带天然阔叶林林层特征研究.北京:中国林业科学研究院博士学位论文.
	Zhuang C Y. 2016. Study on the strata charcteristics of natural broad-leaved forests in middle-subtropical zone. Beijing: PhD thesis of Chinese Academy of Forestry.[in Chinese]
	Halle F, Oldeman R A A, Tomlinson P B. 1978. Tropical trees and forests: an architectural analysis. Berlin, Heidelberg: Springer-Verlag, 331-333.
	Newman I V . Locating strata in tropical rain forests. Journal of Ecology, 1954. 42 (1): 218- 219. doi: 10.2307/2256988
	Rodríguez F , Pemán J , Aunós Á . A reduced growth model based on stand basal area:a case for hybrid poplar plantations in northeast Spain. Forest Ecology and Management, 2010. 259 (10): 2093- 2102. doi: 10.1016/j.foreco.2010.02.021

变量 Variable	建模数据Modelling data				检验数据Validation data
变量 Variable	最小值 Min.	最大值 Max.	平均值 Mean	标准差 SD	最小值 Min.	最大值 Max.	平均值 Mean	标准差 SD
G/(m²·hm^-2)	14.4	59.4	30.1	10.7	14.6	47.7	30.3	10.4
N/(tree·hm^-2)	722.0	4 118.0	1 518.2	810.7	961.5	3 913.8	1 556.3	841.3
D_g/cm	9.1	26.6	16.7	3.9	9.6	21.4	16.4	3.7
H/cm	9.2	23.2	13.7	2.7	10.5	17.3	12.8	1.8
V/m³	69.9	368.0	179.7	52.4	80.1	362.8	182.3	58.4
P	0.63	0.90	0.79	0.06	0.7	0.88	0.78	0.06
HT/m	11.1	27.5	16.4	3.2	13.1	20.5	15.2	2.1
T/a	18	72	38	14.0	21	56	35.7	11.8

划分方法 Division method	变量 Variable	主林层Main storey				次林层Second storey
划分方法 Division method	变量 Variable	最小值 Min.	最大值 Max.	平均值 Mean	标准差 SD	最小值 Min.	最大值 Max.	平均值 Mean	标准差 SD
全树高聚类法 Whole tree height clustering	D_g/cm	10.2	69.1	22.8	9.3	5.7	19.7	11.6	2.9
	H/cm	10.5	27.5	15.3	3.1	6.5	12.9	9.3	1.4
	V/m³	51.5	271.6	150.1	57.3	30.5	107.1	48.9	19.4
	P	0.21	0.76	0.47	0.14	0.20	0.74	0.32	0.14
国际林联标准 IUFRO standard	D_g/cm	9.9	58.6	23.2	8.9	5.5	20.3	11.8	2.9
	H/cm	10.7	33.3	16.7	4.7	6.9	19.7	10.2	2.3
	V/m³	47.4	280.4	140.3	56.1	32.0	139.3	58.3	26.5
	P	0.29	0.75	0.45	0.15	0.20	0.78	0.34	0.15
光竞争高度法 TRSRAT	D_g/cm	10.6	69.1	22.9	9.5	5.8	20.3	11.9	3.2
	H/cm	11.1	29.7	16.3	3.9	6.9	14.7	9.8	1.8
	V/m³	38.9	279.0	140.3	57.2	35.2	132.5	56.2	26.8
	P	0.23	0.75	0.46	0.16	0.21	0.70	0.33	0.15

划分方法 Division method	蓄积 Volume/(m³·hm^-2)			平均胸径 Mean DBH/cm			平均高 Average height/m				郁闭度 Crown density
划分方法 Division method	上层林 Upper layer	下层林 Lower layer	F	上层林 Upper layer	下层林 Lower layer	F	上层林 Upper layer	下层林 Lower layer	F	高差/% Relative elevation	上层林 Upper layer	下层林 Lower layer	F
全树高聚类法 Whole tree height clustering	161.3±83.5	43.0±12.9	**	22.8±9.3	11.6±2.9	**	15.3±3.1	9.3±1.4	**	40.0±7.5	0.41±0.17	0.38±0.17	**
国际林联标准 IUFRO standard	146.6±81.6	57.2±24.9	**	24.1±8.9	12.2±3.2	**	15.8±3.2	9.8±1.5	**	37.0±8.0	0.42±0.16	0.37±0.16	**
光竞争高度法 TRSRAT	146.1±88.2	57.3±25.4	**	23.6±9.3	12.1±3.2	**	15.6±3.0	9.7±1.7	**	37.0±8.2	0.47±0.14	0.32±0.14	**

划分方法 Division method	模型类型 Model type	参数估计 Parameter estimates						建模精度 Modeling accuracy			检验结果 Validation result
划分方法 Division method	模型类型 Model type	b₀	b₁	b₂	b₃	b₄	b₅	R²	MAE	RMSE	MAE	RMSE
不分层 No stratification	全林分 Whole stand	2.814 4	-47.004 2	1.203 0	-6.553 2	-0.238 6	17.976 3	0.925 9	2.337 5	2.953 3	2.769 4	2.996 4
全树高聚类法 Whole tree height clustering	上层林 Upper layer	-0.234 5	106.498 9	1.057 3	4.979 3	0.954 3	-42.116 7	0.964 5	1.309 3	1.706 3	0.434 5	0.536 1
全树高聚类法 Whole tree height clustering	下层林 Lower layer	3.571 4	-68.343 1	1.426 3	-7.654 3	-0.511 1	24.996 1	0.945 5	1.724 5	2.175 0	1.501 0	2.414 1
国际林联标准 IUFRO standard	上层林 Upper layer	1.991 4	-11.884 2	1.274 5	-11.917 3	0.142 8	4.538 5	0.974 6	1.021 7	1.308 9	1.211 5	1.403 1
国际林联标准 IUFRO standard	下层林 Lower layer	3.300 7	-42.158 9	1.087 7	-0.717 6	-0.376 5	14.003 5	0.980 2	0.599 7	0.862 3	0.376 1	0.499 3
光竞争高度法 TRSRAT	上层林 Upper layer	1.124 8	31.098 1	1.017 4	-3.502 3	0.524 4	-13.767 4	0.967 8	1.265 9	1.630 8	1.026 9	1.196 2
光竞争高度法 TRSRAT	下层林 Lower layer	3.928 0	-57.403 8	1.058 4	-0.780 6	-0.601 6	19.575 6	0.984 6	0.593 3	0.761 4	0.648 4	0.908 4

划分方法Division method	样本量Sample capacity	R²	MAE	RMSE
不分层No stratification	51	0.924 2	2.422 2	2.961 8
全树高聚类法Whole tree height clustering	51	0.943 0	1.887 3	2.568 8
国际林联标准IUFRO standard	51	0.959 0	1.458 8	2.178 6
光竞争高度法TRSRAT	51	0.947 5	1.694 2	2.464 6