Biomass allometric equation (BAE) considered as a simple and reliable method in the estimation of forest biomass and carbon was used widely. In China, numerous studies focused on the BAEs for coniferous forest and pure broadleaved forest, and generalized BAEs were frequently used to estimate the biomass and carbon of mixed broadleaved forest, although they could induce large uncertainty in the estimates. In this study, we developed the speciesspecific and generalized BAEs using biomass measurement for 9 common broadleaved trees (Castanopsis fargesii, C. lamontii, C. tibetana, Lithocarpus glaber, Sloanea sinensis, Daphniphyllum oldhami, Alniphyllum fortunei, Manglietia yuyuanensis, and Engelhardtia fenzlii) of subtropical evergreen broadleaved forest, and compared differences in speciesspecific and generalized BAEs. The results showed that D (diameter at breast height) was a better independent variable in estimating the biomass of branch, leaf, root, aboveground section and total tree than a combined variable (D2H) of D and H (tree height), but D2H was better than D in estimating stem biomass. R2 (coefficient of determination) values of BAEs for 6 species decreased when adding H as the second independent variable into Donly BAEs, where R2 value for S. sinensis decreased by 5.6%. Compared with generalized D and D2Hbased BAEs, standard errors of estimate (SEE) of BAEs for 8 tree species decreased, and similar decreasing trend was observed for different components, where SEEs of the branch decreased by 13.0% and 20.3%. Therefore, the biomass carbon storage and its dynamic estimates were influenced largely by tree species and model types. In order to improve the accuracy of the estimates of biomass and carbon, we should consider the differences in tree species and model types.
全 文 :亚热带常绿阔叶林 9个常见树种的
生物量相对生长模型∗
左舒翟1 任 引1 翁 闲2 丁洪峰2 罗云建1∗∗
( 1中国科学院城市环境研究所城市环境与健康重点实验室, 福建厦门 361021; 2福建省顺昌埔上林场, 福建南平 353205)
摘 要 生物量相对生长模型作为一种简便且有效的生物量估算方法,已得到广泛应用.国
内偏重于针叶林或阔叶纯林的生物量相对生长模型研究,而在估算多树种阔叶林的生物量
时,一般选用混合物种的生物量相对生长模型,这会导致估算结果产生较大误差.本文在亚热
带常绿阔叶林典型分布区随机设置了 33块样地,针对栲树、鹿角锥、钩锥、石栎、猴欢喜、虎皮
楠、赤杨叶、乳源木莲和少叶黄杞 9个常见的树种,构建了单物种及混合物种的生物量相对生
长模型,并探讨了单物种模型及混合物种模型间估算误差的差异.结果表明: 以 D(胸径)和
D2H(胸径的平方乘以树高)为自变量,分别构建混合物种模型,其中树枝、树叶、树根、地上和
整株生物量是以 D为自变量的模型为优,但树干生物量是以 D2H 为自变量的模型为优.将树
高引入以 D为自变量的单物种模型后,6个树种单物种模型的解释能力呈不同程度的下降趋
势,最高下降 5.6%(猴欢喜) .与以 D 和 D2H 为自变量的混合物种模型相比,8 个树种单物种
模型的 SEE(估计值的标准差)出现下降;对不同器官而言,其单物种模型的 SEE 不同程度地
下降,最高达 13.0%和 20.3%(树枝) .不考虑种间和模型形式间的差异,将会严重影响生物量
碳库及其动态评估的准确性.因此,为提高生物量估算的准确性,应综合考虑种间和模型形式
间的差异.
关键词 亚热带; 常绿阔叶林; 生物量估算; 生物量相对生长模型
∗林业公益性行业科研专项(201304205)、福建省科技计划重点项目(2013Y0083)、宁波市科技计划项目(2013A610164)和福建省自然科学基
金项目(2014J05044)资助.
∗∗通讯作者. E⁃mail: yjluo@ iue.ac.cn
2014⁃07⁃07收稿,2014⁃12⁃15接受.
文章编号 1001-9332(2015)02-0356-07 中图分类号 S718.5 文献标识码 A
Biomass allometric equations of nine common tree species in an evergreen broadleaved forest
of subtropical China. ZUO Shu⁃di1, REN Yin1, WENG Xian2, DING Hong⁃feng2, LUO Yun⁃
jian1 ( 1Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese
Academy of Sciences, Xiamen 361021, Fujian, China; 2Shunchang Pushang Forest Farm of Fujian
Province, Nanping 353205, Fujian, China) . ⁃Chin.J.Appl.Ecol., 2015, 26(2): 356-362.
Abstract: Biomass allometric equation (BAE) considered as a simple and reliable method in the
estimation of forest biomass and carbon was used widely. In China, numerous studies focused on the
BAEs for coniferous forest and pure broadleaved forest, and generalized BAEs were frequently used
to estimate the biomass and carbon of mixed broadleaved forest, although they could induce large
uncertainty in the estimates. In this study, we developed the species⁃specific and generalized BAEs
using biomass measurement for 9 common broadleaved trees (Castanopsis fargesii, C. lamontii, C.
tibetana, Lithocarpus glaber, Sloanea sinensis, Daphniphyllum oldhami, Alniphyllum fortunei,
Manglietia yuyuanensis, and Engelhardtia fenzlii) of subtropical evergreen broadleaved forest, and
compared differences in species⁃specific and generalized BAEs. The results showed that D (diame⁃
ter at breast height) was a better independent variable in estimating the biomass of branch, leaf,
root, aboveground section and total tree than a combined variable (D2 H) of D and H ( tree
height), but D2H was better than D in estimating stem biomass. R2(coefficient of determination)
values of BAEs for 6 species decreased when adding H as the second independent variable into D⁃
only BAEs, where R2 value for S. sinensis decreased by 5.6%. Compared with generalized D⁃ and
应 用 生 态 学 报 2015年 2月 第 26卷 第 2期
Chinese Journal of Applied Ecology, Feb. 2015, 26(2): 356-362
D2H⁃based BAEs, standard errors of estimate (SEE) of BAEs for 8 tree species decreased, and
similar decreasing trend was observed for different components, where SEEs of the branch decreased
by 13.0% and 20.3%. Therefore, the biomass carbon storage and its dynamic estimates were influ⁃
enced largely by tree species and model types. In order to improve the accuracy of the estimates of
biomass and carbon, we should consider the differences in tree species and model types.
Key words: subtropical zone; evergreen broadleaved forest; biomass estimation; biomass allomet⁃
ric equation.
森林作为陆地生态系统的主体,在维护区域生
态环境、调节全球碳平衡、减缓大气中温室气体浓度
上升等方面具有不可替代的作用.森林生物量固定
了陆地植被碳储量的 82.5%[1],其不仅是陆地生态
系统碳收支评估中的重要指标[2],也是研究诸多林
业和生态问题 (如物质循环、能量流动等)的基
础[3] .目前,针对森林生物量已经开展了大量研究,
但多集中于地上生物量的测定和估算方面[4] .由于
地下生物量取样度难,导致地下生物量测定和估算
研究相对较少.此外,国内研究偏重于针叶林或阔叶
纯林,缺乏多树种常绿阔叶林的研究,尤其是天然
林[4] .
森林生物量可以通过生物量相对生长模型、生
物量估算参数、遥感反演等途径进行估算[5-6] .其中,
生物量相对生长模型将简单易测的指标,如胸径、树
高等与生物量结合起来,为林木生物量的估算提供
了一种简便且有效的方法[7],可以用于验证生态过
程模型和遥感反演的结果[8-9],目前已得到最为广
泛的应用.以往估算多树种阔叶林的生物量时,往往
采用混合物种的生物量相对生长模型[10-12] .但是,
不同物种或同一物种不同地点的树木生物量分配格
局、形态结构、木材密度等存在较大差异,导致它们
的生物量相对生长模型也存在较大差异[13-14] .如果
采用混合物种或混合地点的相对生长模型估算多树
种常绿阔叶林的生物量,将会引起较大的误差[15] .
随着对估算结果准确性的要求日益提高,为了进一
步提高森林生物量估算的准确性,有必要构建典型
森林类型常见树种的生物量相对生长模型.此外,以
往研究多树种阔叶林生物量时,往往只关注森林的
地上部分[16],而对地下部分关注较少,以致缺乏地
下生物量的相对生长模型.
亚热带常绿阔叶林是亚热带大陆东岸季风气候
区的典型植被类型[17] .它在我国的分布较为广泛,
不仅是我国森林碳储量重要的组成部分,也被认为
是具有较大碳增汇潜力的森林类型[18] .我国拥有详
细的各级森林资源清查资料(包含树木胸径、树高
等统计数据),如果建立了亚热带常绿阔叶林常见
树种的生物量相对生长模型,则能为亚热带常绿阔
叶林生物量的准确估算提供科学依据.因此,本文在
亚热带常绿阔叶林的典型分布区随机设置了 33 块
样地,选取 9个常见树种,构建单物种及混合物种的
生物量相对生长模型,探讨单物种模型及混合物种
模型间估算误差的差异,以期提高亚热带常绿阔叶
林生物量碳库的准确性,进而提升我国森林碳收支
评估的水平.
1 研究地区与研究方法
1 1 研究区概况
研究区位于福建省顺昌县埔上林场(26°56′8″ N,
117°47′5″ E),处于低山丘陵地带,海拔 200~400 m,
坡度多在 20° ~ 30°.属于典型的亚热带海洋性季风
气候,年均温 18.5 ℃,年均降雨量 1880 mm,年均日
照时数 1699 h,无霜期>260 d.土壤以红壤为主,土
层深厚,理化性能良好.
1 2 研究方法
1 2 1样木选择 在踏查林区的基础上,选择具有
代表性的地段,随机设置 20 m×20 m的样地共计 33
块,即 15~20 a的样地 9块、20~30 a 的样地 12 块、
30 ~ 40 a 的样地 12 块,林分密度为 1000 ~ 3800
株·hm-2 .为了能够更好地建立亚热带常绿阔叶林
常见树种的生物量相对生长模型,样木的选择依据
李宁等[19]的选择标准,即通过调查样地的林木组
成,筛选优势种;尽量找到各树种不同径级的个体.
根据以上原则,本研究选定 9个优势树种,共 140 株
样木.这 9个树种分别为栲树(Castanopsis fargesii)、
鹿角锥(C. lamontii)、钩锥(C. tibetana)、石栎(Lith⁃
ocarpus glaber)、猴欢喜 ( Sloanea sinensis)、虎皮楠
(Daphniphyllum oldhami)、赤杨叶(Alniphyllum fortu⁃
nei)、乳源木莲(Manglietia yuyuanensis)和少叶黄杞
(Engelhardtia fenzlii).各树种样木的基本情况见
表 1.
1 2 2样木生物量的测定 采用国家林业局发布的
《国家森林资源连续清查森林生物量模型建立暂行
7532期 左舒翟等: 亚热带常绿阔叶林 9个常见树种的生物量相对生长模型
办法(试行)》 [20],测定样木的生物量.具体如下:
1)树干:将样木伐倒后,采用分层切割法,按
1~2 m区分段将树干进行切割(即树高≥10 m 时,
采用 2 m区分段;树高<10 m时,采用 1 m区分段),
现场测定树干(含树皮)的鲜质量.
2)树枝和树叶:将树冠分上、中、下 3 层,按顺
序测定每个带叶枝条的鲜质量,计算每层的平均带
叶枝鲜质量.按各层平均带叶枝鲜质量分别选取 3
个标准枝,对标准枝摘叶后,分别测定枝质量和叶质
量,根据每层标准枝鲜质量推算出各层枝、叶的鲜质
量,进而推算整个树冠枝和叶的鲜质量.
3)树根:以树干基部为中心,向四周小心清除
表土,确定树根的水平伸展范围,然后从表层开始逐
渐向下挖掘,使树根(含根桩)全部露出,最后将树
根全部挖出,并称其鲜质量.
采集各器官的鲜样 500 g 左右带回实验室,在
85 ℃下烘干至恒量,测定样品的含水率,然后将各
器官的鲜质量换算成干质量(即生物量).各树种样
木整株生物量的情况见表 1.
1 3 数据处理
幂函数 y = αxβ 是描述生物体不同器官大小或
属性间相对生长关系且具有植物生长机理的模
型[21-22] .式中,α和 β为模型系数.研究表明,胸径和
树高是推算树木生物量最重要的变量[17] .在查阅大
量文献[9,17,23-24]的基础上,选用以下2种最具代表
表 1 9种树种样木信息的统计
Table 1 Statistics of sample trees for the nine tree species
树种
Species
样本量
Sample
size
胸径
D
(cm)
树高
H
(m)
生物量
Biomass
(kg)
栲树 Castanopsis fargesii 21 5.2~35.5 6.3~19.5 6.2~1204.2
鹿角锥 Castanopsis lamontii 15 5.8~31.7 6.7~17.7 13.2~895.3
钩锥 Castanopsis tibetana 15 5.2~30.6 5.3~13.4 10.4~586.2
石栎 Lithocarpus glaber 15 5.4~23.0 5.9~14.8 10.9~307.8
猴欢喜 Sloanea sinensis 15 5.8~26.0 2.9~13.8 11.1~285.9
虎皮楠 Daphniphyllum oldhami 14 6.7~22.9 5.9~17.2 18.8~299.1
赤杨叶 Alniphyllum fortunei 15 5.1~21.0 8.1~21.3 10.5~301.4
乳源木莲 Manglietia yuyuanensis 15 5.6~25.0 6.2~15.9 8.1~299.2
少叶黄杞 Engelhardtia fenzlii 15 5.5~21.0 5.2~16.1 9.5~209.3
性的幂函数模型,研究各树种的生物量与胸径及树
高的数量关系.
W=aDb
W=a(D2H) b
式中:W 为器官或整株的生物量 ( kg);D 为胸径
(cm);H为树高(m);a 和 b 为拟合系数.模型效果
的评价采用 R2(决定系数)和 SEE(估计值的标准
差)指标.数据分析和图形制作分别采用 SPSS 16.0
和 OriginPro 8.6软件.
2 结果与分析
不考虑树种差异时,以 D 和 D2H 为自变量,构
建了混合物种的生物量相对生长模型(P<0.001)
(图1、表2) .结果表明,树枝、树叶、树根、地上和整
图 1 以胸径为自变量的混合物种生物量相对生长模型
Fig.1 Generalized biomass allometric equations with D (diameter at breast height) as the predictor variable.
A: 树干 Stem; B: 树枝 Branch; C: 树叶 Foliage; D: 树根 Root; E: 地上 Aboveground; F: 整株 Total.
853 应 用 生 态 学 报 26卷
表 2 不同树种各器官的生物量相对生长模型
Table 2 Biomass allometric equations for different tree components by tree species
树种
Species
器官
Component
模型Ⅰ
Model Ⅰ
R2 SEE 模型Ⅱ
Model Ⅱ
R2 SEE
混合物种 树干 Stem WS = 0.1413D2.2696 0.92 0.3088 WS = 0.0627(D2H) 0.8861 0.93 0.2991
Mixed species 树枝 Branch WB = 0.0059D2.6825 0.84 0.5604 WB = 0.0033(D2H) 0.9976 0.76 0.6739
树叶 Foliage WL = 0.0034D2.9029 0.85 0.5789 WL = 0.0017(D2H) 1.0851 0.78 0.6940
地上 Aboveground WA = 0.1381D2.3771 0.94 0.2855 WA = 0.0632(D2H) 0.9185 0.92 0.3173
树根 Root WR = 0.0238D2.4851 0.90 0.3874 WR = 0.0123(D2H) 0.9388 0.85 0.4801
整株 Total WT = 0.1646D2.3916 0.94 0.2788 WT = 0.0772(D2H) 0.9196 0.92 0.3286
栲树 树干 Stem WS = 0.1234D2.3429 0.98 0.1986 WS = 0.0356(D2H) 0.9680 0.98 0.1706
Castanopsis fargesii 树枝 Branch WB = 0.0043D2.8734 0.90 0.5577 WB = 0.0011(D2H) 1.1676 0.87 0.6218
树叶 Foliage WL = 0.0020D3.1498 0.90 0.6210 WL = 0.0004(D2H) 1.2829 0.88 0.6789
地上 Aboveground WA = 0.1117D2.4898 0.98 0.1800 WA = 0.0307(D2H) 1.0216 0.98 0.1935
树根 Root WR = 0.0126D2.7281 0.97 0.2644 WR = 0.0032(D2H) 1.1156 0.96 0.3268
整株 Total WT = 0.1230D2.5277 0.99 0.1738 WT = 0.0335(D2H) 1.0388 0.98 0.2018
鹿角锥 树干 Stem WS = 0.1178D2.3426 0.94 0.2815 WS = 0.0418(D2H) 0.9481 0.97 0.1814
Castanopsis lamontii 树枝 Branch WB = 0.0025D3.0512 0.89 0.5184 WB = 0.0009(D2H) 1.1943 0.86 0.5746
树叶 Foliage WL = 0.0015D3.2150 0.91 0.4709 WL = 0.0004(D2H) 1.2861 0.93 0.4356
地上 Aboveground WA = 0.0921D2.5354 0.94 0.2897 WA = 0.0316(D2H) 1.0195 0.97 0.2207
树根 Root WR = 0.0265D2.5053 0.88 0.4454 WR = 0.0090(D2H) 1.0104 0.90 0.3944
整株 Total WT = 0.1221D2.5211 0.94 0.3043 WT = 0.0418(D2H) 1.0149 0.96 0.2332
钩锥 树干 Stem WS = 0.1915D2.0980 0.96 0.2347 WS = 0.1046(D2H) 0.8239 0.97 0.2039
Castanopsis tibetana 树枝 Branch WB = 0.0047D2.7655 0.93 0.4488 WB = 0.0024(D2H) 1.0702 0.91 0.5038
树叶 Foliage WL = 0.0056D2.7888 0.90 0.5502 WL = 0.0029(D2H) 1.0734 0.87 0.6194
地上 Aboveground WA = 0.1753D2.2494 0.97 0.2345 WA = 0.0940(D2H) 0.8799 0.97 0.2308
树根 Root WR = 0.0366D2.3897 0.95 0.3180 WR = 0.0207(D2H) 0.9219 0.93 0.3899
整株 Total WT = 0.2149D2.2747 0.97 0.2251 WT = 0.1169(D2H) 0.8868 0.97 0.2458
石栎 树干 Stem WS = 0.2927D2.0354 0.89 0.3545 WS = 0.1508(D2H) 0.7956 0.86 0.4138
Lithocarpus glaber 树枝 Branch WB = 0.0281D2.1397 0.90 0.3652 WB = 0.0143(D2H) 0.8336 0.85 0.4373
树叶 Foliage WL = 0.0113D2.4880 0.82 0.5861 WL = 0.0048(D2H) 0.9789 0.80 0.6257
地上 Aboveground WA = 0.3525D2.0581 0.92 0.3128 WA = 0.1804(D2H) 0.8043 0.88 0.3805
树根 Root WR = 0.0585D2.2129 0.94 0.2769 WR = 0.0313(D2H) 0.8513 0.87 0.4099
整株 Total WT = 0.4086D2.0880 0.93 0.2956 WT = 0.2116(D2H) 0.8130 0.88 0.3788
猴欢喜 树干 Stem WS = 0.1635D2.2012 0.89 0.2957 WS = 0.1291(D2H) 0.7756 0.94 0.2273
Sloanea sinensis 树枝 Branch WB = 0.0180D2.2849 0.71 0.5693 WB = 0.0255(D2H) 0.7213 0.59 0.6680
树叶 Foliage WL = 0.0052D2.7326 0.76 0.5889 WL = 0.0082(D2H) 0.8585 0.64 0.7283
地上 Aboveground WA = 0.1790D2.2704 0.92 0.2494 WA = 0.1614(D2H) 0.7802 0.92 0.2518
树根 Root WR = 0.0594D2.1316 0.88 0.3090 WR = 0.0613(D2H) 0.7144 0.83 0.3598
整株 Total WT = 0.2358D2.2483 0.94 0.2241 WT = 0.2180(D2H) 0.7691 0.93 0.2412
虎皮楠 树干 Stem WS = 0.1369D2.3000 0.94 0.2471 WS = 0.0951(D2H) 0.8418 0.94 0.2388
Daphniphyllum 树枝 Branch WB = 0.0061D2.6251 0.84 0.4638 WB = 0.0046(D2H) 0.9442 0.82 0.4994
oldhami 树叶 Foliage WL = 0.0018D3.0804 0.84 0.5508 WL = 0.0014(D2H) 1.1001 0.80 0.6101
地上 Aboveground WA = 0.1370D2.3783 0.95 0.2307 WA = 0.0975(D2H) 0.8657 0.94 0.2437
树根 Root WR = 0.0286D2.3983 0.92 0.2852 WR = 0.0209(D2H) 0.8688 0.91 0.3108
整株 Total WT = 0.1726D2.3686 0.96 0.2009 WT = 0.1239(D2H) 0.8611 0.95 0.2205
赤杨叶 树干 Stem WS = 0.1114D2.4465 0.92 0.3012 WS = 0.0345(D2H) 0.9519 0.95 0.2344
Alniphyllum fortunei 树枝 Branch WB = 0.0042D2.5393 0.78 0.5626 WB = 0.0015(D2H) 0.9686 0.78 0.5690
树叶 Foliage WL = 0.0008D3.3098 0.84 0.6107 WL = 0.0002(D2H) 1.2580 0.83 0.6316
地上 Aboveground WA = 0.1121D2.4896 0.92 0.3008 WA = 0.0346(D2H) 0.9666 0.95 0.2411
树根 Root WR = 0.0344D2.3037 0.84 0.4284 WR = 0.0147(D2H) 0.8630 0.80 0.4717
整株 Total WT = 0.1548D2.4354 0.93 0.2882 WT = 0.0514(D2H) 0.9393 0.94 0.2576
乳源木莲 树干 Stem WS = 0.0825D2.4562 0.94 0.2940 WS = 0.0377(D2H) 0.9295 0.95 0.2821
Manglietia 树枝 Branch WB = 0.0190D2.1851 0.93 0.2889 WB = 0.0097(D2H) 0.8238 0.93 0.2940
yuyuanensis 树叶 Foliage WL = 0.0234D2.0938 0.92 0.3027 WL = 0.0125(D2H) 0.7871 0.91 0.3165
地上 Aboveground WA = 0.1231D2.3836 0.95 0.2532 WA = 0.0584(D2H) 0.9003 0.95 0.2503
树根 Root WR = 0.0213D2.4132 0.91 0.3688 WR = 0.0103(D2H) 0.9080 0.90 0.3809
整株 Total WT = 0.1463D2.3845 0.95 0.2595 WT = 0.0696(D2H) 0.9002 0.95 0.2592
少叶黄杞 树干 Stem WS = 0.1958D2.0293 0.86 0.3681 WS = 0.0941(D2H) 0.7998 0.85 0.3747
Engelhardtia fenzlii 树枝 Branch WB = 0.0021D3.1156 0.84 0.6119 WB = 0.0008(D2H) 1.2105 0.81 0.6641
树叶 Foliage WL = 0.0021D3.1522 0.88 0.5197 WL = 0.0007(D2H) 1.2335 0.86 0.5571
地上 Aboveground WA = 0.1406D2.2912 0.87 0.3998 WA = 0.0636(D2H) 0.8982 0.85 0.4211
树根 Root WR = 0.0174D2.5009 0.85 0.4661 WR = 0.0074(D2H) 0.9777 0.83 0.4949
整株 Total WT = 0.1583D2.3232 0.88 0.3895 WT = 0.0709(D2H) 0.9105 0.86 0.4128
9532期 左舒翟等: 亚热带常绿阔叶林 9个常见树种的生物量相对生长模型
株的模型是以 D为自变量的模型为优,其拟合效果
(R2和 SEE)均优于以 D2H为自变量的模型,但树干
的模型则是 D2H为自变量的模型为优(图 1、表 2).
此外,使用同一种模型拟合时,树枝与树叶模型的解
释能力(R2)明显低于其他器官.
考虑树种差异并以 D 为自变量时,猴欢喜的模
型解释能力(R2)最小,平均为(0.85±0.04);栲树模
型的最高,平均为(0.95±0.02) (表 2).将树高引入
单物种模型后,发现:鹿角锥、虎皮楠和赤杨叶的模
型解释能力分别提高了 1.7%、0.3%和 0.3%,其他树
种均呈现不同程度的下降,最高达 5.6%(猴欢喜);
除树干的模型解释能力有所提高外,其余林木器官
均呈现不同程度的降低(图 2).
由图 3可知,进一步比较单物种及混合物种的
生物量相对生长模型,混合物种模型在预测不同树
种及其器官生物量时存在较大的偏差.与以 D 和
D2H为自变量的混合物种模型相比,少叶黄杞模型
的 SEE分别增加了 19.7%和 10.9%,但其余 8 个树
种的 SEE呈现不同程度的下降.此外,林木器官模型
的 SEE均呈现不同程度的降低,其中树枝的变化最
大,分别降低了 13.0%和 20.3%.
图 2 随着树高变量的引入生物量相对生长模型 R2变化百
分比
Fig.2 Percentage variation in R2 (determination of coefficient)
when tree height was added into the D⁃only biomass allometric
equation.
A: 树种 Tree species; B: 器官 Tree component. CF: 栲树 Castanopsis
fargesii; CL: 鹿角锥 Castanopsis lamontii; CT: 钩锥 Castanopsis tibet⁃
ana; LG: 石栎 Lithocarpus glaber; SS: 猴欢喜 Sloanea sinensis; DO:
虎皮楠 Daphniphyllum oldhami; AF:赤杨叶 Alniphyllum fortunei; MY:
乳源木莲 Manglietia yuyuanensis; EF: 少叶黄杞 Engelhardtia fenzlii.
下同 The same below.
图 3 单物种及混合物种生物量相对生长模型 SEE 变化百
分比
Fig.3 Percentage variation in SEE (standard error of the esti⁃
mates) between species⁃specific and generalized biomass allo⁃
metric equation.
Ⅰ: W=aDb; Ⅱ: W=a(D2H) b .
3 讨 论
通常认为,单物种生物量模型的准确度要优于
混合物种模型,尤其是树叶、树枝等物质转换周期短
的器官[25] .但是,本研究发现,并非所有树种的单物
种模型都优于混合物种模型,少叶黄杞和赤杨叶以
D为自变量的单物种模型 SEE 大于混合物种模型,
而石栎和少叶黄杞在引入树高后,其单物种模型的
SEE大于混合物种模型.增加 H 常能提高生物量模
型的精度[26-27] .然而,本研究发现,树枝、树叶、树
根、地上和整株的生物量模型是以 D 为自变量的模
型要优于以 D2H为自变量的模型,但树干则是 D2H
为自变量的模型为优(图 1、表 2),这可能是由于树
高作为树干纵向生长最直接的指标引起的.此结论
与针叶树生物量模型结论相似,引入 H 作为第二变
量,对于松、杉等针叶树种的树干生物量预测效果有
所提升,但对树根生物量影响不大.
理论上,可一次性测定树木胸径 (D)和树高
(H).但是,准确测量 H 存在一定难度,目测法在实
际生产作业中使用较多,但误差较大.现在应用全站
仪、Vertex Ⅳ测高仪等设备可以准确测量树木 H.由
于林区地形复杂,道路崎岖等因素,而且在常绿阔叶
天然林中,树枝错综复杂,难以精准寻找树木最高
点,要获取准确的 H数据需要投入大量的人力物力.
063 应 用 生 态 学 报 26卷
本文发现,将树高引入单物种模型后,模型解释能力
的变化幅度最大仅为 5.6%(图 2).因此,若对估算
精度要求不高或者估算地上和整株生物量的情况
下,可采用以 D为自变量的生物量相对生长模型.
构建生物量相对生长模型时,自变量除了常用
的胸径和树高之外还可选用其他测量指标,如林
龄[28]和材积[29] .考虑到不同生长阶段树木生物量分
配存在显著差异,一些学者建议构建生物量相对生
长模型时,可增加树木年龄来提升模型的估算精
度[28,30] .本文的研究对象是常绿阔叶次生林,涵盖了
林龄 15~40 a的林分,其树木年龄的确定较为困难.
如果将年龄纳入生物量模型中,也将大大降低模型
的实用性,尤其是推算次生林的生物量时.因此,生
物量相对生长模型的选择应根据实际情况、目的和
不同变量的可获取性来决定,模型中引入较多变量
固然可以提供模型预测精度[28],但往往会增加调查
的难度,降低了相对生长模型的实用性,故而需要综
合考虑准确性与实际需求之间的平衡点[6] .
本文采用 2 种不同的生物量相对生长模型,研
究了亚热带常绿阔叶林 9种常见阔叶树的生物量模
型.本研究不仅完善了亚热带阔叶林主要树种的生
物量模型,而且发现若不考虑种间和模型形式间的
差异,将会严重影响生物量碳库及其动态评估的准
确性.因此,为提高生物量估算的准确性,应综合考
虑种间和模型形式间的差异.当然,若将本文建立的
生物量模型应用于其他区域时,可能需进行适当
校正.
参考文献
[1] Sabine CL, Heimann M, Artaxo P, et al. Current status
and past trends of the global carbon cycle / / Field CB,
Raupach MR, MacKenzie SH, eds. The Global Carbon
Cycle: Integrating Humans, Climate and the Natural
World. Washington DC: Island Press, 2004: 17-44
[2] IPCC. 2006 IPCC Guidelines for National Greenhouse
Gas Inventory (vol. 4): Agriculture, Forestry and Other
Land Use [ EB / OL]. ( 2006⁃10⁃01) [ 2014⁃07⁃03].
http: / / www. ipcc⁃nggip. iges. or. jp / public / 2006gl / vol4.
html
[3] West PW. Tree and Forest Measurement. New York:
Springer, 2009
[4] Luo Y⁃J (罗云建), Wang X⁃K (王效科), Zhang X⁃Q
(张小全), et al. Biomass and Its Allocation of Forest
Ecosystems in China. Beijing: China Forestry Publishing
House, 2013 (in Chinese)
[5] Somogyi Z, Cienciala E, Mäkipää R, et al. Indirect
methods of large⁃scale forest biomass estimation. Euro⁃
pean Journal of Forest Research, 2007, 126: 197-207
[6] Zhao M (赵 敏), Zhou G⁃S (周广胜). Forest inven⁃
tory data (FID)⁃based biomass models and their pros⁃
pects. Chinese Journal of Applied Ecology (应用生态学
报), 2004, 15(8): 1468-1472 (in Chinese)
[7] Salis SM, Assis MA, Mattos PP, et al. Estimating the
aboveground biomass and wood volume of savanna wood⁃
lands in Brazil’s Pantanal wetlands based on allometric
correlations. Forest Ecology and Management, 2006,
228: 61-68
[8] Wang CK. Biomass allometric equations for 10 co⁃occur⁃
ring tree species in Chinese temperate forests. Forest
Ecology and Management, 2006, 222: 9-16
[9] Zianis D, Muukkonen P, Mäkipää R, et al. Biomass
and stem volume equations for tree species in Europe.
Silva Fennica Monographs, 2005, 4: 63
[10] Chen D⁃X (陈德祥), Li Y⁃D (李意德), Liu HP, et
al. Biomass and carbon dynamics of a tropical mountain
rain forest in China. Science China Life Science (中国科
学: 生命科学), 2010, 40(7): 596 - 609 ( in Chi⁃
nese)
[11] Zhang L (张 林), Luo T⁃X (罗天祥), Deng K⁃M
(邓坤枚), et al. Biomass and net primary productivity
of secondary evergreen broadleaved forest in Huangmian
Forest Farm. Chinese Journal of Applied Ecology (应用
生态学报), 2004, 15(11): 2029-2033 (in Chinese)
[12] Zhang G⁃B (张国斌), Li X⁃Q (李秀芹), She X⁃S
(佘新松), et al. Biomass characteristics of dominant
tree species ( group) at Lingnan Forest Farm in Anhui
Province. Scientia Silvae Sinicae (林业科学), 2012,
48(5): 136-140 (in Chinese)
[13] Peichl M, Arain MA. Allometry and partitioning of
above⁃ and belowground tree biomass in an age⁃sequence
of white pine forests. Forest Ecology and Management,
2007, 253: 68-80
[14] Landsberg J, Sands P. Physiological Ecology of Forest
Production: Principles, Processes and Models. London:
Academic Press, 2011
[15] Ngomanda A, Obiang NLE, Lebamba J, et al. Site⁃
specific versus pantropical allometric equations: Which
option to estimate the biomass of a moist central African
forest? Forest Ecology and Management, 2014, 312: 1-
9
[16] Yang T⁃H (杨同辉), Song K (宋 坤), Da L⁃J (达
良俊), et al. The biomass and aboveground net primary
productivity of Schima superba - Castanopsis carlesii in
forests in east China. Science China Life Science (中国
科学: 生命科学), 2010, 40(7): 610-619 ( in Chi⁃
nese)
[17] Feng Z⁃W (冯宗炜), Wang X⁃K (王效科), Wu G
(吴 刚). Biomass and Productivity of Forest Ecosys⁃
tems in China. Beijing: Science Press, 1999 ( in Chi⁃
nese)
[18] Fang JY, Chen AP, Peng CH, et al. Changes in forest
biomass carbon storage in China between 1949 and
1998. Science, 2001, 292: 2320-2322
[19] Li N (李 宁), Xu W⁃B (徐武兵), Lai J⁃S (赖江
山), et al. The coarse root biomass of eight common tree
1632期 左舒翟等: 亚热带常绿阔叶林 9个常见树种的生物量相对生长模型
species in subtropical evergreen forest. Chinese Science
Bulletin (科学通报), 2013, 58 ( 4): 329 - 335 ( in
Chinese)
[20] State Forestry Administration (国家林业局). Methods
for Establishing Forest Biomass Models in National For⁃
est Inventories ( Trial ) [ EB / OL ]. ( 2008⁃12⁃20 )
[ 2014⁃07⁃03 ]. http: / / www. forestry. gov. cn / portal /
main / govfile / 13 / govfile_1377.html (in Chinese)
[21] Niklas KJ. Plant Allometry: The Scaling of Form and
Process. Chicago: University of Chicago Press, 1994
[22] Warton DI, Wright IJ, Falster DS, et al. Bivariate line⁃
fitting methods for allometry. Biological Reviews, 2006,
81: 259-291
[23] Ter⁃Mikaelian MT, Korzukhin MD. Biomass equations
for sixty⁃five North American tree species. Forest Ecolo⁃
gy and Management, 1997, 97: 1-24
[24] Parresol BR. Assessing tree and stand biomass: A re⁃
view with examples and critical comparisons. Forest Sci⁃
ence, 1999, 45: 573-593
[25] Buech RR, Rugg DJ. Biomass relations of shrub compo⁃
nents and their generality. Forest Ecology and Manage⁃
ment, 1989, 26: 257-264
[26] Dong L⁃H (董利虎), Li F⁃R (李凤日), Jia W⁃W (贾
炜玮), et al. Compatible biomass models for main tree
species with measurement error in Heilongjiang Province
of Northeast China. Chinese Journal of Applied Ecology
(应用生态学报), 2011, 22(10): 2653 - 2661 ( in
Chinese)
[27] Ketterings QM, Coe R, van Noordwijk M, et al. Reduc⁃
ing uncertainty in the use of allometric biomass equations
for predicting above⁃ground tree biomass in mixed
secondary forests. Forest Ecology and Management,
2001, 146: 199-209
[28] Li Y (李 燕), Zhang J⁃G (张建国), Duan A⁃G (段
爱国), et al. Selection of biomass estimation models for
Chinese fir plantation. Chinese Journal of Applied Ecolo⁃
gy (应用生态学报), 2010, 21(12): 3036-3046 ( in
Chinese)
[29] Xing Y⁃Q (邢艳秋), Wang L⁃H (王立海). Compati⁃
ble biomass estimation models of natural forests in
Changbai Mountains based on forest inventory. Chinese
Journal of Applied Ecology (应用生态学报), 2007, 18
(1): 1-8 (in Chinese)
[30] Wang J⁃S (汪金松), Zhang C⁃Y (张春雨), Fan X⁃H
(范秀华), et al. Biomass allocation patterns and allo⁃
metric models of Abies nephrolepis Maxim. Acta Ecologica
Sinica (生态学报), 2011, 31(14): 3918-3927 ( in
Chinese)
作者简介 左舒翟,女,1984年生,助理研究员. 主要从事森
林生态系统碳循环研究. E⁃mail: sdzuo@ iue.ac.cn
责任编辑 孙 菊
263 应 用 生 态 学 报 26卷