王怡 湯景明 孫擁康
摘 要:為研究青岡櫟人工林樹(shù)高和胸徑的生長(zhǎng)關(guān)系模型,以九峰國(guó)家森林公園的青岡櫟人工林為研究對(duì)象,將484組樹(shù)高和胸徑數(shù)據(jù)分成建模數(shù)據(jù)和檢驗(yàn)數(shù)據(jù),選用10種常見(jiàn)樹(shù)高曲線模型分別對(duì)建模數(shù)據(jù)進(jìn)行擬合。結(jié)果顯示,10種樹(shù)高曲線模型均達(dá)到極顯著水平(P<0.01),其中二次曲線模型的殘差平方和(SSE)、均方根誤差(RMSE)最小,決定系數(shù)(R2)最大,符合最優(yōu)模型的篩選原則,最優(yōu)模型的殘差分析無(wú)異常值出現(xiàn),并且將檢驗(yàn)數(shù)據(jù)代入最優(yōu)模型得到的樹(shù)高預(yù)測(cè)值和實(shí)測(cè)值之間也無(wú)顯著差異(Sig.= 0.159>0.05)。結(jié)論表明,二次曲線模型可以對(duì)青岡櫟人工林樹(shù)高進(jìn)行有效預(yù)測(cè)。
關(guān)鍵詞:青岡櫟;胸徑;樹(shù)高曲線;模型研究
中圖分類(lèi)號(hào):S758 ? ?文獻(xiàn)標(biāo)識(shí)碼:A ? 文章編號(hào):1006-8023(2021)01-0001-05
Abstract:In order to study the growth relation model between tree height and DBH of Cyclobalanopsis glauca plantation, Cyclobalanopsis glauca plantation in Jiufeng National Forest Park was taken as the research object. 484 groups of tree height and DBH data were divided into modeling data and test data, and 10 common tree height curve models were selected to fit the modeling data respectively. The results showed that all the 10 tree height curve models reached extremely significant level (P<0.01), among which the quadratic curve model had the smallest SSE and RMSE and the largest R2, which accorded with the screening principle of the optimal model. The residual analysis of the optimal model showed no significant difference, and there was no significant difference between the predicted and measured tree height values obtained by substituting test data into the optimal model (Sig.= 0.159>0.05). The conclusion showed that the quadratic curve model could effectively predicted the height of Cyclobalanopsis glauca plantation.
Keywords:Cyclobalanopsis glauca; DBH; tree height curve; model research
0 引言
樹(shù)高和胸徑是森林調(diào)查和經(jīng)營(yíng)過(guò)程中最基本的測(cè)量因子,是用來(lái)計(jì)算材積、立地指數(shù)以及預(yù)估林木生長(zhǎng)和收獲的重要數(shù)據(jù)[1]。在實(shí)際調(diào)查中,胸徑屬于易測(cè)因子,而樹(shù)高的獲取則相對(duì)困難,且精度不高。通過(guò)建立樹(shù)高曲線模型,對(duì)胸徑和樹(shù)高的相關(guān)關(guān)系進(jìn)行描述,利用胸徑準(zhǔn)確預(yù)測(cè)樹(shù)高,對(duì)于生產(chǎn)實(shí)踐和科學(xué)研究都具有非常重要的意義[2-3]。
青岡櫟(Cyclobalanopsis glauca)屬殼斗科青岡屬,是良好的珍貴用材樹(shù)種,其適生范圍廣泛,是我國(guó)亞熱帶常綠闊葉林帶的主要優(yōu)勢(shì)樹(shù)種和造林樹(shù)種[4]。當(dāng)前已有許多關(guān)于各樹(shù)種樹(shù)高和胸徑相關(guān)模型的研究[5-7],但未見(jiàn)有關(guān)青岡櫟樹(shù)高和胸徑模型的研究報(bào)道。本研究以青岡櫟人工林為研究對(duì)象,以樹(shù)高和胸徑實(shí)地調(diào)查數(shù)據(jù)為基礎(chǔ),建立青岡櫟樹(shù)高-胸徑模型,以期能為該地區(qū)青岡櫟的經(jīng)營(yíng)和研究提供科學(xué)依據(jù)。
1 研究區(qū)概況
研究區(qū)位于武漢市九峰國(guó)家森林公園,地處武漢市東郊,距武漢市中心12 km ,地理中心位置為114°29′50" E, 30°31′4" N,占地面積約330 hm2。森林公園境內(nèi)山巒蜿蜒,山間盆地地勢(shì)平坦,屬湖北省內(nèi)典型的丘陵地勢(shì)地貌景觀。森林公園地處亞熱帶季風(fēng)性濕潤(rùn)氣候區(qū),四季分明、光照充足、熱量豐富、雨量充沛、水熱同季、干濕明顯,全年平均氣溫16.3 ℃ ,極端最高氣溫為41 ℃,極端最低氣溫-17.6 ℃ ,年日照時(shí)數(shù)1 600 h左右,無(wú)霜期達(dá)240 d,年降雨量1 200~1 400 mm ,年平均相對(duì)濕度79% 。研究區(qū)大多數(shù)為人工林,主要樹(shù)種有馬尾松(Pinus massoniana)、杉木(Cunninghamia lanceolata)、濕地松(Pinus elliottii)、青岡櫟(Cyclobalanopsis glauca)、楓香(Liquidambar formosana)和栓皮櫟(Quercus variabilis)等。
2.2 模型的選取與擬合
根據(jù)建模數(shù)據(jù),繪制以胸徑為橫坐標(biāo),以樹(shù)高為縱坐標(biāo)的散點(diǎn)圖(圖1),通過(guò)對(duì)散點(diǎn)分布趨勢(shì)的觀察發(fā)現(xiàn),胸徑和樹(shù)高具有正相關(guān)性,根據(jù)前人研究經(jīng)驗(yàn)[8-12],本文選用10種常見(jiàn)的相關(guān)模型對(duì)數(shù)據(jù)進(jìn)行擬合,公式如下。
3 結(jié)果與分析
3.1 曲線模型優(yōu)選
利用青岡櫟建模數(shù)據(jù)進(jìn)行樹(shù)高曲線模型擬合,模型擬合結(jié)果見(jiàn)表2,結(jié)果顯示10種相關(guān)模型均為P<0.01,表明所有方程都達(dá)到了極顯著的水平,但是最優(yōu)模型的篩選原則是殘差平方和和均根方誤差較小,而決定系數(shù)較大。其中,二次曲線模型的SSE=502.654 82,RMSE=1.223 109 4,R2=0.722 168 2,符合最優(yōu)模型的篩選原則,因此可以確定二次曲線模型為最優(yōu)樹(shù)高曲線模型。
3.2 最優(yōu)曲線模型殘差分析
通過(guò)二次曲線模型的回歸方程可計(jì)算出樹(shù)高預(yù)測(cè)值,殘差(ei)即樹(shù)高預(yù)測(cè)值與實(shí)測(cè)值之差,將其與胸徑作圖(圖2),可以發(fā)現(xiàn)殘差值隨著胸徑的增加呈現(xiàn)不規(guī)律分布的狀態(tài),且主要在ei=0附近波動(dòng),預(yù)測(cè)值和實(shí)際測(cè)量值非常接近,沒(méi)有異常值的出現(xiàn),由此可以認(rèn)為建模樣本的數(shù)據(jù)基本正常,模型的擬合效果較好[15]。
3.3 最優(yōu)曲線模型檢驗(yàn)
將檢驗(yàn)樣本數(shù)據(jù)中的胸徑值(D)代入最優(yōu)樹(shù)高曲線模型,即代入二次曲線模型H =2.068 0+0.848 7×D-0.016 5×D2之中,求出樹(shù)高預(yù)測(cè)值,利用SPSS軟件對(duì)樹(shù)高實(shí)測(cè)值和預(yù)測(cè)值進(jìn)行成對(duì)樣本T檢驗(yàn),檢驗(yàn)數(shù)據(jù)樣本統(tǒng)計(jì)及檢驗(yàn)結(jié)果見(jiàn)表3,結(jié)果顯示樹(shù)高實(shí)測(cè)值和預(yù)測(cè)值之間無(wú)顯著差異(Sig.=0.159>0.05),說(shuō)明二次曲線模型可以較好的預(yù)測(cè)青岡櫟的樹(shù)高。
4 結(jié)論
(1)本文選用10種常見(jiàn)的樹(shù)高-胸徑方程分別擬合了青岡櫟的樹(shù)高和胸徑關(guān)系模型,以SSE、RMSE和R2等擬合統(tǒng)計(jì)量作為比較和評(píng)價(jià)備選模型的標(biāo)準(zhǔn),最終從中篩選出二次曲線模型為最優(yōu)樹(shù)高曲線模型,模型回歸方程如下:
H = 2.068 0 + 0.848 7×D-0.016 5×D2。
(2)最優(yōu)樹(shù)高曲線模型的所有殘差在ei=0附近隨機(jī)波動(dòng),且分布在一條幅度變化不大的區(qū)帶內(nèi),滿足回歸方程的基本假設(shè)。利用檢驗(yàn)樣本數(shù)據(jù)對(duì)最優(yōu)模型進(jìn)行檢驗(yàn)的結(jié)果表明,二次曲線模型在所有備選模型中,可以較好地模擬青岡櫟樹(shù)高和胸徑的關(guān)系。
(3)本研究只以胸徑因子為條件對(duì)樹(shù)高曲線模型進(jìn)行了研究,而對(duì)林木生長(zhǎng)的影響因素還有很多,如環(huán)境因子、立地條件、林齡和競(jìng)爭(zhēng)等[16-20],另外,由于數(shù)據(jù)有限且存在不可避免的誤差,因此,對(duì)于青岡櫟樹(shù)高曲線模型的適用性和模擬精度還有待于進(jìn)一步的研究。
【參 考 文 獻(xiàn)】
[1]孟憲宇.測(cè)樹(shù)學(xué)[M].2版.北京:中國(guó)林業(yè)出版社,1996.
MENG X Y. Forest mensuration[M]. 2nd ed. Beijing: China Forestry Publishing House, 1996.
[2]王明亮,唐守正.標(biāo)準(zhǔn)樹(shù)高曲線的研制[J].林業(yè)科學(xué)研究,1997,17(3):36-41.
WANG M L, TANG S Z. Research on universal height diameter curves[J]. Rorest Research, 1997, 17(3): 36-41.
[3]李??ɡ?基于分級(jí)的全國(guó)主要樹(shù)種樹(shù)高-胸徑曲線模型[J].林業(yè)科學(xué),2011,47(10):83-90.
LI H K, FA L. Height-diameter model for major tree species in China using the classified height method[J]. Scientia Silvae Sinicae, 2011, 47(10): 83-90.
[4]NI J, SONG Y C. Relationships between geographical distribution of Cyclobalanopsis glauca and climate in China[J]. Acta Botanica Sinica, 1997, 39(5): 451-460.
[5]顧鳳岐,王艷.基于紅松胸徑與樹(shù)高相關(guān)模型的研究[J].森林工程,2013,29(4):39-41.
GU F Q, WANG Y. The research of the regression model between tree height and diameter at breast height of Korean pine[J]. Forest Engineering, 2013, 29(4): 39-41.
[6]趙俊卉,劉燕,張慧東,等.長(zhǎng)白山天然林不同樹(shù)種樹(shù)高曲線對(duì)比研究[J].浙江林學(xué)院學(xué)報(bào),2009,26(6):865-869.
ZHAO J H, LIU Y, ZHANG H D, et al. Comparison of height-diameter models for main species in natural forest of Changbai Mountains[J]. Journal of Zhejiang Forestry College, 2009, 26(6): 865-869.
[7]李明華,肖舜禎.上海市主要造林樹(shù)種的胸徑-樹(shù)高模型研究[J].江西農(nóng)業(yè)大學(xué)學(xué)報(bào),2019,41(3):501-511.
LI M H, XIAO S Z. Diameter-height models of main tree species in Shanghai[J]. Acta Agriculturae Universitatis Jiangxiensis, 2019, 41(3): 501-511.
[8]童潔,石玉立.加格達(dá)奇3種森林類(lèi)型樹(shù)高-胸徑的曲線擬合[J].東北林業(yè)大學(xué)學(xué)報(bào),2017,45(2):6-11.
TONG J, SHI Y L. Tree height-diameter model in Jiagedaqi[J]. Journal of Northeast Forestry University, 2017, 45(2): 6-11.
[9]張敏,顧鳳歧,董希斌.帽兒山林區(qū)主要樹(shù)種樹(shù)高與胸徑之間的關(guān)系分析[J].森林工程,2014,30(6):1-4.
ZHANG M, GU F Q, DONG X B. Analysis of height-diameter relationship of main tree species in Maoershan forest region[J]. Forest Engineering, 2014, 30(6): 1-4.
[10]馬小欣,姜鵬,馬嬌嬌,等.沿壩地區(qū)華北落葉松胸徑-樹(shù)高生長(zhǎng)模型的研究[J].林業(yè)資源管理,2015(1):44-48.
MA X X, JIANG P, MA J J, et al. Study on DBH-tree height growth model of Larix principis-rupprechtii along the dam ecotone[J]. Forest Resources Management, 2015(1): 44-48.
[11]王曉林,郭斌.柞樹(shù)樹(shù)高與胸徑相關(guān)關(guān)系的研究[J].森林工程,2012,28(6):18-21.
WANG X L, GUO B. Study on the correlation between height and diameter at breast height for Quercus mongolica[J]. Forest Engineering, 2012, 28(6): 18-21.
[12]蘭潔,張毓?jié)?,李吉玫,?新疆天山不同區(qū)域云杉樹(shù)高、胸徑、年齡相關(guān)關(guān)系研究[J].西部林業(yè)科學(xué),2019,48(2):135-140.
LAN J, ZHANG W T, LI J M, et al. Correlation analysis on tree height, DBH and age of Picea spp. in different areas of Tianshan Mountain[J]. Journal of West China Forestry Science, 2019, 48(2):135-140.
[13]朱玉杰,馮國(guó)紅,董春芳.大興安嶺落葉松不同采伐強(qiáng)度的最優(yōu)樹(shù)高曲線模型[J].東北林業(yè)大學(xué)學(xué)報(bào),2015,43(8):54-58.
ZHU Y J, FENG G H, DONG C F. Optimal tree height curve model of larch in Daxingan mountains under different logging intensity[J]. Journal of Northeast Forestry University, 2015, 43(8): 54-58.
[14]趙俊卉,亢新剛,劉燕.長(zhǎng)白山主要針葉樹(shù)種最優(yōu)樹(shù)高曲線研究[J].北京林業(yè)大學(xué)學(xué)報(bào),2009,31(4):13-18.
ZHAO J H, KANG X G, LIU Y. Optimal height-diameter models for dominant coniferous species in Changbai Mountain, northeastern China[J]. Journal of Beijing Forestry University, 2009, 31(4): 13-18.
[15]何曉群,劉文卿.應(yīng)用回歸分析[M].2版.北京:中國(guó)人民大學(xué)出版社,2007.
HE X Q, LIU W Q. Applied regression analysis[M]. 2nd ed. Beijing: China Renmin University Press, 2007.
[16]熊斌梅,汪正祥,李中強(qiáng),等.七姊妹山自然保護(hù)區(qū)黃杉年齡胸徑樹(shù)高的相關(guān)性研究[J].林業(yè)資源管理,2016,38(4):41-46.
XIONG B M, WANG Z X, LI Z Q, et al. Study on the correlation among age, DBH and tree height of the Pseudotsuga sinensis in Qizimei Mountain nature reserve[J]. Forest Resources Management, 2016,38(4): 41-46.
[17]胡志偉,沈澤昊,呂楠,等.地形對(duì)森林群落年齡及其空間格局的影響[J].植物生態(tài)學(xué)報(bào),2007,31(5):814-824.
HU Z W, SHEN Z H, LV N, et al. Impacts of topography on the spatial pattern of the age of forest community[J]. Journal of Plant Ecology, 2007, 31(5): 814-824.
[18]雷澤勇,周晏平,趙國(guó)軍,等.競(jìng)爭(zhēng)對(duì)遼寧西北部樟子松人工固沙林樹(shù)高生長(zhǎng)的影響[J].干旱區(qū)研究,2018,35(1):144-149.
LEI Z Y, ZHOU Y P, ZHAO G J, et al. Effects of competition on tree height growth of sand-fixation plantations of Pinus sylvestris var. mongolica in northwest Liaoning Province[J]. Arid Zone Research, 2018, 35(1): 144-149.
[19]楊晏平,周娟,趙江萍,等.滇西地區(qū)濕地松人工林生長(zhǎng)過(guò)程分析[J].西部林業(yè)科學(xué),2020,49(1):128-132.
YANG Y P, ZHOU J, ZHAO J P, et al. Growth process analysis of Pinus elliottii in western Yunnan[J]. Journal of West China Forestry Science, 2020, 49(1):128-132.
[20]付晗,魏亞偉,殷有,等.基于小班和樹(shù)高生長(zhǎng)方程比較不同立地條件對(duì)樹(shù)木生長(zhǎng)的差異[J].林業(yè)資源管理,2017,46(2):46-52.
FU H, WEI Y W, YIN Y, et al. Analysis on the differences of tree growth with different site conditions based on sub-compartments and tree height growth model[J]. Forest Resources Management, 2017, 46(2): 46-52.