Fine roots refilling process in an artificial gap in a Picea mongolica forest Fine roots refilling process in an artificial gap in a Picea mongolica forest

Fine roots refilling process in an artificial gap in a Picea mongolica forest

  • 期刊名字:中国林学(英文版)
  • 文件大小:572kb
  • 论文作者:Zou Chun-jing,Ma Yong-liang,Zh
  • 作者单位:Shanghai Key Laboratory of Urbanization and Ecological Restoration,Institute of Applied Ecology
  • 更新时间:2020-11-11
  • 下载次数:
论文简介

For. Stud. China, 2007, 9(1): 19- -26 .RESEARCH ARTICLEDOI 10. 1007/s 11632-007-0004-4Fine roots refilling process in an artificial gap in aPicea mongolica forestZou Chun-jing”Ma Yong-liang'Zhang Chao'Xu Wen-duo1 Shanghai Key Laboratory of Urbanization and Ecological Restoration, East China Normal University, Shanghai 200062, P. R. China2 Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, P R. China .Abstract Picea mongolica is an endemic but endangered species in China. The spruce forest is only found in sandy forest-steppeecotones. In this study, we examined the initial response of the quantity and refilling process of fine roots in an artificial canopy gapwith a diameter of 36 m in a P. mongolica forest. Under the canopy, the fine root length densities of trees, shrubs and herbs were2,622, 864 and 3,086 m:m 2, respectively. The fine root biomass of trees, shrubs and herbs were 148, 62 and 65 gm 2, respectively.In the gap, the fine root length density of trees was 15 1 m:m . The mean fine root densities of shrubs and herbs in the gap were 756and 2,568 m'm 2. The fine root biomass of trees, shrubs and herbs were 9, 52 and 47 gm 2, respectively. Two growing seasons afterthe gap creation, hardly any fine tree roots were found in the middle of the gap. The living tree roots in the gap edge zone weremainly located within a 4.5 m distance from the standing trees. Indices developed to show the influence of trees on fine root lengthdensity clearly revealed the effect of the vicinity of living trees on fine root length density. The root densities of shrubs and herbs didnot show a clear response to gap creation despite the increase of their foliage. Our results suggest that in P. mongolica forests a gapdisturbance creates a distinct tree root gap and that the gap edge trees do not extend their root systems rapidly into the formed rootKey words gap disturbance, Picea mongolica, root system, sandy forest steppe ecotone1 IntroductionHowever, minor disturbances due to insect damagend timber harvesting create smaller gaps and arePicea mongolica is an endemic but endangered spenore common in the sandy forests in the north ofcies in Inner Mongolia Autonomous Region of ChinaChina (Zou et al., 1998). The mean size of the gaps is(Xu et al., 1994). The community is one of several0.1 hm- in these P. mongolica forests. Guidelines forspecial forest ecosystems on the continent because it isa sandy forest type (Xu et al, 1993; Xu and Zheng,1993). The ecosystem is located at the ecotone be-°,200kmtween forest and steppe zones (Fig. 1). As far as culti-Russiavation is concerned, the forest lies in the ecotone be-tween agricultural and pastoral areas. The area is alsoin the transitional zone from the Great Xing anMountains to the Hunshandake Sandy Lands (Xu andMongoliaHarbinZou, 1998). It has played an important role in sandand desertification control as well as in the establish-ment of the“Three Northern Protection Forest Sys-tems”. However, these sandy natural forests have beenseriously destroyed and are at the brink of extinction上40°, Beijingdue to mismanagement, forest fires, plant diseases andinsect pests. Therefore, it is urgent that we try to con-Chinaserve and extend these sandy natural forests.359EYellow SeaDisturbances that create gaps of various sizes are aYellow Rivercommonecological feature of boreal coniferous for-中国煤化工130° .ests (Bergeron et al., 1998; Taskinen et al., 2003). Innatural forests, stand-replacing fires create large gaps.Fig. 1 LocationMHCNMHGAuthor for correspondence. E-mail: cjzou@bio.ecnu.edu.cn.2(Forestry Studies in China, Vol.9, No.1, 2007preservation of valuable habitats increase the length oftion of a canopy gap in a P. mongolica forest in thethe edge considerably (Xu and Zou, 1998). A gap dis-sandy forest-steppe ecotone. Specific questions were: .turbance opens space and releases resources that initi-1) What were the root length densities of fine rootsate vegetation succession and facilitate tree regenera-under the canopy? 2) How did the fine roots respondtion (Phillips and Shure, 1990; Lertzman, 1992; Baz-to the gap creation? 3) How far from the gap edge didzaz and Wayne, 1994; Kuuluvainen, 1994; Li et al.,the fine roots of trees extend?1997; McGuire et al, 2001; DeChantal et al, 2003).Many investigations on various types of forests havefocused on the effect of gaps on plant species compo-2 Study area and methodssition (Collins and Pickett, 1988) on availability ofresources, particularly light (Messier and Puttonen, .2.1 Study site1995; Brockway and Outcalt, 1998) and on microcli-mate (Chen et al, 1993, 1995; Tan et al., 2000). Espe-The study was conducted in Baiyinaobao Nature Re-cially in the vicinity of the gap edge, the surroundingserve (43930'- 43936' N, 117906'-117°16' E at an ele-trees have a strong effect, the so-called edge effect, onvation of 1,420 m) in the east of the Inner Mongoliaenvironmental conditions and processes in the gapAutonomous Region of China. The area is covered byarea. In smaller gaps most of the gap area is likelysand with a thickness of 10- -100 m. The climate is ainfluenced in some way or another by the surroundingtypical temperate continental steppe climate. Theforest. In general, the effects of gap disturbance onmean annual temperature is -1.49C. The annual ac-above-ground processes are better understood thancumulated temperature of>5°C is 1,942°C (Xu, 1985).those in underground processes (Liu et al, 2002).The mean annual precipitation is 448.9 mm and theAccording to our study, the root system of P. mon-potential evaporation 1,526 mm. The length of thegolica is characterized by shallow lateral roots withfrost-free season is 65 days.sinkers (Xu et al, 1998; Xu and Zou, 1998). In addi-The community is dominated by Picea mongolication, it is heterorhizic, having structural long roots andwith scattered individuals of Betula platyphylla Suk.short, normally ectomycorrhizal roots. Most of theand Populus davidiana Dode. The mean age of theroot biomass is located in the upper 30 cm of the min-stand was 110 years and it originated from naturaleral soil and in the humus layer. The long roots de-regeneration. The mean height of trees, weighted byvelop from rapidly growing pioneer roots and frombasal area, was 17 m. The stand had 950 stems perthe first and second order branches that are formedhm^ and stem volume was 450 m' hm~ . The experi-more or less at constant intervals. The growth of themental stand is located on a dune formation of finepioneer roots is not much affected by soil conditions,sand (particle size 0.2- 0.02 mm). The zonal soils ofbut their secondary thickening is related to soil condi-the area are mainly black and brown soils on dunes,tions through development of branches and fine roots.which are podzolic. There also are meadow soils,As a result the final shape of the root system is ir-marsh soils and salt soils on riverbanks and meadows.regular (Zou et al., 2001).The groundwater table is at a depth of 5- 8 m. TheVarious studies in coniferous forests have shownmean thickness of the humus layer is ca.50 mm. Thethat fine root biomass and production change withshrub layer consists mainly of Ostryopsis davidianastand development and succession (Vogt et al, 1995).Decne. Cotoneaster melanocarpus Lodd. DasiphoraThe root systems of long lived trees are known to at-fruticosa (L.) Rydb. and Rosa davurica Pall. The herbtain considerable size in order to absorb water andlayer consists largely of Aneurolepidium chinensenutrients (Stone and Kalisz, 1991). Unlike crowns,(Trin.) Kitag, Agropyron cristatum Gaertner.,which are normally segregated in the canopy space,Calamagrostis epigeios (L.) Roth. and Artemisia fri-their root systems often intermingle with the root sys-gida Wild. The ground layer is dominated bytems of nearby trees having intraspecies root connec-Phytidium rugosum (Hedw.) Kindbs. andtion (Caldwell, 1987). The death of a single tree or aPhytidiadphs trguetrus (Hedw.) Warnst.cluster may not create a functional root gap(Wilczynski and Pickett, 1993; Parsons et al., 1994;Campbell et al., 1998; Jones et al, 2003). Fine root2.2 Experimental areadistribution and dynamics in gaps are keys for under-standing the below-ground competitive interference ofIn the middle of a mature Picea mongolica stand, ansurrounding trees.relation to gaps have received less attention, especiallythe trees were mapped in the experimental area and inin boreal forests (Roberts, 1976; Taskinen et al,its close vicinity, using a tachymeter. In April 1999,2003). .when the soil was frozen, a circular gap with a diame-The objective in our research was to examine theterof36mw:中国煤化iexperimentalinitial response of the amount and distribution of finearea (Fig. 2). Bthe cut treesroots (i) of trees, (i) shrubs and (ili) herbs to the crea-were removed:MHCNMHG.Zou Chun-jing et al: Fine roots rfillig process in an artificial gap in a Picea mongolica forest212.3 Root sampling0rThe soil sampling was based on a 10 mx 10 m grid laidacross the 70 mx70 m experimental area. Soil sampleswere taken with a soil auger (diameter 50 mm) to a0depth of 0.5 m. The root sampling was carried outover a period of three consecutive years: before the和0gap creation in August 1998, after the gap creation inAugust 2000 and in August 2001 (Fig. 2). Sampling宫30A口。was not done in 1999 following the harvesting, be-。0/9cause it was thought likely that the root system oftrees stay alive for some time after they have been cutdown. Another reason was that the separation of deadand live roots is always problematic, especially whenthe dying process is underway, since visual appear-ance may be misleading. The period of two years was1020 3(40 50 6C70considered to be long enough to kill most of the rootsystems of the removed trees, so that the effect of theFig. 2 Map of the tree crowns in the experimental area show-experimental gap on root distribution could be de-ing the location of the gap edge. The lighter dots denote thetected. When new samples were taken using the 10trees that were cut to create the artificial gap, darker dots showmx10 m grid, they were taken 0.5 m apart from pre-the remaining living trees. The size of the dot is directly re-vious samples. Because the separation of the rootslated to the width of the tree crown. The location of the rootfrom the soil was extremely time-consuming, it wassamples is marked as△(samples in 1998), + (samples inultimately possible to analyze only some of the sam-2000), 0 (samples in 2001). The solid line marks the gap.ples. Root length was measured from 30 samplestaken in 1998, from 70 samples taken in 2000 andfrom 53 samples taken in 2001. Given its relatively2.4 Response of understory vegetation in asmall sample size, the samples in 1998 do not evenlyPicea mongolica forestcover the experimental area. Nevertheless, this samplewas considered to give a reasonable estimate of theThe response of the understory vegetation to the gappre-treatment situation, given that the forest wascreation was quantified using a plant canopy analyzer,structurally rather homogeneous. Soil samples wereLAI 2000. The measurements of the gap fraction, thatstored in a freezer until analyzed. The soil samplesis, the proportion of sky not shaded by plants, werewere divided into three layers: 1) a humus layer, 2) thetaken late in the growing season in the beginning oftop 10 cm layer of mineral soil, and 3) the 10- -20 cmJuly of both years 2000 and 2001 at 64 systematicallylayer of mineral soil. In the laboratory the roots werelocated points in the study area 50 mm above the soilseparated from the soil using wet sieving. The livesurface and using a view angle of 180°. Readingsroots were manually classified into three vegetationwere first taken above the ground vegetation, thengroups: 1) tree roots, 2) shrub roots and 3) herb roots.below and again above. The“relative transmittance'The separation was based on color and morphologicalwas calculated as a ratio of below to the average ofdifferences. The separation was carried out under aabove readings.microscope. Tough and flexible roots were classifiedinto living roots and dead roots lacking cohesion be-tween the cortex and periderm. Pieces shorter than 42.5 Predicting the effect of gap creation onmm were not separated. Roots were placed one by onefine rootson a transparent film and covered with a thin plasticfilm in order to prevent drying. The length of the rootsThe reduction of the root length density in the edgeand the diameter distribution were measured. In orderzone between the gap and the forest can be expectedto analyze the response of roots to gap creation, theto be related to the proportion of the trees that weresoil samples were divided into two groups, namely,not logged around a given sample location. Four indi-samples taken under the tree canopy and samplesces were used to describe the joint effect of trees ontaken in the gap. This separation was based oIfine root length density in the sample location. Thewhether their location was closer to a living than to asimplest index was the proportion of living treesremoved tree. A total of 103 samples were taken underamong the nearest 25 trees (Io). In addition, tree dis-the canopy, that is, the sample location was closer to atance and tree size dependent indices (I- I3) wereliving tree than to a removed tree. Fifty samples weretested, since中国煤化工wn to affecttaken in the canopy gap.both the total rCYH: extent of theroot system (ICNMHG).Allthein-.2Forestry Studies in China, Vol.9, No.1, 2007dices have 1 added to the denominator in order toThe distance to the nearest living tree (Eq. (1)) ex-force the functions to have finite values at zero dis-plained 35% of the variation of the fine tree roottance. The computed tree functions,affecting roots,length density in the gap.The derived indices affecting roots improved theestimates of root length density in the gap and in theIn _"living(1)edge. The proportion of the living trees (Io, Eq. (1))nallexplained 47% of the variation in fine root lengthdensity (Table 2). Index I (Eq. ()), in which the trees5 living_ 1二1+2are weighted by inverse distance, explained 52% of(2) the variation in tree fine root length density. Indices I2Sll .and I, which include also diameter, further increased1+12the R' to 35% and 58%, respectively and linearised therelationship between I3 and root density (Fig. 4). TheSliving_ dresiduals (3 versus fine root length) spread nicely and_ 1+1(3)their variability is directly related to I3 as expected,Sall。due to the small-scale variability. Figure 5 shows the2" 1+1spatial distribution of the relative root length densitypredicted by I3. The tree root gap was only slightlys iving.《2smaller than the canopy gap.I3 =-(4)The root length densities of shrubs, as well as ofSall_d-herbs, did not show a clear response to gap creationcompared with trees. As well, the root densities ofwhere l is the distance to the tree and d the diameter atthese vegetation groups showed a lot of variability(Table 1). Although the foliage of the herb layer re-1.3 m height.sponded strongly to gap creation, their root systemsdid not show a similar response (Table 3).2.6 Statistical analysis3.2 Fine root length density under forest canopyThe effects of sampling year and position in the gapand in the gapand under the canopy were estimated by means ofANOVA. Post-hoc Kramer- -Tukey HSD multipleThe mean fine root length density of trees was 2,622comparisons were used to test the differences betweenm'm-2 (148 g'm“) under the forest canopy. In the gapmeans. The root densities were transformed by a loga-the fine root length density of trees was 151 m:m~rithmic transformation (n+1) in order to homogenize(9.0 g'm ). The mean fine root densities of shrubs andthe variances before testing. Spearman rank correla-nerbs under the canopy were 864 m:m- (62 g'm-tion coefficients were used to describe the intensity ofassociation. Linear regression models were fitted us-Table 1 The fine root length density (m:m7 3) of the three spe-ing ordinary least squares. The statistical analysis wascies groups, i.e. trees, shrubs and herbs in the gap and undercarried out using Microsoft Excel 2003 and SPSS.the canopy in a Picea mongolica forest, Baiyinaobao, InnerMongolia Autonomous RegionSpecies Under the canopyIn the gap3 Results199820012000(n=30)_ (n=48) (n=25)__ (n=22) (n=28)3.1 Response of fine roots to the gapTrees4,099* 2,763* 2,296*309b101bShrubs627843* 312° 1 ,067*2978The fine root (less than 2 mm in diameter)Herbs6,021° 3,123be 1,108b 2,543c 2,974alength/density ratio (m:m~") of trees in the gap wasNote: samples contained the humus layer and top 20 cm of thereduced two years after the gap creation, comparedmineral soil. The 1998 sampling was carried out before thewith the pre-harvest situation. On the other hand, theregap creation. Trees contain roots of Picea mongolica, Betulawas no significant change in fine root length densityplatyphylla and Populus davidiana; shrubs include largelyin the gap between two and three growing seasonsroots of Ostryopsis davidiana, Cotoneaster melanocarpus,after the gap creation (p>0.05, n= =50) (Table 1)Dasiphora fruticosa, and Rosa davurica and herbs contain allHardly any fine tree roots were found in the middle ofthe other roots. The logarithms of the root densities on thethe gap. The living tree roots in the gap were mainlysame row marked with a similar letter do not differ signifi-concentrated within 4.5 m of the standing gap edgecantly at the 0.05 risk level according to the Tukey- Kramertrees. All the same, some roots were found up to 10 mtest. Differences中国煤化工iable fllowldaway from the edge of the gap into the gap (Fig. 3).by the same lettet, pS0.05).YHCNMHG.Zou Chun-jing et al: Fine roots rfillig process in an artificial gap in a Picea mongolica forest23Table 4 Fine root length density of three species groups sepa-length density using the derived tree indices (Eqs. (1)- (4)) andrately processed by soil horizons from samples collcted in adistance to the nearest living tree (Eq. (1))Picea mongolica forest between 1998 and 2001 (unit: m:m 9)_Explanatory variableDataStratumLayerIn the gapUnder the canopy1,2980.35(n=50)(n=103)1,2020.47TreeHumus57+191,204+2131,1760.520-10 cm48+231,120+1870.5510- 20 cm298+793_1,0560.58Shrub523+124601+98Note: In deriving the models only samples which have re-0--10cm.201+65187+84moved and living trees among the 25 nearest trees were used,10- -20 cm32+976+19n=92.Herb1,986+237923+106746+120321+76264+76Table 3 Gap fractions of the shrub and herb canopies in a151+342,622+ 199Picea mongolica forestTotal756+121864+89Year Under the canopy (n=45)In the gap (n=26)2,568+1033,086+193,2000 0.82*0.47”2001 0.73*Note: The gap of 36 m diameter was cleared in April 1999.Table 5 Spearman correlation coefficients of the fine rootThe gap fractions marked with the same letter do not differlength density between three soil layersStratum Layerferences between values of the same variable followed by the0.18same letter are not significant (Tukey test, p<0.05).0.41**0.39*0.59***0.40**0.43**and 3,086 m'm- (65 gmi ). The mean fine root den-0-10cm0.78***sities of shrubs and herbs in the gap were 756 m:m~0.45**0.44**(52 g*m7) and 2,568 m*m2 (47 gm 3). The fine rootNote: All the samples located under the forest canopy. Corre-length densities of shrubs and herbs were not signifi-lations marked with ** are staistically significant p<0.01 andcantly affected by the gap (Table 4). Roots less than 2*** have p<0.001 (n=94).mm in diameter comprised 94% of the total tree rootlengths. All herb roots and 95% of the shrub roots .were less than 2 mm in diameter.4 DiscussionAccording to our previous study (Xu and Zou,1998), root systems of plant in sandy land are shallow.The estimate of fine root biomass of a Picea mongo-In this research, 88% of the fine tree roots, 93% of thelica stand agreed with the findings of Xu and Zoushrub roots and 90% of the herb roots were found in(1998), 200- 350 gm 2 for roots<2 mm in the sprucethe humus layer and in the 0 -10 cm mineral soil layer.forest. The results suggest that in the studied PiceaThe fine root diameter of shrubs and herbs did changemogolica forest the“ffective" extent of fine roots ofsignificantly with depth in the soil. However, thetrees was limited to the zone of about 4.5 m from themean diameter of tree fine roots was larger (0.8 mm)gap edge into the gap. Since the mean stem to stemin the 10 -20 cm mineral soil layer than in the humusdistance in the forest stand was 4.5 m, our results im-layer (0.6 mm).ply that the effective root system of an individual treeThe root length density showed high, small-scaledoes not extend much further than to the adjacentspatial, variability. The root densities of nearby sam-neighboring trees. Similar results have also been ob-ples were not correlated although their distance wastained in other studies. In a dense lodgepole pine for-only 0.4 m. Neither did the distance to the nearest liv-est Parsons et al. (1994) found that the root systeming tree correlate with root length density in any stra-extended 4 5 m from a tree and that 40% of the finetum. The root densities of the three plant groups, trees,roots within the rooting zone of an individual treeshrubs and herbs were not correlated under the canopy.were its own. Ammer and Wagner (2002) estimatedHowever, the root densities were correlated betweenthat the roots of Norway spruce extend about 8 mlayers. The only exception was tree fine root lengthfrom the stem in a stand with a density of 545 stemsdensity between humus and the top 10 cm of the min-per hm' but close to 15 m in a stand where densityeral soil layer. The correlation was stronger betweenwas 392 stems per hm*. In a beech forest Bauhus andhumus and the top 10 cm mineral soil than betweenBartsch (1996)中国煤化工m dstancethe mineral soil layers (Table 5).from the gapdensity wasone-third of thaYHCNMHGrestandthat.24Forestry Studies in China, Vol.9, No.1, 2007500070400060三30005C0.90.4(豆10003020Distance (m)10Fig. 3 The effect of the nearest living tree on tree fine rootlength density in the combined humus layer and the top 20 cm)10203040506070of mineral soil. The solid line indicates the kernel smoothedFig.5 The estimated distribution of relative fine root length ofroot length density (m*m ) in the gap and the dashed linetrees strata in the canopy gap using index 13. Open circles (inindicates the kernel smoothed root length under the canopy.the shade) denote cut trees and the gray circles (outside theshade) living trees. The size of the circle is directly related tocrown size.6000、5000index, I3, was based on the idea that fine root biomassg 4000is directly related to the cross-sectional area of the3000 Istem and distributed over the area within distance. The2000data were not very effective for testing the models,since there was not much variation in tree diameter ordensity, making the testing of more complicated root-ing models fruitless. The small-scale variation under0.20.40.6 0.8 1.0the intact forest is large and this variance componentcannot be explained by locations and dimensions ofFig. 4 Relationship between the tree index I3, affecting roots,the trees. Thus the coefficients of determination areand fine root length density of the tree strataunlikely to be much higher with any index formula-tion using only tree positions and sizes as long as thetest data are balanced regarding the index. Besides, theroot length density was very low at 10 m from theactual root systems have an irregular shape.edge of the gap. However, the extent of the root sys-The foliage of shrub and herb layer respondedtem may vary as a function of site type, tree speciesstrongly to the creation of the gap, but their root sys-and stand structure (Stone and Kalisz, 1991). The ex-tems did not show a similar response (Tables 1 and 3). .tent of the root system can be larger in poor soils andOne reason for this may be that the root:shoot ratio ofsparse stands than in fertile soils and in dense stands.these plants changed radically, indicating a higher rootFor example, the root systems of Scots pine growingactivity per unit length. In addition, the large spatialon dry sites have been reported to extend 5- 10 m fromvariation and unknown effect of gap creation on tem-individual seed trees. Brockway and Outcalt (1998)poral variability and turnover of herb root length den-found that root competition precludes regenerationsity may also play a role in these results. Root densi-within a 12-16 m distance from the edge in Pinusties of trees and herbs were similar under an intactpalustris forests. Coarse woody debris may also haveforest canopy probably due to the open canopy. Herbsan effect on the extent of lateral roots (Vogt et al.,are known to be able to achieve much higher root1995).densities than conifers, indicating that ground layerUnder the canopy, root length density was inde-vegetation was not far from the maximum root densi-pendent of the proximity of the nearest tree, indicatingties (Liao et al., 2000; Liang et al, 2001).a rather homogeneous root length density on the scaleThe rooting profiles of the three vegetation groupsof 0.4 5 m. Most of the horizontal variation of rootwere similar and shallow, although the root system oflength density was already present on the scale of 0.4the trees was slightly deeper than that of shrubs andm in an intact forest. This agrees with the absence ofherbs (Table 4). The vegetation groups occupied theany correlation between tree proximity and root lengthsame soil layer in the Picea mongolica forest, whichdensity of any strata. The distance to the nearest treecontrasts with Pinus sylvestris stands, where tree rootscould not explain all the gap scale variation of rootare deeper tha中国煤化王(Makkonenlength density. The other indices, I-, performed and HelmisaarHallbaicken,better but differed only slightly (Table 2). The“best”1999). The meMHCNMHGreesalsoin-.Zou Chun-jing et al: Fine roots rfillig process in an artificial gap in a Picea mongolica forest25creased with soil depth. Since the separation of theNos. 39900019 and 30070129). We thank Prof. Liaoroots from mineral soil was easier than from humus,Liping for English corrections.the increase in root diameter was not likely to be aresult of the loss of the highest order of fine rootbranches.ReferencesThe correlation between root length density and soillayers was weak for the tree roots, which was proba-Ammer C, Wagner S.2002. Problems and options in modellingbly due to differences in the architecture of the rootfine-root biomass of single mature Norway spruce trees atsystems, given that herbs often form hummocks whilegiven points from stand data. Can. J. For. Res., 32: 581-590the tree root system is formed in layers. The root sys- Bauhus J, Bartsch N. 1996. Fine-root growth in beech (Fagustems of most boreal tree species seem to be stronglysylvatica) forest gaps. Can. J. For. Res. 26, 2,153- 2,159laterally spread out. The absence of correlation be-Bazzaz F A, Wayne P M. 1994. Coping with environmentaltween root densities of vegetation groups was surpris-heterogeneity: the physiological ecology of tree seedling re-ing, since the root systems of many plant associationsgeneration across the gap-understory continum. In: Cald-are known to be segregated. However, a simple corre-well M M, Pearcy R W (eds.), Exploitation of Environmentallation coefficient is perhaps an inappropriate indicatorHeterogeneity by Plants. San Diego: Academic Press,of the attraction or repulsion of root systems of dif-349- -390ferent species, given that most of the variation mightBergeron Y, Engelmark O, Harvey B, Morin H, Sirois L.1998.be due to small-scale soil heterogeneity of nutrient andKey issues in disturbance dynamics in boreal forests. J.water availability. The soil organic matter contentVeget. Sci, 9: 463- -610correlates with water holding capacity and amount ofBrockway D G, Outcalt K W.1998. Gap- phase regeneration inavailable nutrients and the amount of roots over a longlongleaf pine wiregrass ecosystems. For. Ecol. Manag, 106:time period. The formation of soil heterogeneity is a125- -139result of continuous interaction between soil, plantCaldwell M M. 1987. Competition between root systems inroots and soil fauna. Root segregation is known to beplant communities. Soc. Exp. Biol. Semin. Ser, 30: 167-185most common in water-limited ecosystems.CampbellJ J, Finer L, Messier C.1998. Fine-root production inNo sign of recovery of the tree root system in thesmall experimental gaps in successional mixed boreal forests.gap was detected during follow-up inspections two toJ. Veget. Sci, 9: 537- -542three years after the gap creation. This suggests that Chen J, Franklin J F, Spies T A. 1993. Contrasting microcli-the edge trees are not very rapid and aggressive inmates among clearcut, edge, and interior of old-growthextending their root systems into the formed root gap.Douglas-fir forest. Agric. For. Meteorol, 63: 219- -237This may be due to several factors. First, the edgetrees may be adversely affected by the gap due to in-climatic gradients from clearcut edges into old-growthcreased bending stress (Dean, 2001 ), changed micro-Douglas-fir forests. Ecol. App,5: 74 -86climate and increased number of herbivors. Second,Collins B S, Pickett S T A.1988. Demographic responses ofgap creation is likely to reduce rooting density close toherb layer species to experimental canopy gaps in a northernthe edge trees, providing easily accessible soil re-hardwoods forest. J. Ecol, 76: 437- 450sources in the vicinity of the trees. In addition, higherDean T J. 2001. Potential effect of stand structure on below-precipitation and elevated soil temperatures at the gapground allocation. For. Sci, 47: 69-76edge may increase the availability of nutrients, thusDeChantal M, Leinonen K, Kuuluvainen T, Cescatti A. 2003.reducing the need for root foraging for nutrients fur-Early response of Pinus sylvestris and Picea abies seedlingsther away. As a consequence, the edge trees may in-to a experimental canopy gap in a boreal spruce forest. For.crease their rooting density within their previousEcol. Manag, 176: 321-336rooting zone, while the most extensive parts of theDrexhage M, Gruber F. 1999.root system may even die. The results of our studytionships for Picea abies: estimating root system biomasssuggest that in sandy spruce forests in the sandy for-from breast-height diameters. Scand. J. For. Res, 14:est-steppe ecotone a gap disturbance creates a distinct328- 333tree root gap that can be roughly described by a lineconnecting the cut down trees closest to the gap edgenecromass in limed and fertilized Norway spruce (Piceaand that the gap edge trees do not rapidly extend theirabies (L.) Karst.) stands. For. Ecol. Manag., 119: 99- 110root systems into the formed root gap.Jones R H, Mitchell R J, Stevens G N, PecotS D.2003. Con-trols of fine root dynamics across a gradient of gap sizes in apine woodland. Oecologia, 134: 132- 143AcknowledgementsKuuluvainen T. 1994. Gap disturbance, ground microtopogra-phy, and the regeneration dynamics of boreal coniferous for-We thank Mr. Liu Guangtian and the staff of Baiyi-ests in Finland中国煤化:35-51naobao Nature Reserve. This study was funded by the Lertzman K P.L placement in aNational Natural Science F oundation of China (Grantsubalpine, old-MHCNMHG69.2Forestry Studies in China, Vol.9, No.1, 2007LiX G, He W M, Dong M.1997. A preliminary study on gap22(6): 64 68 (in Chinese with an English abstract)dynamics of Gordonia acuminata population in Jinyun Taskinen O, Ilvesniemi H, Kuuluvainen T, Leinonen K. 2003.Mountain. Acta Ecol. Sin, 17(5): 543- -548 (in Chinese withResponse of fine roots to an experimental gap in a borealan English abstract)Picea abies forest. Plant Soil, 255: 503- -512Liang X D, Ye W H, Yi W M. 2001. Gap dynamics and theVogt K A, Vogt D J, Asbjornsen H, Dahlgren R A. 1995. Roots,maintenance of biodiversity. Chin. J. Ecol, 20(5): 64- 68 (innutrients and their relationship to spatial patterns. Plant Soil,Chinese with an English abstract)168: 113-123LiaoL P, GaoH, YuX J, Wang S L. 2000. Nutrient retranslo-Wilczynski C J, Pickett S T A.1993. Fine root biomass withincation in fine roots of Cunninghamia lanceolata, Alnus cre-experimental canopy gaps: evidence for a below-ground gap.mastogyne and Kalopanax septemlobum in the mixed planta-J. Veget. Sci,4:571-574tions. Chin. J. Appl. Ecol, 11(2): 161-164 (in Chinese withXu W D, Chang Y. 1992. Model of zonal climax vegetationtypes in Northeast China. Chin. J. Appl. Ecol., 3: 215- 222LiuJ J, Wang D X, Lei R D. 2002. Turnover process and en-(in Chinese with an English abstract)ergy change of fine roots of Pimus tabulaeformis and Quer- Xu W D, Li W D, Zheng Y. 1994. The taxonomy of Piceacus aliena var. acuteserrata natural forests in Qinlingmongolica in Inner Mongolia. Bull. Bot. Res, 14(1): 59- -68Mountains. Sci. Silv. Sin, 38(4): 25- -30 (in Chinese with an(in Chinese with an Engish abstract)English abstract)Xu W D, Liu GT, Duan P S, ZouC J. 1998. Study on PiceaMakkonen K, Helmisaari H. 1998. Seasonal and yearly varia-mongolica Forest Ecosystem in Baiyinaobao Natural Re-tions of fine-root biomass and necromass in a Scots pineserve, Inner Mongolia. Beijing: China Forestry Publishing(Pinus sylvestris L.) stand. For. Ecol. Manag, 102: 283- -290House (in Chinese)McGuire J P, Mitchell R J, Moser E B, Pecot S D, Gjerstad D Xu W D, Zheng Y R, Liu G T. 1993. Relationship betweenH, Hedman C W. 2001. Gaps in a gappy forest: plant re-growth and ecological conditions of spruce in sandy land.sources, longleaf pine regeneration, and understory responseChin. J. Appl. Ecol, 4(4): 368- -373 (in Chinese with an Eng-to tree removal in longleaf pine savannas. Can. J. For. Res.,lish abstract)31: 765- -778Xu W D, Zheng Y R. 1993. Relationship between seedlingMessier C, Puttonen P.1995. Spatial and temporal variation ingrowth and dry matter production of spruce in sandy land.the light environment of developing Scots pine stands: TheChin. J. Appl. Ecol, 4(1): 1-6 (in Chinese with an Englishbasis for a quick and efficient method of characterizing light.Can. J. For. Res., 25: 343- -354Xu W D, Zou C J. 1998. Sandy Forest Ecosysterms of China.Parsons W F J, Miller S L, Knight D H. 1994. Root-gap dy-eiing: China Forestry Publishing House (in Chinese)namics in a lodgepole pine forest: ectomycorrhizal ancXu W D. 1985. Kira's thermal index and its application onnonmyeorrhizal fine root activity after experimental gapvegetation in China. Chin. J. Ecol, 4: 35 -39 (in Chinese)formation. Can.J. For. Res., 24: 1,531-1,538Zou C J, Han S J, Xu W D. 2001. Modular dynamics andPhillips D L, Shure D J. 1990. Patch-size efects on early suc-structure of Picea mongolica. J. Wuhan Bot. Res, 19(5):cession in southern Appalachian forests. Ecol, 71: 204 -212369- -376 (in Chinese with an English abstract)Roberts J. 1976. A study of root distribution and growth in aPimus sylvestris L. (Scots pine) plantation in East Anglia.and overcompensation of Picea mongolica seedlings dam-Plant Soil, 44: 607- 621aged by defoliating insects. Chin. J. Appl. Ecol, 22(3):Stone E L, Kalisz P J.1991. On the maximum extend of tree269- 274 (in Chinese with an English abstract)Tan X, SunX Y, Yan H P. 2000. Canopy gaps in the pine for-(Received July 6, 2006 Accepted October 18, 2006)ests of the West Mountain in Beijing. J. Beijing For. Univ,中国煤化工MHCNMH G.

论文截图
版权:如无特殊注明,文章转载自网络,侵权请联系cnmhg168#163.com删除!文件均为网友上传,仅供研究和学习使用,务必24小时内删除。