research article effect of disturbance regimes on spatial ...research article effect of disturbance...

Research ArticleEffect of Disturbance Regimes on Spatial Patterns of TreeSpecies in Three Sites in a Tropical Evergreen Forest in Vietnam

Do Thi Ngoc Le12 Nguyen Van Thinh13 Nguyen The Dung2 and Ralph Mitloumlhner1

1Tropical Silviculture amp Forest Ecology Georg-August-Universitat Gottingen Busgenweg 1 37077 Gottingen Germany2Vietnam Forestry University Xuan Mai Town Chuong My District Hanoi 100000 Vietnam3Silvicultural Research Institute (SRI) Vietnamese Academy of Forest Sciences Duc Thang Bac Tu Liem DistrictHanoi 100000 Vietnam

Correspondence should be addressed to DoThi Ngoc Le dothingocle81yahoocom

Received 7 August 2015 Revised 3 November 2015 Accepted 17 December 2015

Academic Editor Scott D Roberts

Copyright copy 2016 DoThi Ngoc Le et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

The effects of disturbance regimes on the spatial patterns of the five most abundant species were investigated in three sites in atropical forest at Xuan Nha Nature Reserve VietnamThree permanent one-ha plots were established in undisturbed forest (UDF)lightly disturbed forest (LDF) and highly disturbed forest (HDF) All trees ge5 cm DBH were measured in twenty-five 20m times 20msubplots A total of 57 tree species belonging to 26 families were identified in the three forest types The UDF had the highestbasal area (30m2 haminus1) followed by the LDF (17m2 haminus1) and the HDF (130m2 haminus1) The UDF also had the highest tree density(751 individuals haminus1) while the HDF held the lowest (478 individuals haminus1) Across all species there were 417 ldquojuvenilesrdquo 267ldquosubadultsrdquo and 67 ldquoadultsrdquo in the UDF while 274 ldquojuvenilesrdquo 230 ldquosubadultsrdquo and 36 ldquoadultsrdquo were recorded in the LDF 238ldquojuvenilesrdquo 227 ldquosubadultsrdquo and 13 ldquoadultsrdquo were obtained in the HDF The univariate and bivariate data with pair- and mark-correlation functions of intra- and interspecific interactions of the fivemost abundant species changed in the three forest typesMostspecies indicated clumping or regular distributions at small scale but a high ratio of negative interspecific small-scale associationswas recorded in both the LDF and HDF sites These were however rare in the UDF

1 Introduction

A forest stand comprises a set of trees characterized by theirlocations and sizes Tree diameter distributions can provideinformation on tree sizes but cannot address tree locationsHowever tree diameters are associated with tree positionsand growth is sensitive to both spatial interaction among trees[1] and local habitat characteristics [2] The theory of markedpoint processes provides a formal framework for an analysisof the spatial characteristics of tree diameter distributionsin which the points indicate tree locations and the marksdenote particular tree characteristics such as diameter atbreast height tree height and growth during a given timespan [3 4] Ecologists have become increasingly interested instudying spatial patterns in ecology [5ndash8] In addition treespecies associations at different life stages or age classes havealready received considerable attention [5 9ndash11] The spatialpattern of a particular species especially the adult-juvenile

relationship provides useful information on the speciesrsquoregeneration process [10 12 13]The spatial pattern of trees isan important indicator of standhistory population dynamicsand species interaction in forests [14 15] all of which arehelpful in understanding mechanistically the processes andpatterns of plant communities [16]

Spatial distributions and the spatial patterns (both verti-cal and horizontal) of trees in forests are important elementsfor understanding forest ecosystem dynamics [17] howeverthe potential for ecological understanding has not yet beenfully recognized [18] A spatial point pattern is a set oflocations distributed within a region of interest which havebeen generated by some spatial process [19] Many methodsand indices have been developed in order to interpret andassess spatial distributions [20ndash23] The spatial distributionpattern within plant populations is influenced by variousecological and evolutionary processes which take place dur-ing the life history of a plant such as seed dispersal intra- and

Hindawi Publishing CorporationInternational Journal of Forestry ResearchVolume 2016 Article ID 4903749 16 pageshttpdxdoiorg10115520164903749

2 International Journal of Forestry Research

interspecific competition and environmental heterogeneityThe spatial structure of a forest is largely determined by therelationships within neighboring groups of trees [24] Themain factors in a forest structure are the spatial distributionspecies diversity and variations in tree dimensions [25] Inrecent decades several structural indices such as the Clarkand Evans aggregation index [20] pair-correlation function[26] and species diversity indices [13] have been developedto quantify spatial forest structure

Tropical rain forests have been used in recent studies todemonstrate the effects of high diversity on spatial distribu-tion [5 27]This explains why interspecific species have func-tional similarity and may adapt to average environment con-ditions [28] One approach to exploring spatial plant dynam-ics is to use a point pattern analysis of fully mapped plantlocations [29] Spatial statistics like Ripleyrsquos 119870 function [23]and the pair-correlation function [26 30] quantify the small-scale spatial correlation structure of a pattern which containsinformation on positivenegative interactions among plantsIn addition point pattern analysis is ideally suited to controlfor environmental heterogeneity and focus on neighborhoodprocesses Spatial patterns of individuals within populationsare closely linked to ecological processes consequently eco-logical processes may be deduced from spatial patterns [31ndash33]

There is little published information on the spatial pat-terns of tree species in tropical evergreen forests as affectedby different disturbance regimes in Vietnam In the presentstudy univariate 119892(119903) statistics with the null model of com-plete spatial randomness and the mark-correlation functionwere used for point pattern analysis at different scales Weanalyzed the intraspecific and interspecific association (rela-tionship or correlation) of five abundant tree species in threeforest types in a tropical evergreen forest in Xuan Nha NatureReserve VietnamThis study will providemeaningful knowl-edge for predicting the spatial patterns of the most dominanttree species after disturbance and could have further silvi-cultural implications for management practice (eg selectingnative tree species and choosing site conditions for enrich-ment planting) in restoration zones of the nature reserve

2 Materials and Methods

21 Study Site The study site (Figure 1) is located in XuanNha Nature Reserve (20∘361015840ndash20∘481015840N 104∘291015840ndash104∘501015840E)The reserve has a total area of 27100 ha of which 15300 aretropical evergreen forests and 2600 are tropical forests onlimestone [34]Thenature reserversquos elevation ranges from 300to 2000m above sea level The topography is composed oftwo shallow-sided valleys which run across the reserve fromthe Laotian border Xuan Nha Nature Reserve is located ina tropical monsoon climate where the mean temperature is213∘C The rainy season lasts from May to October (witha maximum of 3324mm of rain in August) while the dryseason goes from November to April (with a minimum of324mm in February) Mean annual precipitation is about1673mm Six major soils are found in the study site (1) thicklayers of dark yellow ferralsols at 700ndash1700m in altitude thatdevelop on shale or metamorphic rocks with a medium

QuynhNhai

Thuan Chau

Muong La

Bac YenPhu Yen

East Sea

Moc Chau

Yen Chau

Mai SonSong Ma

Sop Cop

Son La City

Xuan Nhanature reserve

Cambodia

LAOS

N China

Figure 1 The location of Xuan Nha Nature Reserve in VietnamDark blue yellow and green colors indicate the undisturbed forest(UDF) lightly disturbed forest (LDF) and highly disturbed forest(HDF) respectively

texture (2) thin layers of yellow-brown ferralsols at 700ndash1700m altitude which develop on limestone ormetamorphiclimestone and have coarse tomedium texture (3) thick layersof light yellow ferralsols at 700ndash1700m on shale ormetamor-phic rockswithmedium texture (4) thick layers of lightgrey-yellow ferralsols that develop at 300 to 1000m in low hilly ormountainous regions on slate siltstone sandstone and con-glomerate with a coarse to medium texture (5) grey-yellowferralsols modified by paddy fields with medium texture thatappear in the surrounding villages and (6) alluvial soilsdeposited at the foot of mountains river banks and streams[35]

In accordance with the collected data approximately 43of the forest area (app 787220 hectares) is disturbed while428 (78214 ha) is classified as undisturbed (Table 1) Thisdisturbed area is divided into two categories lightly disturbedand heavily disturbed Undisturbed and disturbed forestsrefer to different disturbance regimes where ldquodisturbancerdquo isdefined as the impact level of human activities in the forestDisturbance in this context refers to selective timber harvest-ing felling small-sized trees for nontimber products and past(pre-2003) illegal logging

Undisturbed forests (primary or rich forests) are areasthat do not show evidence of damage from human activitiesthese are relatively stable forests not yet (or less) influenced byhumans or natural disasters Because such forests are limitedto inaccessible and protected areas they are not used for pro-duction purposes UDFs are extremely rich forests with a tim-ber volume of standing trees between 201 and 300m3 haminus1LDFs (average forest) are influenced by humans or naturaldisasters leading to changes in their structure This typedemonstrates low and minor damage from human beingsand has a timber reserve of standing trees between 101 and200m3 haminus1 Forests may be classified as LDFs if past logging

International Journal of Forestry Research 3

Table 1 Stratification of vegetation types and characteristics of three main forest types at Xuan Nha Nature Reserve Vietnam

Forest type Area (ha) Species richness (119873ha) Family richness (119873ha) Tree density (119873ha) Basal area (m2ha)Undisturbed (rich forest) 78214 49 25 751 298Lightly disturbed (average forest) 22885 42 21 540 169Heavily disturbed (poor forest) 21795 30 18 478 130Wood and bamboo mixed forest 4832 mdash mdash mdash mdashBamboo forest 29210 mdash mdash mdash mdashLimestone forest 15494 mdash mdash mdash mdashNonforested lands 10245 mdash mdash mdash mdashTotal 182675

has resulted in the removal of only 6 to 10 of trees abovea minimum harvestable size (40ndash50 cm DBH) HDFs (pos-texploitation forests poor forests) are degraded as a resultof human activities that have severely impacted their canopystructure productivity and volume at various levels Suchforests have been exploited for their timber or other forestproducts 10 to 30 of trees above a minimum harvestablesize (40ndash50 cm DBH) have been removed and they have areserve of standing trees between 10 and 100m3 haminus1

The reserve has a high tree species diversity with 173families and 1074 species The main vegetation observedat Xuan Nha Nature Reserve is monsoon evergreen broad-leaved forest numerous rare species of flora have been foundhere including 65 rare tree species that account for 61 ofthe total species listed in Vietnamrsquos Red Book [35] At theUDF Lithocarpus ducampii is the most dominant speciesfollowed by Syzygium cuminii Vatica odorata Toxicodendronsuccedaneum and Cinnamomum parthenoxylon Lithocarpusducampii Vatica odorata and Syzygium cuminii dominatedin the LDF while Engelhardtia roxburghiana and Nepheliummelliferum could be found in the HDF In the UDF the fivemost dominant and species-rich families were LauraceaeFagaceae Myrtaceae Dipterocarpaceae and AnacardiaceaeFagaceae was the most important family in the LDF followedby Lauraceae Sapindaceae Dipterocarpaceae and Myr-taceae Anacardiaceae dominated only in the UDF Juglan-daceae was restricted to the HDF The top family in the HDFwas Lauraceae followed by Dipterocarpaceae SapindaceaeJuglandaceae and Fagaceae Across the three forest typesLauraceae was the most dominant but Fagaceae Myrtaceaeand Dipterocarpaceae were also common families The treedensity of the LDF was lower than that of the UDF whichhad the highest density (751 trees haminus1) of all treesge5 cmDBHamong the three forest sites the HDF held the lowest treedensity (478 stems haminus1)

22 Sampling Design Three permanent one-ha plots (100mtimes 100m each) were established in undisturbed lightly dis-turbed and highly disturbed forests (one plot each) in 2003The corners of each plot were marked with a concrete postand a GPS device Each one-ha plot was further divided intotwenty-five 400m2 subplots (20m times 20m) (Figure 2)

23 DataCollection Datawere collected in 2013Within eachplot all living trees larger than 5 cm in diameter at breast

I II

1 2 3 4 50

6 7 8 9 10

0

11 12 13 14 15

16 17 18 19 20

21 22 23 24 25

IV III

N1

N2

y1x1

y2

x2

20m20m

20m

20m

Figure 2 Sampling design (sample size 1 ha) Symbols I II III andIV are the four corners of the sampled plot 25 subplots (20m times20m) were set up in order to measure the coordinates of all treeswith DBH ge5 cm 119873

1(1199091 1199101) 1198732(1199092 1199102) are the tree coordinates

sampled

height (DBH 13m) were marked with paint and their DBHand height weremeasured All individuals were classified intothree life history stages ldquojuvenilesrdquo (5 cm le DBH lt 10 cm)ldquosubadultsrdquo (10 cm le DBH lt 30 cm) and ldquoadultsrdquo (DBH ge30 cm) [36] Tapes were used to measure the coordinates oftrees as follows the starting point in each subplot (400m2)was fixed at the northwestern cornerTheposition of each tree(119883119884 coordinates)was determined bymeasuring the distanceto the two edges of the plot (Figure 2) All trees were recordedand labelled in order to avoid missing any ones Tree specieswere identified in the field with specialistsrsquo help unidentifi-able specimenswere taken to theVietnamForestryUniversityHerbarium for identification

24 Data Analysis Tree density (stemhaminus1) was calculatedfrom the count of all individuals from the 25 subplots treebasal area (BAm2 haminus1) was calculated by using the followingequation BA = (314 times (DBH)2)4 (m2) The total basal areaper ha was calculated by the sum of the BA of all trees inthe 25 subplots Species richness was taken by counting thenumber of species occurring in all subplots of each forest type[22] The R-package software was used to measure nearestneighbor distance (NND) [37] Only the five most abundanttree species in the three forest types were used to simulatespatial distributions and the relationships among them Pair-correlation functions 119892(119903) mark-correlation functions andnullmodels were used to analyze spatial patterns in this study


241 Pair-Correlation Functions The pair-correlation func-tion was used as a summary statistic to quantify the spatialstructure of the uni- and bivariate patterns [8 26] Based onintertree distances the univariate pair-correlation function11989211(119903) can be used to determine whether a point distribution

is random aggregatedclustered or regular at distance 119903 inwhich those patterns occur The parameter 119892(119903) indicateswhether a pattern is random (complete spatial randomness(CSR) 119892(119903) = 1) clumped (119892(119903) gt 1) or regular (119892(119903) lt 1) ata given radius 119903

The pair-correlation function 11989211(119903) for the univariate

pattern of species 1 can be defined based on the neighborhooddensity 119874

11(119903) = 120582

111989211(119903) that is the mean density of trees

belonging to species 1 surrounded by rings with radius 119903and width 119889(119903) [8] where 120582

1is the intensity (number of

species 1rsquos trees in the plot) Under the null model of completespatial randomness (CSR) the points are independently andrandomly distributed over the entire plot [8] The values of11989211(119903) within the 95 confidence envelope indicate that the

spatial structure of the given distances does not differ signif-icantly from the CSR Values of 119892

11(119903) above 95 confidence

envelope indicate that a distance class ismore aggregated thanunder CSR values of 119892

11(119903) below the 95 confidence enve-

lope indicate lower aggregation (approaching a more regularpattern) at that scale [38] The pair-correlation function forbivariate patterns (composed of trees from species 1 and 2)follows where the quantity 119892

12(119903) is the ratio of the observed

mean density of species 2rsquos trees in the rings around species 1rsquostrees to the expectedmean density of species 2rsquos trees in thoserings [8] The corresponding neighborhood density functionyields 119874

12(119903) = 120582

211989212(119903) 11989212(119903) = 1 shows independence

(no interaction) 11989212(119903) gt 1 indicates attraction between

two point patterns at distance 119903 and 11989212(119903) lt 1 indicates

repulsioninhibition between the two patterns at distance 119903

242 Effect of Interspecific Competition on Tree Growth Amark-correlation function (MCF) 119870

119898119898(119903) [26 39] using

DBH as marks was used to analyze the distance-dependentsize correlation of trees for distances up to 50mThe similar-ity or dissimilarity between the DBH marks of two trees at adistance 119903 apart is quantified by the equation 119891(119898

1 1198982) =

1198981times 1198982 where 119898

1and 119898

2are the DBH values of two

neighboring trees119870119898119898(119903) is defined as the normalized mean

value of 119891(1198981 1198982) for all marks at distance 119903 Marks are con-

sidered independent and positively or negatively correlated atdistance 119903 if 119870

119898119898(119903) = 1 119870

119898119898(119903) gt 1 or 119870

119898119898(119903) lt 1

respectively

243 Null Models The univariate statistic is used to analyzethe spatial pattern of one object while the bivariate statisticis used to analyze the spatial association of two objects(pattern 1 and pattern 2) [8 40] Based on a homogeneitytest on the spatial pattern of all adult trees (DBH ge 10 cm)we applied complete spatial randomness (CSR) as the nullmodel [8 39] A Monte-Carlo approach was used to test forsignificant departures from the null models Each of the 199simulations of a point process underlying the null modelgenerates a summary statistic simulation envelopes with 120572 lt005 were calculated from the fifth highest and lowest values

of the 199 simulations [26] All analyseswere performedusingthe software Programita [8]

3 Results

31 Population Structure of Individuals in theThree Life StagesThe adults of all tree species were used to test for environ-mental heterogeneity in the three forest types The five mostabundant species (Engelhardtia roxburghiana Lithocarpusducampii Syzygium cuminii Nephelium melliferum and Cin-namomumparthenoxylon Table 2) were analyzed for all otherassociations (eg intra-and interspecific) Across all speciesthere was a higher proportion of ldquojuvenilesrdquo ldquosubadultsrdquoand ldquoadultsrdquo in the UDF than in the LDF or HDF All treespecies across the three life stages are mapped in Figure 3

The diameter distributions of the five most abundantspecies are shown in Figure 4 The frequency distribution ofall tree species demonstrated an inverted J-shaped curve inwhich the number of individuals gradually decreased withincreasing diameter classes a trend found across all threeforest sites The distribution of Engelhardtia roxburghianavaried among the three forest types a unimodal distributionwas found in the HDF (Figure 4(c)) whereas the UDF andLDF had reverse J-shapes (Figures 4(a) and 4(b)) ConverselySyzygium cuminii displayed a unimodal distribution in allthree forest types Both species had no individuals in the sizeclass greater than 30 cm and 884 of the total tree densitycame from the 5ndash20 cm diameter class However a very highabundance of 5ndash10 cm DBH individuals of L ducampii and ahigh number of 10 cm S cuminii trees were observed in theUDF

32 Intraspecific Interactions The spatial patterns of the fivemost abundant species in the three forest types are shown inFigure 5 the univariate 119892(119903) statistics with the null modelof complete spatial randomness (CSR) displayed differentspatial patterns for these species at various scales There wereno statistically significant departures from randomness forEngelhardtia roxburghiana in the HDF (Figure 6(a)) butin the UDF it was clumped at scales of 6m and 9-10m(Figure 5(a))Cinnamomum parthenoxylonwas aggregated atscales of 13 and 28m in the UDF (Figure 5(d)) and randomlydistributed at all distances in both the HDF and LDF sites(Figures 6(c) and 6(d))Engelhardtia roxburghiana in the LDFand Cinnamomum parthenoxylon in the HDF both occurredin larger spatial clusters of trees (gt40m) and were thusevidence for clustering at larger spatial scales Lithocarpusducampii was significantly aggregated at a scale of 9m in theHDF (Figure 5(b)) but predominantly random in the LDF(Figure 5(c)) and UDF (Figure 6(b)) Syzygium cuminii wassignificantly aggregated at a large scale of 39m in the HDF(Figure 5(e)) but showed an independent distribution at allscales in the LDF (Figure 6(e)) no significant pattern wasfound at all scales in the UDF (Figure 6(f)) Nephelium mel-liferum in contrast was randomly distributed at all distancesin the HDF (Figure 6(g)) but tended to have a regular distri-bution at a larger scale (gt50m) in the LDF (Figure 6(h))Thisspecies displayed clumped distribution at two scales of 4 and16m in the UDF (Figure 5(f)) Cinnamomum parthenoxylon


Table 2 The five most abundant tree species of each forest site in Xuan Nha Nature Reserve Vietnam

Species Tree density [119873ha] Basal area [m2ha] Mean DBH [cm] Maximum DBH [cm] Median NN [m]Undisturbed forest

S cuminii 103 217 155 282 1233L ducampii 78 491 224 1195 1495N melliferum 53 175 177 603 1194C parthenoxylon 50 273 231 618 1368E roxburghiana 46 261 212 701 1298

Lightly disturbed forestL ducampii 96 280 154 554 1099S cuminii 60 117 149 282 970C parthenoxylon 50 146 178 414 1158E roxburghiana 44 128 169 427 1329N melliferum 42 127 175 417 1200

Heavily disturbed forestE roxburghiana 61 180 173 570 566L ducampii 50 160 176 420 1270S cuminii 49 089 142 420 1134N melliferum 47 126 170 280 1176C parthenoxylon 42 132 189 320 1245

and Nephelium melliferum in both the HDF and LDF sitesshowed a random spatial distribution at all distances up to50m

The mark-correlation function showed different spatialdistributions of the five most abundant tree species ForEngelhardtia roxburghiana only a negative correlation wasdetected at scales of up to 2m in the HDF (data not shown)but there was a significant positive association in the rangeof 14-15m in the LDF (Figure 7(a)) no statistically significantcorrelationwas found in theUDF Lithocarpus ducampiihad apositive correlation at scales of 26m in theHDF (Figure 7(b))and 1 and 5-6m in the LDF (Figure 7(c)) in the UDF itexhibited a negative association at 16-17 and 48-49m (Fig-ure 7(d)) Cinnamomum parthenoxylon displayed a similarnegative correlation at different scales of 22-23m in the HDF(Figure 7(e)) 3ndash6m in the LDF (Figure 7(f)) and 8m in theUDF (Figure 7(g)) Only two instances of relationship wereobserved for Syzygium cuminii one was positive at 7 and9m the other was negative at a distance of 12m in the HDF(Figure 7(h)) Nephelium melliferum demonstrated positivetrends at a scale of 4m and a negative correlation at a distanceof 8m in the HDF (Figure 7(i)) In the LDF it tended to havea negative correlation at the particular scales of 24 and 47m(Figure 7(j)) and a positive association in the UDF at thescales of 28-29m (Figure 7(k))

33 Interspecific Interactions The results of the bivariatespatial pattern analysis using pair- and mark-correlationfunctions for the five abundant species in the three foresttypes are shown in Figure 8 A total of 20 (5 times 4) bivariatepoint pattern analyses were simulated for pairs of the fivemost abundant tree species Engelhardtia roxburghiana andLithocarpus ducampii showed a trend for positive association(statistical attraction) at scales of 17 and 42m in the HDF

(Figure 8(a)) in the UDF they demonstrated a positivecorrelation at scales of 6 and 15m and a negative associationat a distance of 2-3m (Figure 8(b))However this relationshipwas independently distributed at all scales in the LDF Therewas a significant positive association (attraction) betweenEngelhardtia roxburghiana and Cinnamomum parthenoxylonat distances of 6 and 28m in the LDF (Figure 8(c)) and39m (Figure 8(d)) in the HDF but no such association wasfound at all scales in the UDF Engelhardtia roxburghianashowed no interaction with Syzygium cuminii at all distancesfor both the LDF and HDF however a significant positiveassociationwas detected between the two at the specific scalesof 4 7 10 16 and 21m in the UDF (Figure 8(e)) BetweenEngelhardtia roxburghiana and Nephelium melliferum therewas only a negatively associated trend at scales of 33 and 8min the HDF and UDF (Figures 8(f) and 8(g)) Two instancesof statistically significant departures from randomness wereobtained between Lithocarpus ducampii and Cinnamomumparthenoxylon in the HDF one was towards repulsion distri-bution at a small distance of 5m the other towards positiveassociation at a scale of 22m (Figure 8(h)) In the LDF nospatial interaction of Lithocarpus ducampii and Cinnamo-mumparthenoxylonwas recorded at all distances Lithocarpusducampii and Syzygium cuminii were significantly segregated(attraction) at a small distance of 7m and negatively asso-ciated (repulsion) at 11m in the HDF (Figure 8(i)) incomparison this relationship in the LDF showed significantrepulsion at the scales of 4 and 33m (Figure 8(j)) The onlyclear significant aggregation obtained was for the interactionbetween Lithocarpus ducampii and Syzygium cuminii at thesmall scales of 4 and 7m as well as the larger distances of16 and 28m in the UDF (Figure 8(k)) A similar result wasrecorded for Lithocarpus ducampii andNepheliummelliferumat different distances in the HDF the relationship exhibited a


X (m)

Types

AdultsSubadultsJuveniles

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

Diameter class (cm)10

20

30

40

50

60

70

80

90

100

110

120

130

(a)

X (m)

Types



20

30

40

50

60

70

80

90

100

110

120

130

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

(b)

X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(c)

Figure 3 Distribution of all tree individuals in the three life stages ldquojuvenilesrdquo ldquosubadultsrdquo and ldquoadultsrdquo Symbol size is proportional to theDBH of the individuals with an actual size from 60 to 1292 cm The unit of (119883119884) axes is meter where (a) is the undisturbed forest (b) thelightly disturbed forest and (c) the highly disturbed forest


5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

Engelhardtia roxburghianaLithocarpus ducampiiCinnamomum parthenoxylon

Syzygium cuminiiNephelium melliferum

(a) Undisturbed forest

5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

(b) Lightly disturbed forest



All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)

(c) Highly disturbed forest

Figure 4 Diameter distributions of all living trees and the five major tree species in each forest type in Xuan Nha Nature Reserve Vietnam

significant positive spatial association at scales of 5 and 31mwith a significant negative spatial association at 7 9 and 30m(Figure 8(l))

The complementary analyses using the mark-correlationfunction are shown in Figure 9 Engelhardtia roxburghianaand Lithocarpus ducampii were spatially uncorrelated at allscales in the three forest types but the former displayed anegative spatial interaction with Cinnamomum parthenoxy-lon at scales of 18ndash20m in the HDF (Figure 9(a)) thisinteraction was strongly positive at 7ndash9 21ndash24 and 36m inthe UDF (Figure 9(b)) A significant negative association wasobserved between Lithocarpus ducampii and Cinnamomumparthenoxylon at small scales of 3m and greater than 32-33min the HDF (Figure 9(c)) whereas no spatial interaction wasrecorded in the UDF In the HDF however two tendencieswere obtained between Lithocarpus ducampii and Nepheliummelliferum a positive association at scales of 6 and 8-9mand a significant repulsion at a scale greater than 30m (Fig-ure 9(d)) Cinnamomum parthenoxylon showed a negative

spatial interaction with Syzygium cuminii at the range of 24-25m in the UDF (Figure 9(e)) whereas in the HDF andLDF sites there was only a spatially independent interactionbetween them A similar result was obtained for Syzygiumcuminii and Nephelium melliferum in different scales theyshowed a significant repulsion at spatial distances of 6 10ndash12and 15-16m in the UDF (Figure 9(f))

4 Discussion

41 Intraspecific Competition Environmental heterogeneityand dispersal limitation which usually influence the large-scale spatial patterns of trees are important processes thatlead to an aggregation pattern [41 42] Our investigationrevealed that there was change in the spatial distributionfrom aggregation to randomness (and even to regularity) forthe five most abundant species these displayed a significantdegree of spatial aggregation at several small distances of10 30 and 40m (Figure 4) suggesting the importance of


g11(r)

10 20 30 400Scale (m)

0123456789

(a) E roxburghiana (UDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)

(b) L ducampii (HDF)

g11(r)

10 20 30 400Scale (m)

00051015202530354045

(c) L ducampii (LDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)

(d) C parthenoxylon (UDF)

0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)

(f) N melliferum (UDF)

Figure 5 Spatial patterns of the five most abundant tree species shown by the pair-correlation function 119892(119903) with the null model of CSRBlack lines are observed patterns and grey lines are approximate 95 confidence envelopes

intraspecific competition in governing the spatial distribu-tion and traits of tree populations in these forests Of thefive most abundant species Engelhardtia roxburghiana andSyzygium cuminii had a widely distributed aggregation in thisarea Lithocarpus ducampii was only found to have a regu-lar distribution as succession progressed providing furtherevidence of competition in the development of an evergreenforest structure Hubbell [12] and Condit et al [43] suggestedthat poor seed dispersal is related to greater clumping

intensities of populations however our results showed thataggregations of abundant species were relatively loose [44]At a given scale of observation it has been found that theaggregation pattern occurs more frequently than randomand regular patterns in natural tropical forests where largetrees organize themselves in a regular way [5] However thisgeneral rule did not apply to the Xuan Nha forests

The question of whether trees compete in larger sizeclasses [5 45] is important for our general understanding of


10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)

(a) E roxburghiana (HDF)

g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)

(b) L Ducampii (UDF)

g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)

(c) C parthenoxylon (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)

(d) C parthenoxylon (LDF)

g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789

(g) N Melliferum (HDF)

g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(h) N Melliferum (LDF)



10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)

(a) E roxburghiana (LDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(d) L ducampii (UDF)

10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)

(e) C parthenoxylon (HDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)

(f) C parthenoxylon (LDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(g) C parthenoxylon (UDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)

(i) N melliferum (HDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)

(j) N melliferum (LDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)

(k) N melliferum (UDF)

Figure 7 Spatial patterns of the five most abundant tree species shown by mark-correlation function with the null model of uncorrelatedmarks Black lines are observed patterns and grey lines are approximate 95 confidence envelopes

tropical forest ecology Three different patterns (regularityrandomness and clumping with a tendency towards regular-ity) were found by Pelissier [46] among large adult trees inthree different plots in tropical India This was also recordedin our results but these patterns tended toward randomnessWe found evidence for competition among trees within thesmaller size classes of the three forest types but this competi-tion was lower than expected given their relatively high den-sities indeed there was only slight regularity or aggregationat a radius of less than 10mClustered spatial distributions areoften typical in naturally regenerated stands [47] In very oldforests at Wind River on the Pacific coast of the USA Northet al [48] found that trees were clustered at all distancesClustering also appeared to be present at shorter distancesalthough this was not seen as differing from a random spatialprocess [49]

42 Interspecific Competition In this study we conducteda comprehensive spatial pattern analysis to assess speciesassociations among the fivemost abundant tree species acrossthree different tropical forest types in Northern VietnamThese species comprised 439 541 and 521 of all treesin the UDF LDF and HDF sites respectively Analogousanalyses of data from tropical forests in Sinharaja (Sri Lanka)

and BarroColorado Island (Panama) also revealed significantsmall-scale associations [27] However we cannot excludethe possibility that these differences may owe as much tovariation in other factors such as environmental heterogene-ity range of conditions or species richness Further com-parisons across sites would help to assess the effects of thesepossibilities Our analyses therefore indicate that there weresome small differences between the spatial structures of theChangbaishan (CBS) and Xuan Nha forests for example wefound that most species cooccurred in small neighborhoodsless often than expected Only 8 of all species pairs sharedroughly the same plot areas in the CBS forests [50] whereasmost XuanNha plot species pairs interacted both at small andlarge scales or else had no association Individuals of differentspecies thus showed a clear tendency towards independenceof one another at all distances

The analyses of spatial association patterns among the fiveabundant species in the three different forest types presentin Xuan Nha Nature Reserve revealed a balance betweenthe attraction and repulsion interaction of spatial structuresThe selective analysis of small-scale effects found surprisingresults especially because more than one-third of all speciespairs had independent association at all scalesWe also foundthat the positive small-scale association patternwas caused by


g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7

(a) E roxburghiana and L ducampii (HDF)

g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(b) E roxburghiana and L ducampii (UDF)

g12(r)

10 20 400 30Scale (m)

0123456789

(c) E roxburghiana and C parthenoxylon (LDF)

g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)

(d) E roxburghiana and C parthenoxylon (HDF)

g12(r)

10 20 30 400Scale (m)

00051015202530354045

(e) E roxburghiana and S cuminii (UDF)

g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)

(f) E roxburghiana and N melliferum (HDF)

g12(r)

0123456789

10 20 30 400Scale (m)

(g) E roxburghiana and N melliferum (UDF)

g12(r)

0123456789

10

10 20 30 400Scale (m)

(h) L ducampii and C parthenoxylon (HDF)

Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)

(i) L ducampii and S cuminii (HDF)

g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

(j) L ducampii and S cuminii (LDF)

g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40

(k) L ducampii and S cuminii (UDF)

g12(r)

00051015202530354045

10 20 30 400Scale (m)

(l) L ducampii and N melliferum (HDF)

Figure 8 Spatial correlations among the five most abundant tree species are shown by the bivariate pair-correlation The observed patterns(dark line) that lie beyond the confidence envelopes (grey lines) indicate significant departures from the null models of toroidal shift (spatialindependence)

the aggregation of neither small individuals nor larger treesSmall-scale attraction rather than competition appeared to bea critical factor in shaping spatial distribution patterns posi-tive associations at small distances can thus be expectedwhenspecies have similar requirements for establishment Like-wise there are positive interspecific neighborhood effects ontree survival (facilitation) andor species are dispersed acrossthe same microhabitats [51]

5 Conclusion

Our analysis based on a systematic design may have beenconstrained by a lack of observation (pseudoreplication) Weused a single one-ha permanent sample plot per forest typeand while this may reflect current conditions it does notfacilitate a detailed statistical analysis Other more efficientmethods are recommended for a better understanding oftree speciesrsquo distributiondensity and disturbance regimesOur results indicate that high intensity human disturbanceadversely affected floristic composition tree species densityand diversity among the three forest types The basal areasrecorded in the three forest types are quite low indicatingthat these forests should be conserved and protected and alltimber cutting and logging activities (legal or illegal) shouldbe avoided until the forest recovers in terms of quantity andquality Enrichment planting for species with low densities in

the disturbed forests should be carried out the promotion ofnatural regeneration by tending weeding and making baresoil in this area will also accelerate the recovery process

The main goals of the present study were to detect theeffects of disturbance degrees on the spatial patterns anddistributions of tree species Through spatial point-patternanalyses our results clearly show the impacts of disturbancedegrees on horizontal structure and intra- and interspecificinteractions The univariate and bivariate data with pair-and mark-correlation functions of intra- and interspecificinteractions within the five most abundant tree species indi-cate that most species demonstrated clumping and regulardistributions at small scales A high proportion of negativeinterspecific small-scale association (inhibition) and positiveassociation (attraction) were recorded in both the LDF andHDF sites but were rare in the UDF Based on the presentresults silvicultural implications such as the site selectionof enrichment planting and the promotion of natural regen-eration could be applied in lightly and highly disturbedforests We suggest that some species characterized by anegative interspecific association for example Lithocarpusducampii and Cinnamomum parthenoxylon should not beplanted in homogeneous environments This competitionand difference in habitat existed in the LDF and HDF siteswhile other habitat associations or dispersal limitationsexisted in theUDFThe results obtained from this study could


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)

(a) E roxburghiana and C parthenoxylon (HDF)

10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)

(b) E roxburghiana and C parthenoxylon (UDF)

10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)

(c) L ducampii and C parthenoxylon (HDF)

02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)

(d) L Ducampii and N melliferum (HDF)

02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)

(e) C parthenoxylon and S cuminii (UDF)

10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)

(f) S Cuminii and N Melliferum (UDF)

Figure 9 Spatial correlations among the fivemost abundant tree species are shown by the bivariate mark-correlation functionsThe observedpatterns (dark line) that lie beyond the confidence envelopes (grey lines) indicate significant departures from the null models of independentmarks

be improved if the series data regarding recruitment andmortality were available in order to explain the underlyingdynamic processes

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are grateful to the Forest Inventory and PlanningInstitute (FIPI) for providing data of the permanent sampleplots They wish to thank all the people who contributed tothis work especiallyMr Dung andMr Huongwho identifiedtree species for them in the field They profoundly thank theforest ranger Mr Phung Huu Thu who worked with them


under difficult conditions in the field Thanks also go toProfessor Dr J Saborowski who provided them with sta-tistical advice at times of critical need They thank MsTinaMarie Joaquimwho corrected and checked grammaticalerrors in the paperThey acknowledge support by theGermanResearch Foundation and theOpenAccess Publication Fundsof the Gottingen University

References

[1] G Kunstler C H Albert B Courbaud et al ldquoEffects of compe-tition on tree radial-growth vary in importance but not in inten-sity along climatic gradientsrdquo Journal of Ecology vol 99 no1 pp 300ndash312 2011

[2] M Kariuki M Rolfe R G B Smith J K Vanclay and R MKooyman ldquoDiameter growth performance varies with speciesfunctional-group andhabitat characteristics in subtropical rain-forestsrdquo Forest Ecology and Management vol 225 no 1ndash3 pp1ndash14 2006

[3] J Lancaster ldquoUsing neutral landscapes to identify patterns ofaggregation across resource pointsrdquo Ecography vol 29 no 3pp 385ndash395 2006

[4] L Parrott and H Lange ldquoUse of interactive forest growth simu-lation to characterise spatial stand structurerdquo Forest Ecology andManagement vol 194 no 1ndash3 pp 29ndash47 2004

[5] F He P Legendre and J V LaFrankie ldquoDistribution patternsof tree species in a Malaysian tropical rain forestrdquo Journal ofVegetation Science vol 8 no 1 pp 105ndash114 1997

[6] J A Logan P White B J Bentz and J A Powell ldquoModelanalysis of spatial patterns in mountain pine beetle outbreaksrdquoTheoretical Population Biology vol 53 no 3 pp 236ndash255 1998

[7] C Skarpe ldquoSpatial patterns and dynamics of woody vegetationin an arid savannardquo Journal of Vegetation Science vol 2 no 4pp 565ndash572 1991

[8] TWiegand and K AMoloney ldquoRings circles and null-modelsfor point pattern analysis in ecologyrdquo Oikos vol 104 no 2 pp209ndash229 2004

[9] S Barot J Gignoux and J-C Menaut ldquoDemography of asavanna palm tree predictions from comprehensive spatialpattern analysesrdquo Ecology vol 80 no 6 pp 1987ndash2005 1999

[10] D N Hamill and S J Wright ldquoTesting the dispersion ofjuveniles relative to adults a new analytic methodrdquo Ecology vol67 no 4 pp 952ndash957 1986

[11] D A King S J Wright and J H Connell ldquoThe contribution ofinterspecific variation in maximum tree height to tropical andtemperate diversityrdquo Journal of Tropical Ecology vol 22 no 1pp 11ndash24 2006

[12] S P Hubbell ldquoTree dispersion abundance and diversity in atropical dry forestrdquo Science vol 203 no 4387 pp 1299ndash13091979

[13] R W Sterner C A Ribic and G E Schatz ldquoTesting for lifehistorical changes in spatial patterns of four tropical treespeciesrdquo Journal of Ecology vol 74 no 3 pp 621ndash633 1986

[14] P Haase ldquoSpatial pattern analysis in ecology based on RipleyrsquosK-function introduction and methods of edge correctionrdquoJournal of Vegetation Science vol 6 no 4 pp 575ndash582 1995

[15] S Vacek and J Leps ldquoSpatial dynamics of forest decline the roleof neighbouring treesrdquo Journal of Vegetation Science vol 7 no6 pp 789ndash798 1996

[16] J H Hou X C Mi C R Liu and K P Ma ldquoSpatial patternsand associations in a Quercus-Betula forest in northern ChinardquoJournal of Vegetation Science vol 15 no 3 pp 407ndash414 2004

[17] T T Veblen D H Ashton and F M Schlegel ldquoTree regener-ation strategies in a lowland Nothofagus-dominated forest insouth-central Chilerdquo Journal of Biogeography vol 6 no 4 pp329ndash340 1979

[18] J F Franklin T A Spies R V Pelt et al ldquoDisturbances andstructural development of natural forest ecosystems with silvi-cultural implications using Douglas-fir forests as an examplerdquoForest Ecology and Management vol 155 no 1ndash3 pp 399ndash4232002

[19] P J Diggle Statistical Analysis of Spatial Point Patterns Aca-demic Press London UK 1983

[20] P J Clark and F C Evans ldquoDistance to nearest neighbor as ameasure of spatial relationships in populationsrdquoEcology vol 35no 4 pp 445ndash453 1954

[21] P Greig-smith ldquoThe use of random and contiguous quadratsin the study of the structure of plant communitiesrdquo Annals ofBotany vol 16 no 2 pp 293ndash316 1952

[22] E C Pielou ldquoThe use of point-to-plant distances in the study ofthe pattern of plant populationsrdquo Journal of Ecology vol 47 no3 pp 607ndash613 1959

[23] B D Ripley ldquoTests of lsquorandomnessrsquo for spatial point patternsrdquoJournal of the Royal Statistical Society Series B Methodologicalvol 41 no 3 pp 368ndash374 1979

[24] C Hui L C Foxcroft D M Richardson and S MacFadyenldquoDefining optimal sampling effort for large-scale monitoring ofinvasive alien plants a Bayesian method for estimating abun-dance and distributionrdquo Journal of Applied Ecology vol 48 no3 pp 768ndash776 2011

[25] A Pommerening ldquoApproaches to quantifying forest structuresrdquoForestry vol 75 no 3 pp 305ndash324 2002

[26] D Stoyan and H Stoyan Fractals Random Shapes and PointFields Methods of Geometrical Statistics Wiley Chichester UK1994

[27] T Wiegand S Gunatilleke and N Gunatilleke ldquoSpecies asso-ciations in a heterogeneous Sri Lankan dipterocarp forestrdquoTheAmerican Naturalist vol 170 no 4 pp 77ndash95 2007

[28] S P Hubbell ldquoNeutral theory and the evolution of ecologicalequivalencerdquo Ecology vol 87 no 6 pp 1387ndash1398 2006

[29] S Getzin andKWiegand ldquoAsymmetric tree growth at the standlevel random crown patterns and the response to sloperdquo ForestEcology and Management vol 242 no 2-3 pp 165ndash174 2007

[30] J Illian A Penttinen H Stoyan and D Stoyan StatisticalAnalysis and Modelling of Spatial Point Patterns John Wiley ampSons 2008

[31] P Greig-Smith ldquoPattern in vegetationrdquo The Journal of Ecologyvol 67 no 3 pp 755ndash779 1979

[32] J Leps ldquoCan underlying mechanisms be deduced fromobserved patternsrdquo in Spatial Processes in Plant CommunitiesF Krahulec A D Q Agnew and J H Willems Eds pp 1ndash11SPB Academic Publishing The Hague The Netherlands 1990

[33] A S Watt ldquoPattern and process in the plant communityrdquo TheJournal of Ecology vol 35 no 12 pp 1ndash22 1947

[34] A W Tordoff B Q Tran T D Nguyen and H M LeSourcebook of Existing and Proposed Protected Areas in VietnamBirdlife International in Indochina andMinistry of Agricultureand Rural Development Hanoi Vietnam 2004

[35] Plan of conservation andsustainable development Xuan NhaNature Reserve until 2020 Son La Protection of Department2013


[36] N H Hai K Wiegand and S Getzin ldquoSpatial distributions oftropical tree species in northern Vietnam under environmen-tally variable site conditionsrdquo Journal of Forestry Research vol25 no 2 pp 257ndash268 2014

[37] B Adrian and T Rolf ldquoSpatstat an R package for analyzingspatial point patternsrdquo Journal of Statistical Software vol 12 no6 pp 1ndash42 2005

[38] M J Lawes M E Griffiths J J Midgley S Boudreau H A CEeley and C A Chapman ldquoTree spacing and area of compet-itive influence do not scale with tree size in an African rainforestrdquo Journal of Vegetation Science vol 19 no 5 pp 729ndash7382008

[39] S Getzin T Wiegand K Wiegand and F He ldquoHeterogeneityinfluences spatial patterns and demographics in forest standsrdquoJournal of Ecology vol 96 no 4 pp 807ndash820 2008

[40] F Goreaud and R Pelissier ldquoAvoiding misinterpretation ofbiotic interactions with the intertype K 12-function populationindependence vs random labelling hypothesesrdquo Journal ofVegetation Science vol 14 no 5 pp 681ndash692 2003

[41] S Getzin C Dean F He J A Trofymow K Wiegand and TWiegand ldquoSpatial patterns and competition of tree species in aDouglas-fir chronosequence on Vancouver Islandrdquo Ecographyvol 29 no 5 pp 671ndash682 2006

[42] G Shen M Yu X-S Hu et al ldquoSpecies-area relationshipsexplained by the joint effects of dispersal limitation and habitatheterogeneityrdquo Ecology vol 90 no 11 pp 3033ndash3041 2009

[43] R Condit P S Ashton P Baker et al ldquoSpatial patterns in thedistribution of tropical tree speciesrdquo Science vol 288 no 5470pp 1414ndash1418 2000

[44] Z Luo B Ding X Mi J Yu and Y Wu ldquoDistribution patternsof tree species in an evergreen broadleaved forest in easternChinardquo Frontiers of Biology in China vol 4 no 4 pp 531ndash5382009

[45] H A Peters ldquoNeighbour-regulated mortality the influence ofpositive and negative density dependence on tree populationsin species-rich tropical forestsrdquo Ecology Letters vol 6 no 8 pp757ndash765 2003

[46] R Pelissier ldquoTree spatial patterns in three contrasting plots ofa southern Indian tropical moist evergreen forestrdquo Journal ofTropical Ecology vol 14 no 1 pp 1ndash16 1998

[47] A Van Laar and A Akca Forest Mensuration Managing ForestEcosystems Springer Science amp Business Media 1997

[48] M North J Chen B Oakley et al ldquoForest stand structure andpattern of old-growthwestern hemlockDouglas-fir andmixed-conifer forestsrdquo Forest Science vol 50 no 3 pp 299ndash311 2004

[49] R P Duncan and G H Stewart ldquoThe temporal and spatialanalysis of tree age distributionsrdquo Canadian Journal of ForestResearch vol 21 no 12 pp 1703ndash1710 1991

[50] X Wang T Wiegand Z Hao B Li J Ye and F Lin ldquoSpeciesassociations in an old-growth temperate forest in north-easternChinardquo Journal of Ecology vol 98 no 3 pp 674ndash686 2010

[51] I Martınez T Wiegand F Gonzalez-Taboada and J R ObesoldquoSpatial associations among tree species in a temperate forestcommunity in North-western Spainrdquo Forest Ecology and Man-agement vol 260 no 4 pp 456ndash465 2010

Submit your manuscripts athttpwwwhindawicom

Forestry ResearchInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Environmental and Public Health

Journal of



EcosystemsJournal of


MeteorologyAdvances in

EcologyInternational Journal of


Marine BiologyJournal of


Hindawi Publishing Corporationhttpwwwhindawicom

Applied ampEnvironmentalSoil Science

Volume 2014

Advances in


Environmental Chemistry

Atmospheric SciencesInternational Journal of



Waste ManagementJournal of

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal of

Geophysics


Geological ResearchJournal of

EarthquakesJournal of


BiodiversityInternational Journal of


ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

OceanographyInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Journal of Computational Environmental SciencesHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


ClimatologyJournal of


interspecific competition and environmental heterogeneityThe spatial structure of a forest is largely determined by therelationships within neighboring groups of trees [24] Themain factors in a forest structure are the spatial distributionspecies diversity and variations in tree dimensions [25] Inrecent decades several structural indices such as the Clarkand Evans aggregation index [20] pair-correlation function[26] and species diversity indices [13] have been developedto quantify spatial forest structure

Tropical rain forests have been used in recent studies todemonstrate the effects of high diversity on spatial distribu-tion [5 27]This explains why interspecific species have func-tional similarity and may adapt to average environment con-ditions [28] One approach to exploring spatial plant dynam-ics is to use a point pattern analysis of fully mapped plantlocations [29] Spatial statistics like Ripleyrsquos 119870 function [23]and the pair-correlation function [26 30] quantify the small-scale spatial correlation structure of a pattern which containsinformation on positivenegative interactions among plantsIn addition point pattern analysis is ideally suited to controlfor environmental heterogeneity and focus on neighborhoodprocesses Spatial patterns of individuals within populationsare closely linked to ecological processes consequently eco-logical processes may be deduced from spatial patterns [31ndash33]

There is little published information on the spatial pat-terns of tree species in tropical evergreen forests as affectedby different disturbance regimes in Vietnam In the presentstudy univariate 119892(119903) statistics with the null model of com-plete spatial randomness and the mark-correlation functionwere used for point pattern analysis at different scales Weanalyzed the intraspecific and interspecific association (rela-tionship or correlation) of five abundant tree species in threeforest types in a tropical evergreen forest in Xuan Nha NatureReserve VietnamThis study will providemeaningful knowl-edge for predicting the spatial patterns of the most dominanttree species after disturbance and could have further silvi-cultural implications for management practice (eg selectingnative tree species and choosing site conditions for enrich-ment planting) in restoration zones of the nature reserve

2 Materials and Methods

21 Study Site The study site (Figure 1) is located in XuanNha Nature Reserve (20∘361015840ndash20∘481015840N 104∘291015840ndash104∘501015840E)The reserve has a total area of 27100 ha of which 15300 aretropical evergreen forests and 2600 are tropical forests onlimestone [34]Thenature reserversquos elevation ranges from 300to 2000m above sea level The topography is composed oftwo shallow-sided valleys which run across the reserve fromthe Laotian border Xuan Nha Nature Reserve is located ina tropical monsoon climate where the mean temperature is213∘C The rainy season lasts from May to October (witha maximum of 3324mm of rain in August) while the dryseason goes from November to April (with a minimum of324mm in February) Mean annual precipitation is about1673mm Six major soils are found in the study site (1) thicklayers of dark yellow ferralsols at 700ndash1700m in altitude thatdevelop on shale or metamorphic rocks with a medium

QuynhNhai

Thuan Chau

Muong La

Bac YenPhu Yen

East Sea

Moc Chau

Yen Chau

Mai SonSong Ma

Sop Cop

Son La City

Xuan Nhanature reserve

Cambodia

LAOS

N China

Figure 1 The location of Xuan Nha Nature Reserve in VietnamDark blue yellow and green colors indicate the undisturbed forest(UDF) lightly disturbed forest (LDF) and highly disturbed forest(HDF) respectively

texture (2) thin layers of yellow-brown ferralsols at 700ndash1700m altitude which develop on limestone ormetamorphiclimestone and have coarse tomedium texture (3) thick layersof light yellow ferralsols at 700ndash1700m on shale ormetamor-phic rockswithmedium texture (4) thick layers of lightgrey-yellow ferralsols that develop at 300 to 1000m in low hilly ormountainous regions on slate siltstone sandstone and con-glomerate with a coarse to medium texture (5) grey-yellowferralsols modified by paddy fields with medium texture thatappear in the surrounding villages and (6) alluvial soilsdeposited at the foot of mountains river banks and streams[35]

In accordance with the collected data approximately 43of the forest area (app 787220 hectares) is disturbed while428 (78214 ha) is classified as undisturbed (Table 1) Thisdisturbed area is divided into two categories lightly disturbedand heavily disturbed Undisturbed and disturbed forestsrefer to different disturbance regimes where ldquodisturbancerdquo isdefined as the impact level of human activities in the forestDisturbance in this context refers to selective timber harvest-ing felling small-sized trees for nontimber products and past(pre-2003) illegal logging

Undisturbed forests (primary or rich forests) are areasthat do not show evidence of damage from human activitiesthese are relatively stable forests not yet (or less) influenced byhumans or natural disasters Because such forests are limitedto inaccessible and protected areas they are not used for pro-duction purposes UDFs are extremely rich forests with a tim-ber volume of standing trees between 201 and 300m3 haminus1LDFs (average forest) are influenced by humans or naturaldisasters leading to changes in their structure This typedemonstrates low and minor damage from human beingsand has a timber reserve of standing trees between 101 and200m3 haminus1 Forests may be classified as LDFs if past logging








I II

1 2 3 4 50

6 7 8 9 10

0

11 12 13 14 15

16 17 18 19 20

21 22 23 24 25

IV III

N1

N2

y1x1

y2

x2

20m20m

20m

20m



sampled








11(119903) = 120582












12(119903) = 120582






119898119898(119903) [26 39] using


1 1198982) =

1198981times 1198982 where 119898

1and 119898





119898119898(119903) = 1 119870

119898119898(119903) gt 1 or 119870

119898119898(119903) lt 1

respectively



3 Results















X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(a)

X (m)

Types



20

30

40

50

60

70

80

90

100

110

120

130

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

(b)

X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(c)



5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)






4 Discussion



g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics





















I II

1 2 3 4 50

6 7 8 9 10

0

11 12 13 14 15

16 17 18 19 20

21 22 23 24 25

IV III

N1

N2

y1x1

y2

x2

20m20m

20m

20m



sampled








11(119903) = 120582












12(119903) = 120582






119898119898(119903) [26 39] using


1 1198982) =

1198981times 1198982 where 119898

1and 119898





119898119898(119903) = 1 119870

119898119898(119903) gt 1 or 119870

119898119898(119903) lt 1

respectively



3 Results















X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(a)

X (m)

Types



20

30

40

50

60

70

80

90

100

110

120

130

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

(b)

X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(c)



5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)






4 Discussion



g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics



















11(119903) = 120582












12(119903) = 120582






119898119898(119903) [26 39] using


1 1198982) =

1198981times 1198982 where 119898

1and 119898





119898119898(119903) = 1 119870

119898119898(119903) gt 1 or 119870

119898119898(119903) lt 1

respectively



3 Results















X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(a)

X (m)

Types



20

30

40

50

60

70

80

90

100

110

120

130

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

(b)

X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(c)



5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)






4 Discussion



g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics

























X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(a)

X (m)

Types



20

30

40

50

60

70

80

90

100

110

120

130

0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)

(b)

X (m)

Types


0 20 40 60 80 100

0

20

40

60

80

100

Y(m

)


20

30

40

50

60

70

80

90

100

110

120

130

(c)



5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)






4 Discussion



g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85

All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




5 10 15 20 25 30 35 40 45DBH class (cm)

50 55 60 65 70 75 80 85



All species

0

10

20

30

40

50

60

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85




All species

0

10

20

30

40

50

60

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

050

100150200250

Num

ber o

f tre

es p

er h

ecta

re (N

haminus

1)

5 15 25 35 45 55 65 75 85

DBH class (cm)






4 Discussion



g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

00051015202530354045


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


0

1

2

3

4

5

6

7

8

9

g11(r)

10 20 30 400Scale (m)

(e) S cuminii (HDF)

g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)







10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















10 20 30 400Scale (m)

0

1

2

3

4

5

6

g11(r)


g11(r)

0

1

2

3

4

5

6

7

10 20 30 400Scale (m)


g11(r)

0

2

4

6

8

10

12

10 20 30 400Scale (m)


g11(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12

(e) S cuminii (LDF)

g11(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)

(f) S cuminii (UDF)

g11(r)

10 20 30 400Scale (m)

0123456789


g11(r)

10 20 30 400Scale (m)

0

2

4

6

8

10

12




10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

10

20

30

40

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

04

06

08

10

12

14

16

18

20

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

Km1(r)

(h) S cuminii (HDF)

Figure 7 Continued


10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















10 20 30 40 500Scale (m)

0002040608101214161820

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

35

Km1(r)








g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















g12(r)

10 20 30 400Scale (m)

0

1

2

3

4

5

6

7


g12(r)

0

1

2

3

4

5

6

10 20 30 400Scale (m)


g12(r)

10 20 400 30Scale (m)

0123456789


g12(r)

0

1

2

3

4

5

6

7

8

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00051015202530354045


g12(r)

00

05

10

15

20

25

30

35

40

10 20 30 400Scale (m)


g12(r)

0123456789

10 20 30 400Scale (m)


g12(r)

0123456789

10

10 20 30 400Scale (m)


Figure 8 Continued


g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















g12(r)

00051015202530354045

10 20 30 400Scale (m)


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35


g12(r)

10 20 30 400Scale (m)

00

05

10

15

20

25

30

35

40


g12(r)

00051015202530354045

10 20 30 400Scale (m)




5 Conclusion





10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics















10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

18

Km2(r)


10 20 30 40 500Scale (m)

00

05

10

15

20

25

30

Km2(r)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)


02

04

06

08

10

12

14

16

Km2(r)

10 20 30 40 500Scale (m)


02

04

06

08

10

12

14

Km2(r)

10 20 30 40 500Scale (m)


10 20 30 40 500Scale (m)

02

04

06

08

10

12

14

16

Km2(r)






Acknowledgments




References

























































Journal of












Volume 2014

Advances in









Geophysics
















References

























































Journal of












Volume 2014

Advances in









Geophysics



































Journal of












Volume 2014

Advances in









Geophysics














research article effect of disturbance regimes on spatial ...research article effect of disturbance...

Documents