Development of Allometric Equations for Estimating Above ...downloads.hindawi.com/journals/ijeco/2013/658140.pdf2 InternationalJournalofEcology structure,diversity,andproductivity[13,14].Consequently,

Hindawi Publishing CorporationInternational Journal of EcologyVolume 2013 Article ID 658140 8 pageshttpdxdoiorg1011552013658140

Research ArticleDevelopment of Allometric Equations for EstimatingAbove-Ground Liana Biomass in Tropical Primary andSecondary Forests Malaysia

Patrick Addo-Fordjour12 and Zakaria B Rahmad1

1 School of Biological Sciences University of Science Malaysia 11800 Pulau Penang Penang Malaysia2 Department ofTheoretical and Applied Biology College of Science Kwame Nkrumah University of Science and Technology (KNUST)Kumasi Ghana

Correspondence should be addressed to Patrick Addo-Fordjour paddykay77yahoocom

Received 26 February 2013 Accepted 4 June 2013

Academic Editor Shibu Jose

Copyright copy 2013 P Addo-Fordjour and Z B Rahmad This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

The study developed allometric equations for estimating liana stem and total above-ground biomass in primary and secondaryforests in the Penang National Park Penang Malaysia Using biomass-diameter-length data of 60 liana individuals representing15 species allometric equations were developed for liana stem biomass and total above-ground biomass (TAGB) Three types ofallometric equations were developed models fitted to untransformed weighted and log-transformed (log

10) data There was a

significant linear relationship between biomass and the predictors (diameter length andor their combinations) The same setof models was developed for primary and secondary forests due to absence of differences in regression line slopes of the forests(ANCOVA119875 gt 005)The coefficients of determination values of themodels were high (stem 0861 to 0990 TAGB 0900 to 0992)Generally log-transformed models showed better fit (Furnivalrsquos index FI lt 050) than the other models (FI gt 05) A comparisonof the best TAGB model in this study (based on FI) with previously published equations indicated that most of the equationssignificantly (119875 lt 005) overestimated TAGB of lianas However a previous equation from Southeast Asia estimated TAGB similarto that of the current equation (119875 gt 005)Therefore regional or intracontinental equations should be preferred to intercontinentalequations when estimating liana biomass

1 Introduction

Lianas have great influence on forest ecosystems especiallyin tropical forests [1] They contribute very much to speciesdiversity in the tropics constituting as high as 38 of speciesdiversity [2]Theymay compose of amuch higher percentage(45) with regard to total woody plant stems in the tropics(cf [3]) Lianas serve as an important source of food for forestfauna especially in the dry season [4] They may provideup to about one-third of canopy foliage in the forest (cf[5]) and therefore contribute substantially (up to 36) tototal above-ground leaf biomass in tropical forest ecosystems[6] Lianas compete with trees which may affect tree growth[7 8] Additionally they may have negative influence on seed

production of trees and also impede natural regeneration oftrees [7ndash9]

Tropical forest ecosystems continue to be exploited atalarming rates resulting in their conversion to secondaryforests and many other forms of land use [10] A highproportion of forests in the tropics is made up of secondaryforests [11] In Malaysia many lowland dipterocarp forestshave been converted to secondary forests as a result oflogging farming and other human activities (cf [12]) Withchanges in tropical forest ecosystems as a result of humandisturbance the ability of secondary forests to accumulatebiomassmay be altered Even within primary forests changesin biomass stock may occur from time to time due toother factors which elicit changes in forest composition

2 International Journal of Ecology

structure diversity and productivity [13 14] Consequentlythere is the need to assess biomass levels of primary andsecondary forests from time to time Knowledge of theamount of biomass stored up within forests is importantfor a couple of reasons Firstly biomass data is fundamentalin evaluating the productivity structure and conditions offorests [15] Secondly biomass measurements are useful inestimating the amount of carbon sequestered by forests [16]Furthermore it plays an important role in elucidating theeffects of deforestation and carbon sequestration on globalcarbon balance [16] Although lianas continue to increase inbiomass in tropical forests [17] they have not been factored inmost forest biomass assessments In recent times there havebeen growing calls for more biomass assessment of tropicalforests because of the need to estimate forest carbon stocksand the potential amount of biomass available as a source ofenergy [18] Unfortunately all the calls were made in relationto only tree biomass with no mention of lianas This is quiteunfortunate in view of the fact that lianas can add up to 30of total above-ground biomass in tropical forests with denseliana population (cf [17])

Though there are several ways of determining biomass ofplants in forests [18] two main approaches are commonlyused These are destructive harvesting of plants and thendetermining their biomass and the use of allometric equa-tions to predict biomass of plants However it is preferablethat allometric equations be used to estimate forest biomassas they allow for the estimation of large forest areas and avoiddestruction of forest There are many allometric equationsthat have been developed for the estimation of plant biomassin many tropical forests Nevertheless most of the allometricequations are for trees with only a few on lianasThe scarcityof liana allometric equations is partly due to the difficulty inaccessing the whole length of lianas from trees Currently itis known that lianas can contribute up to 30 of total above-ground biomass in liana-dense tropical forests (cf [17]) Sincelianas are disturbance adapted group of plants [19] they arelikely to increase in abundance in the face of increasingdisturbance in tropical forests and therefore their biomasscould even be higherWith this development the importanceof lianas in the forest would continue to increase withtime and therefore ignoring them in biomass assessmentof forests could lead to underestimation of forest biomass[20] Accordingly lianas should be integrated into tropicalforest biomass assessments so as to obtain the true amountof tropical forest biomass To this end more allometricequations of lianas must be developed for tropical forestecosystems tomake biomass determination of lianas possibleand an integral part of forest biomass assessment

Even though a few liana allometric equations have beendeveloped for the estimation of above-ground biomass intropical forests [3 20 21] there is no allometric equationfor lianas in Malaysia To date only one allometric equationhas been developed for the estimation of liana biomassin the whole of Southeast Asia [22] To know the relativecontribution of above-ground plant parts to total above-ground biomass plant part allometric equations must bedeveloped Most liana allometric equations in use onlymeasure total above-ground biomass without the individual

6∘U

525

kilometers

0

66∘UUUUU

Study site

5∘20

998400U

110∘10

998400T 110∘20

998400T

International boundary

Pulau

N

Penang

Thailand

Sumatera

Straitsof

Malacca

SouthChina

SeaPeninsularMalaysia

100∘T 101

∘T 102∘T 103

∘T 104∘T

5∘U

4∘U

3∘U

2∘U

1∘U

Figure 1Map of Penang State showing the location of the study area(Penang National Park)

components Since above-ground portion of lianas have twomain components (stem and leaf) developing allometricequations for at least one of them should make it possibleto determine the biomass of the other component whenit is used in conjunction with total above-ground biomassequations Therefore the study was aimed at developingallometric equations for the estimation of stem and totalabove-ground biomass of lianas in primary and secondaryforests within the Penang National Park Penang Malaysia

2 Materials and Methods

21 Study Area The study was conducted in the PenangNational Park Penang Malaysia (N 5∘ 275831015840 E 100∘ 123501015840)(Figure 1) which has a total area of 1213 haThe national parkis comprised of primary and secondary forest typesThe parkcontains many plant species (1000 species) the majority ofwhich are from Dipterocarpaceae Fabaceae ApocynaceaeAnacardiaceae Euphorbiaceae and Moraceae A total of 190species of animals have been recorded and these include25 mammal species 53 butterfly species 46 bird speciesand other reptiles insects and amphibians [23] The PenangNational Park was designated as a national park in April 42003 and gazetted under National Park Act 226 of 1980 onApril 10 2003

The secondary forest within the Penang National Parkhas experienced various forms of human disturbance sincethe early years of the twentieth century Logging was a majorform of disturbance in the secondary forest although humanactivities such as farming hunting and gathering occurredin it The secondary forest possesses relics of considerablehistorical land use activities such as clearcutting farminglogging and bush burning due to its close proximity tothe community The secondary forest is also characterisedby some ancient logging routes confirming its extensive use

International Journal of Ecology 3

in the distant past The provision of infrastructure in thesecondary forest of the national park has also contributedto loss of forest cover and resulted in fragmentation insome parts of the forest Human disturbance has significantlyinfluenced liana diversity and structure in the secondaryforest [24]

22 Field Sampling and Biomass Measurements Differentliana species were destructively sampled for themixed speciesallometric equations from April to June 2012 Field samplingoccurred in the rainy season during which leaf biomass wasmaximum [21] To ensure that lianas sampled representedvariation in microhabitats within the national park lianaswere sampled from slopes valleys and flatlands in eachforest type Sampling was conducted in such a way thatthere was even species representation from each forest typeas much as possible Lianas were harvested for biomassdetermination after measuring their diameter (at 13m fromthe rooting base) In addition the length of each harvestedliana individual was measured A total number of 60 lianaindividuals (30 each from the primary and secondary forests)were harvested (Table 1) The characteristics of the lianasharvested are indicated in Table 2 Lianas with clonal stemswere rare in the study area andwere not included in the studyLiana leaves were separated from stems and both were sun-dried for 3 weeks and 2 months respectively for dry weightdetermination

23 Data Analyses Data exploration and preliminary modelfitting were carried out to get models that best fit the dataVarious models were initially developed using the originaluntransformed data but only those that met regressionassumptions (homogeneity of variance linearity normalityand nonautocorrelation) and had high goodness of fit wereretained Subsequently a series of data transformation wascarried out to find out whether better models could bedeveloped from the transformation of data Logarithmictransformation (log

10) and weighting of the independent

variables (diameter and length) and the dependent variable(biomass) were found to be the best transformations whichproduced models that did not violate regression assumptionsand at the same time had high goodness of fit Homogeneityof variance and linearity of data were checked from residualplots whereas autocorrelation and normality were checkedwith Durbin-Watson statistics and probability plot respec-tively

Though logarithmic transformation is reported toincrease the statistical validity of regression analysis byhomogenising variance it introduces a slight downward biaswhen data are back-transformed to arithmetic units [25]To account for the bias the back-transformed results fromlogarithmic unit are usually multiplied by a correction factorConsequently a correction factor (CF) was calculated forall logarithmic equations The CF is given by the followingequation [26]

CF = exp((SEElowast2303)22) (1)

where SEE is the standard error of the estimate

Table 1 Number of individuals of liana species in the primary andsecondary forests used for the allometric equations

Species Number of individualsPrimary forest Secondary forest Total

Agelaea macrophylla 2 2 4

Artabotrys oblongus 2 4 6

Cnestis palala 2 0 2

Coptosapelta parviflora 2 2 4

Cyathostemma hookeri 3 2 5

Dalbergia rostrata 1 4 5

Gnetum latifolium 3 2 5

Rourea rugosa 1 1 2

Salacia sp 1 2 3

Spatholobus ferrugineus 2 2 4

Strychnos curtisii 3 0 3

Strychnos ignatii 2 1 3

Tetracera macrophylla 2 4 6

Willughbeia sp 2 2 4

Willughbeia angustifolia 2 2 4

Table 2 Summary of allometric properties of liana individuals usedin the study

Parameter Minimum Maximum MeanDiameter (cm) 130 1420 539

Length (m) 400 2620 940

Stem biomass (kg) 261 2310 926

Total above-ground biomass (kg) 303 2670 1078

Due to the low abundance of lianas in the forests onlymixed allometric equations were developed Equations weredeveloped using liana diameter [diameter]2 length or acombination of them as estimate parameters

Comparison of models was done using coefficient ofdetermination (1198772) and Furnivalrsquos index (FI) Models withthe same response (dependent) variable were compared using1198772 Formodels with different dependent variables using1198772 to

compare them could produce false results [27] ConsequentlyFurnivalrsquos index [28] which is able to compare models ofdifferent dependent variables or weights was used to comparethe logarithmic and nonlogarithmic models The index iscalculated as follows

FI = 1[1198911015840 (119884)]

radicMSE (2)

where 1198911015840(119884) is the derivative of the dependent variable withrespect to biomass MSE is the mean square error of the fittedequation and the square bracket ([ ]) is the geometric mean

Regression analyses were conducted to determine therelationships between liana biomass and the various inde-pendent variables (diameter length and their combinations)for both untransformed and transformed data Analysisof covariance (ANCOVA) was carried out to examine thepossible effect of forest type on the various regressionmodels


Table 3 Six previously published allometric equations (total above-ground liana biomass) used in comparing the current allometricequationAll the equations are based on themodel form total above-ground biomass = exp [119888 + 120572 ln (diameter)]

Equation c 120572

Gehring et al [21] minus1547 2640

Gerwing and Farias [20] 0147 2184

Putz [29] 0036 1806

Hozumi et allowast [22] minus1347 2391

Beekman [30] minus1459 2566

Schnitzer et al [3] minus1484 2657

lowastBased on a data set collected from Cambodia Southeast Asia

Forest type was used as the main factor whereas the variousindependent variables were used as covariables

A current allometric equation for total above-groundbiomass which was deemed as the best model based onFurnivalrsquos index of fit was compared with six previouslypublished total above-ground biomass allometric equationsof lianas (Table 3) Five of the equations were constructedfrom diameter-biomass data of lianas taken from individualstudies Three of the studies namely Gerwing and Farias[20] Gehring et al [21] and Putz [29] contained publishedallometric biomass regression equations but new versionsof these equations were developed from their data sets bySchnitzer et al [3] Two of the studies [22 30] on the otherhand did not have allometric equations but contained thediameter-biomass data which were used to construct allo-metric equations by Schnitzer et al [3] The sixth allometricequation [3] was constructed from combined data sets of allthe above five studies (424 liana individuals) The equationswere applied to a data set taken from the same forest (PenangNational Park) The data was taken from 78 liana individualscomprising 27 species (diameter ranged from 14 to 14 cm)Paired t-test was run to determine significant differencesamong total above-ground biomass values determined byeach of the six previous equations and the current equation

Regression analyses and t-test were conducted with theGenStat software (VSN International Ltd Hemel Hemp-stead UK) whereas ANCOVA was run with XLSTAT 20122version (Addinsoft SARL Paris France) All analyses wereconducted at a significance level of 5

3 Results and Discussion

A total of seven different models each were developedfor stem and total above-ground biomass of liana species(Tables 4 and 5) These could be categorised into threetypes of models models fitted on the original arithmeticscale (models 1ndash3 and 8ndash10) weighted models fitted on theoriginal arithmetic scale (models 4-5 and 11-12) and modelsfitted on logarithmic transformed data (models 6-7 and 13-14) There was a significant linear relationship between theindependent variables (diameter length [diameter]2 andortheir combinations) and the dependent variables (stem andtotal above-ground biomass) in the allometric equations(Figures 2 3 4 5 and 6)

14121086420

25

20

15

10

5

0

Stem

bio

mas

s (kg

)

Diameter (cm)

Figure 2 Liana allometric relationship between stem diameter andstem biomass of lianas

30

TAG

B (k

g)

14121086420Diameter (cm)

25

20

15

10

5

0

Figure 3 Liana allometric relationship between stem diameter andtotal above-ground biomass (TAGB) of lianas

Generally the models developed in this study exhibitedhigh coefficients of determination (1198772 ge 86) Diameteris the commonest predictor in most biomass allometricequations [31] although length is also used in some studies(eg [12 32]) There were relatively higher coefficients ofdetermination for the nontransformed allometric equationsthat used diameter as liana biomass predictor (1198772 = 989and 992 for stem and total biomass equations resp) thanthose which used liana length as the predictor of lianabiomass (1198772 = 896 and 90) This indicates that lianadiameter was a better predictor of liana biomass than lianalength Addition of liana length as a second estimator to lianaallometric equations was found to improve upon goodnessof fit of allometric equations of lianas in a study conductedin the Amazonian forest [21] However a different patternwas observed in this study whereby there was no significantcontribution of using length as a second estimator of bothstem and total biomass models (119875 = 0066 and 0057 resp)Gehring et al [21] used a much higher sample size (119899 = 439individuals) which could have affected the influence of lianalength on the allometric equations developed in that studyAlthough the weighted form of liana diameter (diameter2)was a significant estimator of liana biomass it yielded a


Table 4 Allometric equations of mixed species for estimating liana stem biomass (kg)

Equation c(plusmnSE) 120572(plusmnSE) 120573(plusmnSE) 1198772 (adjusted) FI

1 Stem biomass = 119888 + 120572D 0154 plusmn 0078 1731 plusmn 0028 mdash 0989 058

2 Stem biomass = 119888 + 120572L 255 plusmn 0189 0416 plusmn 0010 mdash 0896 170

3 Stem biomass = 119888 + 120572D + 120573L 0474 plusmn 0258 0452 plusmn 0037 0394 plusmn 0010 0976 084

4 (Stem biomass)09 = 119888 + 120572D 0782 plusmn 0098 1210 plusmn 0015 mdash 0990 053

5 Stem biomass = 119888 + 1205721198632 0144 plusmn 0089 0349 plusmn 0005 mdash 0861 263

6 Log10 (stem biomass) = 119888 + 120572(log10 D) 0396 plusmn 00033 1086 plusmn 0042 mdash 0981 018

7 Log10 (stem biomass) = 119888 + 120572(log10 D) + 120573(log10 L) 0165 plusmn 0002 0432 plusmn 0063 0431 plusmn 0062 0954 047

equation number D liana diameter L liana length

Table 5 Allometric equations of mixed species for estimating total above-ground biomass (kg) of lianas


8 Total biomass = c + 120572D 0262 plusmn 0181 1934 plusmn 0029 mdash 0992 058

9 Total biomass = c + 120572L 3120 plusmn 0413 0476 plusmn 0021 mdash 0900 192

10 Total biomass = c + 120572D + 120573L 0768 plusmn 0280 0511 plusmn 0040 0450 plusmn 0010 0987 089

11 (Total biomass)09 = c + 120572D 1041 plusmn 0096 1354 plusmn 0015 mdash 0993 053

12 Total biomass = c + 1205721198632 0194 plusmn 0041 0405 plusmn 0007 mdash 0904 241

13 Log10(total biomass) = c + 120572(log10 D) 0490 plusmn 0021 1090 plusmn 0027 mdash 0986 022

14 Log10 (total biomass) = c + 120572(log10 D) + 120573(log10 L) 0275 plusmn 0021 0470 plusmn 0066 0452 plusmn 0044 0960 049


121080604020

15

125

1

075

05

log10(diameter in cm)

log 1

0(T

AGB

in k

g)

Figure 4 Liana allometric relationship between log10stemdiameter

and log10total above-ground biomass (TAGB) of lianas

much lower coefficient of determination (models 5 and 12)compared to its unweighted diameter equations (models 1and 8) Addition of liana length as a second predictor inthe logarithmic models (models 7 and 14) did not improvegoodness of fit significantly (119875 = 0318 and 0423 for stem andtotal biomass resp) just as observed in the raw data models

Based on Furnivalrsquos index (FI) of fit the logarithmictransformed allometric equations performed better than theother models in both stem and total above-ground modelsWhereas the FI values for all the logarithmic transformedmodels were below 05 those of the other models were above05 Considering the fact that measuring liana length on thefield is impossible without harvesting them it is practicallynot feasible to use allometric equations incorporating liana

20

15

10

5

014121086420

Diameter (cm)

(Ste

m b

iom

ass i

n kg

)09

Figure 5 Liana allometric relationship between stem diameter and(stem biomass in kg)09 of lianas

length Additionally the inclusion of length may introducepropagated variance and recirculation of estimation errorsat each subsequent prediction period (cf [33]) For thesereasons the allometric equations which used only lianadiameter as the predictor are recommended for use in lianabiomass determination Out of the equations that used onlydiameter as the independent variable models 6 and 13 couldbe selected as the best models for determining liana stemand total above-ground biomass as they had the lowest FIvalues (018 and 022 resp) in the study The overall supe-riority of the logarithmic transformed models to the othermodels lends support to the finding that log transformationprovides better fit than untransformed data in most cases[33]



20

15

10

5

0

(TAG

B in

kg)

09

Figure 6 Liana allometric relationship between stem diameter and(TAGB in kg)09 of lianas (TAGB total above-ground biomass)

Table 6 Correction factor (CF) of logarithmic models for stem andtotal above-ground biomass of lianas

Model CFStem Total

Log10 (biomass) = c + 120572(log10 D) 1005 1002

Log10 (biomass) = c + 120572(log10 D) + 120573(log10 L) 1011 1011

Due to the systematic bias introduced by log transforma-tion of data correction factor is normally calculated for log-arithmic transformed allometric equations so as to accountfor the bias The correction factor values determined for thevarious log-transformed allometric equations (Table 6) weregenerally low (02ndash11) indicating that the downward bias ofthe equations was marginal

There was no significant effect of forest type on theallometric equations developed in the study (ANCOVA 119875 gt005) indicating that the equations could be used in bothprimary and secondary forests This finding is supported byGehring et al [21] who did not observe significant change inthe goodness of fit of liana allometric equations when forestage was factored in

Comparison of the best total above-ground biomassmodel with previously published equations (Table 3) indi-cated that the current equation differed considerably fromthem All the previous allometric equations overestimatedtotal above-ground biomass when they were applied to adata set taken from the same forest (Penang National ParkMalaysia) (Figure 7) The overestimation of total above-ground biomass of lianas ranged from 29 to 745Themeantotal above-ground biomass per liana species estimated byfive of the previous equations differed significantly (Table 7119875 lt 005) from that of the current study Interestinglythe allometric equation developed from a data set fromSoutheast Asia [3 22] provided the lowest overestimationand its mean total above-ground biomass did not differsignificantly (119875 = 0079) from that of the current allometricequation This indicates that allometric equations from the

250

300

350

0

100

200

50

150

Model 13 (this study)Hozumi et al 1969Beekman 1981Gerwing and Farias 2000

Putz 1983Schnitzer et al 2006

Gerhing et al 2004

TAG

B (k

g)


Figure 7 Comparison of model 13 with previously reportedmodelsfor determination of liana total above-ground biomass (TAGB) intropical forests

same region may be more similar in their estimation ofbiomass Among the previous equations was one that wasdeveloped from data sets of five different sites from fourcountries (Table 3) Consequently this equation is considereda more general equation which could be more useful in otherforest ecosystems in the tropics However this equation evenoverestimated total above-ground biomass of lianas by asmuch as 44 The underestimation of total above-groundbiomass of lianas by the current equation may be due tolow liana wood density in the Penang National Park fromwhich it was developed Equations developed from low wooddensity forests are likely to underestimate plant biomasswhereas equations developed from high wood density forestsmay overestimate plant biomass in some forests [12 32] Thecomparison of the previous equations to the current one hasrevealed the need for site specificmodels to be encouraged foraccurate determination of liana biomass in tropical forestsHowever where site specific allometric equations are notavailable caremust be taken in choosing allometric equationsfor forests In that case equations within the same region orcontinent should be explored first

4 Conclusion

Some allometric relationships were developed for liana stemand total above-ground biomass using liana diameter lengthdiameter squared and a combination of them as predictorsThree types of liana allometric equations were developed (1)models fitted to logarithmic transformed data (2) weightedmodels fitted to data (both dependent and independent data)on original arithmetic scale and (3) models fitted to data


Table 7 Comparison of mean estimated total above-ground bio-mass (per species) between model 13 and six previously publishedallometric equations (see Table 3) with t-testThe value in parenthe-sis represents the mean biomass for the current equation whereasthose outside the parenthesis represent the means of the previousequations

Pair Mean 119875 valueModel 13 versus Gehring et al[21] 12664 (6240) 0023

Model 13 versus Gerwing andFarias [20] 24230 (6240) lt0001

Model 13 versus Putz [29] 10868 (6240) 0019Model 13 versus Hozumi et al[22] 8707 (6240) 0079

Model 13 versus Beekman [30] 11648 (6240) 0034Model 13 versus Schnitzer et al[3] 1403 (6240) 0013

on original arithmetic scale Liana diameter was a betterestimator of liana stem and total above-ground biomasscompared to liana length Hence the models that used onlyliana diameter as the parameter estimate are recommendedfor use Liana allometric equations developed in the studycould be used in both primary and secondary forests as theywere not affected by forest type A comparison of the besttotal above-ground allometric equation developed in thisstudy (model 13) with previously published models indicatedthat the previous equations overestimated total above-groundbiomass of lianas by at least 29

Conflict of Interests

The authors do not have any direct financial relation orwhatsoever with the content of their paper that might leadto a conflict of interests for any of the authors Consequentlythey declare no conflict of interests

Acknowledgments

This study was supported by TWAS-USM PostgraduateFellowship and Research University (RU) Grant (1001PBI-OLOGI815086)The authors are grateful toMrAbuHusin ofthe Forest Research InstituteMalaysia andMrNtimGyakariof the Forestry Commission of Ghana for their assistancein plant identification They finally thank Mr S M Edzhamof the School of Biological Sciences USM Malaysia for hisimmense assistance in the field

References

[1] S A Schnitzer and F Bongers ldquoThe ecology of lianas and theirrole in forestsrdquoTrends in Ecology and Evolution vol 17 no 5 pp223ndash230 2002

[2] P Addo-Fordjour A K Anning E A Atakora and P SAgyei ldquoDiversity and distribution of climbing plants in asemi-deciduous rain forest KNUST Botanic Garden GhanardquoInternational Journal of Botany vol 4 no 2 pp 186ndash195 2008

[3] S A Schnitzer S J DeWalt and J Chave ldquoCensusing andmeasuring lianas a quantitative comparison of the commonmethodsrdquo Biotropica vol 38 no 5 pp 581ndash591 2006

[4] F Bongers M P E ParrenM D Swaine andD Traore ldquoForestclimbing plants ofWest Africa introductionrdquo in Forest ClimbingPlants of West Africa Diversity Ecology and Management FBongers M P E Parren and D Traore Eds pp 5ndash18 CABInternational Oxfordshire UK 2005

[5] K Lertpanich and W Y Brockelman ldquoLianas and environ-mental factors in the Mo Singto Biodiversity Research plotKhao Yai National Park Thailandrdquo Natural History Journal ofChulalongkorn University vol 3 pp 7ndash17 2003

[6] Y Tang R L Kitching and M Cao ldquoLianas as structuralparasites a re-evaluationrdquo Chinese Science Bulletin vol 57 no4 pp 307ndash312 2012

[7] D R Perez-Salicrup ldquoEffect of liana cutting on tree regenera-tion in a liana forest in Amazonian Boliviardquo Ecology vol 82 no2 pp 389ndash396 2001

[8] D R Perez-Salicrup and M G Barker ldquoEffect of liana cuttingon water potential and growth of adult Senna multijuga (Cae-salpinioideae) trees in a Bolivian tropical forestrdquoOecologia vol124 no 4 pp 469ndash475 2000

[9] S A Schnitzer J W Dalling and W P Carson ldquoThe impactof lianas on tree regeneration in tropical forest canopy gapsevidence for an alternative pathway of gap-phase regenerationrdquoJournal of Ecology vol 88 no 4 pp 655ndash666 2000

[10] W F Laurance ldquoReflections on the tropical deforestation crisisrdquoBiological Conservation vol 91 no 2-3 pp 109ndash117 1999

[11] International Tropical Timber Organization (ITTO) ITTOGuidelines For the Restoration Management and Rehabilitationof Degraded and Secondary Tropical Forests ITTO PolicyDevelopment Series No 13 International Tropical TimberOrganization (ITTO) Yokohama Japan 2002

[12] T Kenzo R Furutani D Hattori et al ldquoAllometric equationsfor accurate estimation of above-ground biomass in logged-over tropical rainforests in Sarawak Malaysiardquo Journal of ForestResearch vol 14 no 6 pp 365ndash372 2009

[13] J Chave D Coomes S Jansen S L Lewis N G Swenson andA E Zanne ldquoTowards aworldwidewood economics spectrumrdquoEcology Letters vol 12 no 4 pp 351ndash366 2009

[14] D Tilman P B Reich J Knops D Wedin T Mielke and CLehman ldquoDiversity and productivity in a long-term grasslandexperimentrdquo Science vol 294 no 5543 pp 843ndash845 2001

[15] D B MacKay P M Wehi and B D Clarkson ldquoEvaluatingrestoration success in urban forest plantings in Hamilton NewZealandrdquo Urban Habitats vol 6 no 1 2011

[16] Q M Ketterings R Coe M Van Noordwijk Y Ambagaursquoand C A Palm ldquoReducing uncertainty in the use of allometricbiomass equations for predicting above-ground tree biomass inmixed secondary forestsrdquo Forest Ecology and Management vol146 no 1ndash3 pp 199ndash209 2001

[17] S A Schnitzer and F Bongers ldquoIncreasing liana abundanceand biomass in tropical forests emerging patterns and putativemechanismsrdquo Ecology Letters vol 14 no 4 pp 397ndash406 2011

[18] J R Moore ldquoAllometric equations to predict the total above-ground biomass of radiata pine treesrdquo Annals of Forest Sciencevol 67 no 8 p 806 2010

[19] E E Hegarty and G Caballe ldquoDistribution and abundance inforest communitiesrdquo in The Biology of Vines F E Putz andH A Mooney Eds pp 313ndash335 Cambridge University PressCambridge UK 1991


[20] J J Gerwing and D L Farias ldquoIntegrating liana abundance andforest stature into an estimate of total aboveground biomass foran eastern Amazonian forestrdquo Journal of Tropical Ecology vol16 no 3 pp 327ndash335 2000

[21] C Gehring S Park and M Denich ldquoLiana allometric biomassequations for Amazonian primary and secondary forestrdquo ForestEcology and Management vol 195 no 1-2 pp 69ndash83 2004

[22] K Hozumi K Yoda S Kokawa and T Kira ldquoProductionecology of tropical rain forests in south-western Cambodia IPlant biomassrdquo Oecologia vol 145 pp 87ndash99 1969

[23] H C Wern and C N Weng ldquoThe Potentials threats andchallenges in sustainable development of Penang NationalParkrdquoMalaysian Journal of Environmental Management vol 11pp 95ndash109 2010

[24] P Addo-Fordjour Z B Rahmad and A M S ShahrulldquoEffects of human disturbance on liana community diversityand structure in a tropical rainforest Malaysia implication forconservationrdquo Journal of Plant Ecology vol 5 no 4 pp 391ndash3992012

[25] G L Baskerville ldquoUse of Logarithmic regression in the estima-tion of plant biomassrdquo Canadian Journal of Forest Research vol2 no 1 pp 49ndash53 1972

[26] D G Sprugel ldquoCorrecting for bias in log-transformed allomet-ric equationsrdquo Ecology vol 64 no 1 pp 209ndash210 1983

[27] B R Parresol ldquoAssessing tree and stand biomass a review withexamples and critical comparisonsrdquo Forest Science vol 45 no4 pp 573ndash593 1999

[28] G M Furnival ldquoAn index for comparing equations used inconstructing volume tablesrdquo Forest Science vol 7 pp 337ndash3411961

[29] F E Putz ldquoLiana biomass and leaf area of a ldquotierra firmerdquo forestin the Rio Negro Basin Venezuelardquo Biotropica vol 15 no 3 pp185ndash189 1983

[30] F Beekman Structural and Dynamic Aspects of the Occurrenceand Development of Lianes in the Tropical Rain Forest Depart-ment of Forestry Agricultural University Wageningen TheNetherlands 1981

[31] S T Gower C J Kucharik and J M Norman ldquoDirectand indirect estimation of leaf area index f(APAR) and netprimary production of terrestrial ecosystemsrdquo Remote Sensingof Environment vol 70 no 1 pp 29ndash51 1999

[32] T Kenzo T Ichie D Hattori et al ldquoDevelopment of allometricrelationships for accurate estimation of above- and below-ground biomass in tropical secondary forests in SarawakMalaysiardquo Journal of Tropical Ecology vol 25 no 4 pp 371ndash3862009

[33] C Wang ldquoBiomass allometric equations for 10 co-occurringtree species in Chinese temperate forestsrdquo Forest Ecology andManagement vol 222 no 1ndash3 pp 9ndash16 2006

Submit your manuscripts athttpwwwhindawicom

Forestry ResearchInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2013

Environmental and Public Health

Journal of



EcosystemsJournal of

ISRN Soil Science


EcologyInternational Journal of


Marine BiologyJournal of


Hindawi Publishing Corporationhttpwwwhindawicom

Applied ampEnvironmentalSoil Science

Volume 2013

Volume 2013

ISRN Biodiversity


Atmospheric SciencesInternational Journal of



Waste ManagementJournal of

ISRN Environmental Chemistry



Volume 2013

Geological ResearchJournal of

ISRN Forestry

Volume 2013Hindawi Publishing Corporationhttpwwwhindawicom


Volume 2013

BiodiversityInternational Journal of

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2013

OceanographyInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawi Publishing Corporation httpwwwhindawicom Volume 2013

The Scientific World Journal

ISRN Ecology



ClimatologyJournal of


structure diversity and productivity [13 14] Consequentlythere is the need to assess biomass levels of primary andsecondary forests from time to time Knowledge of theamount of biomass stored up within forests is importantfor a couple of reasons Firstly biomass data is fundamentalin evaluating the productivity structure and conditions offorests [15] Secondly biomass measurements are useful inestimating the amount of carbon sequestered by forests [16]Furthermore it plays an important role in elucidating theeffects of deforestation and carbon sequestration on globalcarbon balance [16] Although lianas continue to increase inbiomass in tropical forests [17] they have not been factored inmost forest biomass assessments In recent times there havebeen growing calls for more biomass assessment of tropicalforests because of the need to estimate forest carbon stocksand the potential amount of biomass available as a source ofenergy [18] Unfortunately all the calls were made in relationto only tree biomass with no mention of lianas This is quiteunfortunate in view of the fact that lianas can add up to 30of total above-ground biomass in tropical forests with denseliana population (cf [17])

Though there are several ways of determining biomass ofplants in forests [18] two main approaches are commonlyused These are destructive harvesting of plants and thendetermining their biomass and the use of allometric equa-tions to predict biomass of plants However it is preferablethat allometric equations be used to estimate forest biomassas they allow for the estimation of large forest areas and avoiddestruction of forest There are many allometric equationsthat have been developed for the estimation of plant biomassin many tropical forests Nevertheless most of the allometricequations are for trees with only a few on lianasThe scarcityof liana allometric equations is partly due to the difficulty inaccessing the whole length of lianas from trees Currently itis known that lianas can contribute up to 30 of total above-ground biomass in liana-dense tropical forests (cf [17]) Sincelianas are disturbance adapted group of plants [19] they arelikely to increase in abundance in the face of increasingdisturbance in tropical forests and therefore their biomasscould even be higherWith this development the importanceof lianas in the forest would continue to increase withtime and therefore ignoring them in biomass assessmentof forests could lead to underestimation of forest biomass[20] Accordingly lianas should be integrated into tropicalforest biomass assessments so as to obtain the true amountof tropical forest biomass To this end more allometricequations of lianas must be developed for tropical forestecosystems tomake biomass determination of lianas possibleand an integral part of forest biomass assessment

Even though a few liana allometric equations have beendeveloped for the estimation of above-ground biomass intropical forests [3 20 21] there is no allometric equationfor lianas in Malaysia To date only one allometric equationhas been developed for the estimation of liana biomassin the whole of Southeast Asia [22] To know the relativecontribution of above-ground plant parts to total above-ground biomass plant part allometric equations must bedeveloped Most liana allometric equations in use onlymeasure total above-ground biomass without the individual

6∘U

525

kilometers

0

66∘UUUUU

Study site

5∘20

998400U

110∘10

998400T 110∘20

998400T

International boundary

Pulau

N

Penang

Thailand

Sumatera

Straitsof

Malacca

SouthChina

SeaPeninsularMalaysia

100∘T 101

∘T 102∘T 103

∘T 104∘T

5∘U

4∘U

3∘U

2∘U

1∘U

Figure 1Map of Penang State showing the location of the study area(Penang National Park)

components Since above-ground portion of lianas have twomain components (stem and leaf) developing allometricequations for at least one of them should make it possibleto determine the biomass of the other component whenit is used in conjunction with total above-ground biomassequations Therefore the study was aimed at developingallometric equations for the estimation of stem and totalabove-ground biomass of lianas in primary and secondaryforests within the Penang National Park Penang Malaysia

2 Materials and Methods

21 Study Area The study was conducted in the PenangNational Park Penang Malaysia (N 5∘ 275831015840 E 100∘ 123501015840)(Figure 1) which has a total area of 1213 haThe national parkis comprised of primary and secondary forest typesThe parkcontains many plant species (1000 species) the majority ofwhich are from Dipterocarpaceae Fabaceae ApocynaceaeAnacardiaceae Euphorbiaceae and Moraceae A total of 190species of animals have been recorded and these include25 mammal species 53 butterfly species 46 bird speciesand other reptiles insects and amphibians [23] The PenangNational Park was designated as a national park in April 42003 and gazetted under National Park Act 226 of 1980 onApril 10 2003

The secondary forest within the Penang National Parkhas experienced various forms of human disturbance sincethe early years of the twentieth century Logging was a majorform of disturbance in the secondary forest although humanactivities such as farming hunting and gathering occurredin it The secondary forest possesses relics of considerablehistorical land use activities such as clearcutting farminglogging and bush burning due to its close proximity tothe community The secondary forest is also characterisedby some ancient logging routes confirming its extensive use



















Rourea rugosa 1 1 2

Salacia sp 1 2 3









Length (m) 400 2620 940






FI = 1[1198911015840 (119884)]

radicMSE (2)





Equation c 120572



Putz [29] 0036 1806










14121086420

25

20

15

10

5

0

Stem

bio

mas

s (kg

)

Diameter (cm)


30

TAG

B (k

g)


25

20

15

10

5

0
























121080604020

15

125

1

075

05


log 1

0(T

AGB

in k

g)





20

15

10

5

014121086420

Diameter (cm)

(Ste

m b

iom

ass i

n kg

)09





20

15

10

5

0

(TAG

B in

kg)

09



Model CFStem Total






250

300

350

0

100

200

50

150



Gerhing et al 2004

TAG

B (k

g)




4 Conclusion











Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology






















Rourea rugosa 1 1 2

Salacia sp 1 2 3









Length (m) 400 2620 940






FI = 1[1198911015840 (119884)]

radicMSE (2)





Equation c 120572



Putz [29] 0036 1806










14121086420

25

20

15

10

5

0

Stem

bio

mas

s (kg

)

Diameter (cm)


30

TAG

B (k

g)


25

20

15

10

5

0
























121080604020

15

125

1

075

05


log 1

0(T

AGB

in k

g)





20

15

10

5

014121086420

Diameter (cm)

(Ste

m b

iom

ass i

n kg

)09





20

15

10

5

0

(TAG

B in

kg)

09



Model CFStem Total






250

300

350

0

100

200

50

150



Gerhing et al 2004

TAG

B (k

g)




4 Conclusion











Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology






Equation c 120572



Putz [29] 0036 1806










14121086420

25

20

15

10

5

0

Stem

bio

mas

s (kg

)

Diameter (cm)


30

TAG

B (k

g)


25

20

15

10

5

0
























121080604020

15

125

1

075

05


log 1

0(T

AGB

in k

g)





20

15

10

5

014121086420

Diameter (cm)

(Ste

m b

iom

ass i

n kg

)09





20

15

10

5

0

(TAG

B in

kg)

09



Model CFStem Total






250

300

350

0

100

200

50

150



Gerhing et al 2004

TAG

B (k

g)




4 Conclusion











Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology

























121080604020

15

125

1

075

05


log 1

0(T

AGB

in k

g)





20

15

10

5

014121086420

Diameter (cm)

(Ste

m b

iom

ass i

n kg

)09





20

15

10

5

0

(TAG

B in

kg)

09



Model CFStem Total






250

300

350

0

100

200

50

150



Gerhing et al 2004

TAG

B (k

g)




4 Conclusion











Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology






20

15

10

5

0

(TAG

B in

kg)

09



Model CFStem Total






250

300

350

0

100

200

50

150



Gerhing et al 2004

TAG

B (k

g)




4 Conclusion











Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology













Acknowledgments


References







































Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology























Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology








Journal of




ISRN Soil Science








Volume 2013

Volume 2013

ISRN Biodiversity









Volume 2013


ISRN Forestry



Volume 2013







ISRN Ecology




Documents

Development of Allometric Equations for Estimating Above ...downloads.hindawi.com/journals/ijeco/2013/658140.pdf2 InternationalJournalofEcology structure,diversity,andproductivity[13,14].Consequently,