UBC Faculty Research and Publications

Pacific silver fir site index in relation to ecological measures of site quality Klinka, Karel 1999-04-08

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


52383-SSES020.pdf [ 171.1kB ]
JSON: 52383-1.0107256.json
JSON-LD: 52383-1.0107256-ld.json
RDF/XML (Pretty): 52383-1.0107256-rdf.xml
RDF/JSON: 52383-1.0107256-rdf.json
Turtle: 52383-1.0107256-turtle.txt
N-Triples: 52383-1.0107256-rdf-ntriples.txt
Original Record: 52383-1.0107256-source.json
Full Text

Full Text

Scientia Silvica Extension Series, Number  20, 1999Pacific Silver Fir Site Index in Relation to EcologicalMeasures of Site QualityIntroductionEcosystem-specific  forest  management requirescomprehension of tree species productivity in managedsettings, and how this productivity varies with the ecologicaldeterminants of site quality, i.e., the environmental factorsthat directly affect the growth of plants: light, heat, soilmoisture, soil  nutrients, and soil  aeration. A goodunderstanding of this variation is necessary for makingspecies- and site-specific silvicultural decisions to maximizeproductivity.  Productivity of a given species is  usuallymeasured by site index (tree height at 50 years at breastheight age). Quantitative relationships between site indexand these measures of site quality provide predictive modelsfor estimating site index.Pacific silver fir (Abies amabilis (Dougl. ex Loud.) Forbes)is an important timber crop species in the coastal forestsof British Columbia. In relation to climate, its range insouthwestern British Columbia extends from sea level toalmost timberline, and from the hypermaritime region onwestern Vancouver Island to the subcontinental region onthe leeward side of the Coast Mountains. In relation tosoils, its range extends from slightly dry to wet sites andfrom very poor to very rich sites. In view of this relativelywide climatic amplitude, a large variability in productivitycan be expected. It is particularly important to considerthe growth performance of Pacific silver fir when decisionsare made regarding whether or not to cut stands on high-elevation sites. In the study summarized here, relationshipsbetween Pacific silver fir site index and selected ecologicalmeasures of site quality were examined, and site indexmodels using these measures as predictors were developed.Study Stands and ProcedureThe study area encompassed the  entire native range ofPacific silver  fir  in  southwestern  British Columbia;Vancouver Island and the adjacent mainland south of theline extending from Port McNeill to Lillooet. The studystands had regenerated naturally after a major disturbance(wind, fire or clearcutting) and were deliberately selectedacross the widest range of climate (measured by elevationand continentality), soil  moisture, and soil  nutrientconditions. A 0.04 ha plot was established in each of the98 study stands, the three largest diameter trees were cutand sectioned, and site index was determined from stemanalysis data. We allocated stands into 4 continentality strataaccording to plot location on biogeoclimatic maps. Thestrata were: (1) maritime (west Vancouver Island and theleeward side of the Coastal Mountains), (2) less-maritime(east Vancouver Island), (3) submaritime (the leeward sideof the Coastal Mountains within the MH zone), and (4)subcontinental (the leeward side of the Coastal Mountainswithin the ESSF zone). Soil moisture regime (SMR) andsoil  nutrient regime  (SNR)  were estimated using acombination of topographic and soil  morphologicalproperties, as well as understory vegetation. Only threeSNRs were used (poor, medium and rich), since no standssampled were identified with very poor SNR, and the fewvery rich sites sampled were combined with rich sites.Regression analysis was applied to examine the relationshipbetween site index and climate (measured by elevation andcontinentality) using a climosequence; i.e.  a  data setcomprised of 42 sites with fresh and moist SMRs and amedium SNR. Covariance analysis was applied to test theeffects of differential availability of soil  moisture andnutrients on site  index.  After adjusting for  elevation,differences in mean site index were tested between sitesstratified according to (1) SMRs across a hygrosequence,i.e., a data set comprised of 45 sites with medium SNR,and (2)  SNRs across a trophosequence,  i.e.,  a data setcomprised of 76 plots with fresh and moist SMRs (Table1).To develop and test predictive models for Pacific silver firsite index, the data set (98 plots) was randomly split into acalibration data set (67 plots) and a test data set (31 plots).Multiple regression  analysis was used to  fit  predictivemodels from climate and/or soil variables, and the precisionof fitted models was tested against independent data andfor bias using the root-mean square prediction error (root-MSPR = the square root of the mean squared differencesbetween predicted and measured site index) and paired t-tests,  respectively.  The best model was compared to  aclimate-specific, polymorphic site  index  model (seeScientia Silvica Number 19).G53G52G52G55 G50G48G47G4cG58G50 G55G4cG46._G4bG36G4cG57G48G3G4cG51G47G48G5bG3GbG50GcG14G13G14G15G14G17G14G19G14G1bG15G13G15G15G15G17Results and DiscussionPacific silver fir site index decreased rapidly with increasingelevation, but the decrease varied with continentality (Figure1). There was little difference between the maritime andless-maritime/submaritime strata; for every 100 m increasein  elevation, site  index  decreased 2.0 m and 2.4 m,respectively. In the subcontinental stratum, where Pacificsilver fir does not occur below 800 m in elevation, thedecrease per 100 m elevation gain was only 0.8 m.However,  for the  same elevation, site index  was muchlower in the less-maritime/submaritime strata, comparedto the maritime and subcontinental strata. This pattern mayreflect drier summers in the less-maritime/submaritimestrata  compared to the  maritime stratum,  or  a shortergrowing season due to higher winter precipitation comparedto the subcontinental stratum.Figure 1.  Regression lines showing the relationship betweensite index and elevation separately for the continentality strata:maritime (R2 = 0.89), less-maritime/submaritime (R2 = 0.89),and subcontinental (R2 = 0.69). All regressions are significantat p <0.001).G28G4fG48G59G44G57G4cG52G51G3GbG50GcG13 G15G13G13 G17G13G13 G19G13G13 G1bG13G13 G14G13G13G13 G14G15G13G13 G14G17G13G13 G14G19G13G13 G14G1bG13G13G36G4cG57G48G3G4cG51G47G48G5bG3GbG50GcG13G18G14G13G14G18G15G13G15G18G16G13G16G18G17G13G36G58G45G46._G52G51G57G4cG51G48G51G57G44G4fG36G58G45G50G44G55G4cG57G4cG50G48G12G57G55G44G51G56G4cG57G4cG52G51G30G44G55G4cG57G4cG50G48Figure 2.  Relationship of site index to (A) soil moistureregimes across a medium hygrosequence (n = 45) and (B) soilnutrient regimes across a fresh and moist trophosequence (n =76) when adjusting for elevation and continentality. Meansrepresent site index at the elevation of 992 m and 969 m,respectively. Error bars represent one standard error of themean.showed that poor sites had significantly lower mean siteindex compared to medium and rich sites (p = 0.001 and<0.001, respectively); however, no significant differencewas shown between medium and rich sites (p  =  0.713)(Figure 2B).  This pattern indicates  that  the  poor SNRrepresents sites with insufficient nutrient supply for theoptimal height growth of Pacific silver  fir.  Thus, forconstructing predictive models three groups of sites wereconsidered: (1) water-deficient (WD - slightly dry) sites,nutrient-deficient (ND - fresh and poor) sites, and siteswith sufficient soil moisture and soil nutrients (SWSN -remaining edatopes in Table 1).After adjusting for elevation, analysis of covariance showeda significant effect of available soil moisture on site index(n = 45, p = 0.02). Multiple comparisons of the marginalmeans using Bonferroni's adjustment indicated that  siteindex on slightly water-deficient sites (slightly dry SMR)was significantly lower than on fresh and moist sites (p =0.002 and 0.006, respectively; Figure 2A), reflecting thesensitivity of Pacific silver fir to water stress. No significantdifference was observed between fresh and moist sites (p= 1.0).Covariance analysis using the trophosequence showed atrend of increasing site index with increasing nitrogen (N)availability (n = 76, p  <0.001). After Bonferroni'sadjustment, multiple comparison of the marginal means(B)G56G4fG4cG4aG4bG57G4fG5cG3G47G55G5c G49G55G48G56G4b G50G52G4cG56G57G36G4cG57G48G3G4cG51G47G48G5bG3GbG50GcG14G13G14G15G14G17G14G19G14G1bG15G13G15G15G15G17(A)The climate model (Eq. [1]) accounted for a slightly smallerportion of the variation of site index and was less precise(lower root-MSPR) than the combined model using climateand soil variables (Eq. [2], Table 2). Both models wereunbiased when tested against independent data, since nosignificant  differences were found  using paired t-testsbetween measured and predicted site index  (p = 0.650 andFigure 3.  Scattergrams comparing the predicted to measured(from the test data, n = 31) site index using (A) the combinedclimate-edatope model (Eq. [2]) and (B) the climate-specificsite index model that used top height and breast height age aswell as continentality strata as predictors. Dashed linesindicate perfect correlation.0.795, respectively). The precision off the climate model(root-MSPR = 3.73 m) and off the combined model (root-MSPR = 2.93 m) was low,  when compared to a modelusing top  height, stand age and continentality strata aspredictors, which had a root-MSPR off 0.76 m (Figure 3)(also see Sfcientia Sfilvica Number 19). Although unbiasedsite index estimates off Paciffic silver ffir can be obtainedffrom the models using site variables, their low precisionmay restrict their application to the fforest rather than thestand-level.G33G55G48G47G4cG46._G57G48G47G3G56G4cG57G48G3G4cG51G47G48G5bG3GbfG50GcG13 G18 G14G13 G14G18 G15G13 G15G18 G16G13 G16G18 G17G13G30G48G44G56G58G55G48G47G3G56G4cG57G48G3G4cG51G47G48G5bG3GbfG50GcG13G18G14G13G14G18G15G13G15G18G16G13G16G18G17G13G33G55G48G47G4cG46._G57G48G47G3G56G4cG57G48G3G4cG51G47G48G5bG3GbfG50GcG13 G18 G14G13 G14G18 G15G13 G15G18 G16G13 G16G18 G17G13G30G48G44G56G58G55G48G47G3G56G4cG57G48G3G4cG51G47G48G5bG3GbfG50GcG13G18G14G13G14G18G15G13G15G18G16G13G16G18G17G13Table 1.  Number of study stands according to edatopes. Thestands used for a climosequence are italicized, for ahygrosequence bolded, and for a trophosequence underlined.Boxes refer to water-deficient (WD), nutrient-deficient (ND),and sites with sufficient water and nutrient supply (SWSN)used in the predictive model [2].Table 2.  Models for predicting site index (SI) from indirectclimatic (Eq. [1]), and indirect climatic (elevation andcontinenality strata as dummy variables) and soil variables (asdummy variables) (Eq. [2]). All models are significant (p<0.001, n = 98). R2adj is adjusted for the number of independentvariables. SEE is standard error of estimate. Root-MSPR is thesquare root of the mean squared differences betweenpredicted and measured site index using the test data (n = 31).ELE is elevation (m), SC is the subcontinental stratum, M isthe maritime stratum, ND and WD are defined in Table 1. Regression model R2adj  SEE  root- MSPR         Eq. [1] SI = 34.240 - 0.02105(ELE) + 0.01301(ELE*SC) + 6.0(M) - 11.035(SC) 0.78 3.87  3.73         Eq. [2] SI = 35.783 - 0.02080(ELE) + 0.01202(ELE*SC) + 4.97(M) - 9.852(SC) - 3.313(ND) - 6.047(WD) 0.84 3.12  2.93 G3(A)(B)ReferenceSplechtna, B. 2000. The growth off Abies amabilis (Dougl.ex Forbes) in relation to climate and soil in southwesternBritish Columbia. Ph.D. Dissertation, Forest SciencesDepartment, University off British Columbia, Vancouver,BC. In progress  Poor  Medium  Rich Total Slightly dry  14  3  - 17 Fresh 11  32  2 45 Moist -  10  21 31 Very Moist  -  -  5 5  25 45  28 98 G3WDND SWSNScientia Silvica  is published by tfhe Forestf Sciences Departfmentf,The Universitfy of Britfish Columbia, ISSN 1209-952XEditor: Karel Klinka (klinka@intferchange.ubc.ca)Researcfh: Bernhard Splechtfna (bsplechtfna@utfanetf.atf)Producftion and design: Christfine Chourmouzis (chourmou@intferchange.ubc.ca)Financfial support: Forestf Renewal Britfish ColumbiaFor more information cfontacft: Bernhard Splechtfna (bsplechtfna@utfanetf.atf)Copies available from: www.forestfry.ubc.ca/klinka, or K. Klinka, Forestf Sciences Departfmentf, UBC, 3036-2424 Main Mall, Vancouver, BC,  V6T 1Z4


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items