Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Analysis of measurement errors associated with variable-radius plot forest sampling Omule, Stephen Agnew Yen’Emurwon 1978-02-24

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

Item Metadata

Download

Media
831-UBC_1978_A6_7 O48.pdf [ 4.81MB ]
Metadata
JSON: 831-1.0075436.json
JSON-LD: 831-1.0075436-ld.json
RDF/XML (Pretty): 831-1.0075436-rdf.xml
RDF/JSON: 831-1.0075436-rdf.json
Turtle: 831-1.0075436-turtle.txt
N-Triples: 831-1.0075436-rdf-ntriples.txt
Original Record: 831-1.0075436-source.json
Full Text
831-1.0075436-fulltext.txt
Citation
831-1.0075436.ris

Full Text

ANALYSIS OF MEASUREMENT ERRORS ASSOCIATED WITH VARIABLE-RADIUS PLOT FOREST SAMPLING by STEPHEN AGNEW VWEMURWON OMULE B.Sc. For. (Hons), Makerere University, 1976 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN THE FACULTY OF GRADUATE STUDIES ( THE DEPARTMENT OF FORESTRY) We accept this thesis as conforming to the required standard THE UNTVERSlTy OF BRITISH COLUMBIA kptill, 1978 © Sttphtn Agnzw Vtn'Emuiwon Omult > J97s In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of f € y Iv vf The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 f Date -ii-ABSTRACT Forest inventories form a basis of decision-making in most aspects of forest resources management. Implicitly, therefore, inventories must be conducted efficiently and with a minimum of error. This thesis analyses the errors of measurement during a variable-radius plot forest inventory cruise, with the objective of determining crew variation and bias in tree count, diameter, and total height measurements. Data for the study were collected at The University of British Columbia Research Forest, Maple Ridge, British Columbia, during the third year forestry students field school in mensuration. In each of the six lots of 10 pre-set sample plot centers, three to six 3~man student crews each established prism and relascope plots, took the tree count, and measured the dia meters at breast height of all the "in" trees and the total heights of the first "in" trees. The crews took their measurements independently of each other, and had to complete the exercise within a period of 8 hours. The author's measurements and observations in the same plot centers formed the control to the study. The crew variation in tree count was calculated by the method of Analysis of Variance, and in diameter and height measurement by pooling the variation for each tree to obtain a weighted average variation per tree. Bias was evaluated by comparing the erew results with those of the control. The following results were obtained from the study: The average tree count was 9.5 trees in the basal area factor (BAF) - 6 prism plots and 7.4 trees in the BAF = 9 relasoope plots. The coefficient of variation of observer tree count v/as 10.445 with the prism and 4.93% with the relascope. The percentage error in the determination of basal area per hectare from 2 5 prism plots was as much as ± 4.09%. and from 25 relascope plots ± 1.93% at the a = .05 probability level. About 37% of the tree counts in the re lascope plots and 2 5% in the prism plots were measured without error. The maximum tree count error v/as ± 6 trees per plot. The average tree diameter v/as 52 .67 cm, and the measurer coefficient of variation was 8.16%. Only 6% of the tree diameter measurements were correct. Accuracy was lower at larger tree diameters. The average tree height was 31.38 m, and the measurer coefficient of variation of the tree height measurements was 21.86%. Crews measured tree heights with a significant bias. Only about 2% of the m.onr."romontf? were correct, with over 15% of the measurements being in error by + 6.0 m or more. Height measurement was subject to larger and more source s o f e rror, -iv-The results suggest that forest resouce managers should be more careful in using untrained crews in forest inventory work. They should establish rigorous field training programs, and outline and implement checkcruising guidelines. - AC TABLE OF CONTENTS Page TITLE PAGE i ABSTRACT ii LIST OF TABLES v LIST OF FIGURES vACKNOWLEDGEMENTS vii INTRODUCTION 1 LITERATURE REVIEW 4 SOURCE OF DATA 10 METHOD OF ANALYSIS AND RESULTS 13 Tree CountDiameter 24 Height 30 DISCUSSION 36 CONCLUSIONS 41 LITERATURE CITED 2 APPENDICES: I. Maps of the Study Area 45 II. Tree Counts in the Relascope (BAF = 9) Plots 47 III. Tree counts in the Prism (BAF = 6) Plots 8 IV. Tree Diameter Measurements ... 49 V. Tree Total Height Measure ments 57 VI. Sample Questionnaire of the Instrument User Preference Survey 9 -vi-LIST OF TABLES Table Page I ANOVA Table for Tree Count . 15 II Important Statistics from the Tree Count ANOVA 17 III ANOVA Table for the Prism Plots 1IV ANOVA Table for the Relascope Plots 18 V ' Tree Count Accuracy in the Prism Plots 19 VI Tree Count Accuracy in the Relascope Plots 19' VII Tree Diameter Measurements Accuracy 26 rIII The Distribution of Diameter Measurement Errors by Diameter Classes 29 IX Tree Total Height Measurement Accuracy 3.1 X The Distribution of Height Measurement Errors by Height Classes 32 -vii-LIST OF FIGURES Figure „ la A Plot of the Residuals against True Counts with the Relascope 2G lb A Plot of the Residuals against True Counts with the Prism 1 2a Relationship between True Count and the Crew (Estimated) Count in the Relascope Plots 22 2b Relationship between True Count and the Crew (Estimated) Count in the Prism Plots 23 3 A Plot of the Residuals against True Diameter Measurement 2 7 4 Relationship between True Diameter and the Crew (Estimated) Diameter Measurement 2 3 5 A Plot of the Residuals against True Total Height Measurement 33 6 Relationship between True Height and the Crew (Estimated) Total Height Measurement 3<1 -viii-ACKNOWLEDGEMENTS I am most grateful to Dr. D.D. Munro for suggesting the problem and supervising me on the study, to the members of my committee - Drs. J.P. Demaerschalk, A. Kozak and D.D. Munro - for reviewing the thesis, and to Dr. S.W. Nash of the Department of Mathematics for recommending the Analysis of Variance Model used herein. I am greatly indebted to Makerere University, Uganda, for the financial support in form of a study fellowship, and to the Ford Foundation for Eastern Africa for administering the fellowship. For the additional financial support in form of the Donald S. McPhee Forestry Award, I acknowledge The University of British Columbia and all those who made it possible. I am also grateful to the Faculty of Forestry 1977 Spring School Class of whom I took advantage to collect data for the study, to all the members of the Biometrics Group, Faculty of Forestry, for their invaluable suggestions, and to The University of British Columbia Computing Centre for the provision of computing facilities. -1-INTRODUCTION Forest inventories form a basic requirement in most aspects of forest resources management. It is therefore necessary to conduct inventories efficiently and with a minimum of error. Variable-radius plot (VRP) sampling is an efficient, unbiased, and valid forest inventory technique; however, significant variations have been observed among the results of independent cruisers. These variations are a result of measurement errors caused by measurer bias or use of faulty instruments. This thesis examines the errors of measurement during a VRP forest cruise, attributable to the measurer. The main parameters measured in a VRP cruise are: the tree count, the diameter at breast height (dbh), and total height. The sources of possible measurement errors are different for each parameter. The common causes of the tree count errors, as a con sequence of observer bias, are the following: a) failure to select "doubtful" trees, b) missing hidden trees, c) failure to make exactly a 360° sweep, d) failure to align instrument with the tree, e) failure to correct for slope, f) moving instrument away from the plot center. A small error in tree count results in a large percentage error in the estimate of basal area per unit area, because VRP samples are based on a relatively small number of trees per plot. In British Columbia, the instrument used for measuring tree dbh is the diameter tape. During measurement systematic errors of random magnitude, occur. They are caused by the following: a) the tilting of the tape such that part or whole of the circumference of the tape is below or above the correct measuring plane, often resulting in an overestimate, b) the taking of measurements at heights other than 1.3m above average ground level (or germination point) of a tree, c) personal judgement in regard to the irregular stem forms. These errors have a significant influence in the estimate of tree volumes, especially if the tree heights are also in accurately determined* The common causes of error in tree height measurement based on trigonometric principles, are the following: a) misreading instruments as a result of unsteadiness, b) difficulty in locating suitable points from which to observe the exact position of the tree top and base, c) horizontal distance measurement errors - not pulling the tape tight, incorrect alignment of the tape, erroneous location of the zero point of tape, ommission of whole tape length, and the use of in correct tape length, d) biases of individual persons as a result of leaning trees, -3-Measurement of tree heights requires skill and care, without which large errors may result. Measurement errors influence the reliability of an inventory result. This influence is often underestimated, mainly because of the difficulty in obtaining well-founded quantitative estimates of the errors. The objective of this study will be to determine the biases and variation among independent cruisers, of tree count, dbh, and total height measurement. This will be used as a basis for the establishment of guidelines for performance and checkcruising, and for revising the forest inventory measurement standards of the British Columbia Forest Service if desired. -4-LITERATURE REVIEW A measurement error is the difference between a true value of a unit and an inexact measurement of the unit. It may be due to the following, FAO (1973) : (a) a constant bias, (b) a variable component relative to the sampling unit being correlated to the exact value of the measurement perameter in the corresponding unit, (c) an arbitrary component with mean zero. The sources, kinds, and influences of measurement errors have been discussed by Loetsch et al. (1973) . They emphasized the importance of measurement errors and classified the error sources as follows: (a) peculiarities of the object being measured, (b) uncertainties in the measuring procedure, (c) inaccuracy of the measuring device, (d) topographical or physical influences, (e) defects in the observer senses. They further distinguished between "true" and "apparent" errors, and stated (p. 12) : "for comparative measurements used in practice for estimating the accuracy and efficiency of a measuring procedure and as a confirmation of the correctness of a statement in general, only "apparent" deviations (errors) can be proven". Measurement errors have also been discussed by Cochran (1973) and Carron (1968) . Ferguson (1975) pointed out that although the measurement errors affecting forest inventory estimates have long been known, there have been few empirical studies on the magnitude of the effect of these errors. Some of these studies will be -5-mentioned below. Kendall and Sayn-Wittgenstein (1959) emphasized that of the sources of error in VRP forest inventory, those attributable to the observer were most important and frequent. Ker, et al_. (19 57) reported biases of untrained students in estimating basal area per acre (ha.) with a 3-diopter (BAF = 2.2) prism, to be between -37% to 1%. They found a tree count variation among the students to be 1 tree per plot and an average negative bias of 1.2% of the actual tally. They also observed that errors were not associated with ground slope, but rather with the age of the stand. Carow (1958) observed that there was a definite personal bias in judging line trees with a relascope, and that the coefficients of variation in basal area determination in creased with increase in BAF. He observed a 38.7% coefficient of variation with a 49.0 minute (BAF = 2.17) and 69.0 minute (BAF = 4.35) angle count, and 42.9% with a 97.4 minute (BAF = 8.71) angle. Kendall and Sayn-Wittgenstein (1959) compared the errors made by four operators taking independent acre (ha ) tree counts at 9 locations (and using different BAFs) with true counts obtained by careful measurement. They obtained an average error of -6% with BAF = 5 (1.14), -3.5% with BAF = 10 (2.29), 1.5% with BAF = 20 (4.59) and 4.5% with BAF = 40 (9.18) . They attributed the large error percent by the observers to the fact that only a small number of sample points was used and that the observers did not check the -6-"doubtful" trees. Stage (1962) reported on tests of doubtful tree judgement made by field crews and forestry students matching photographs. He deduced that the assumption of tallying one-half of marginal trees could not be relied upon as not unbiased. He also noted that qualifying trees were more likely to be missed with smaller BAFs, and that personal bias in prism sampling could be a result of differences in response by different persons to tree bark and background colour. Willingham (196 2) on experiments with prism calibration reported that a greater magnitude of error was likely to occur in the exclusion of trees that should be counted. He emphasized the importance of a personal calibration for each user of a given prism. Sayn-Wittgenstein (1963) reported that there was one tree about which there was some doubt for every two trees the observer was sure about. With the small BAFs, associated with high tree counts per point, the number of doubtful trees was large. This provided an opportunity for bias (mainly an underestimate). Sayn-Wittgenstein also showed that personal bias enters into the errors associated with large BAFs and low tree count per point. A properly executed relascope cruise yields results of only a negligible bias. Kirby (1965) stressed that bias in inventory estimates could be prevented by eliminating as much as possible, human judgement in the field. Munro (1966) reported a coefficient of variation of ± 11.15% among student crew estimates of individual plots. All the cruisers had been instructed to check doubtful trees. The same study showed a negative, statistically insignificant bias of 2.1% in tree count. There was no consistent relation ship between personal error and the number of trees per plot, slope of the plot, size, species or position of trees within the plot. 92% of the plots measured had errors of ± 1 tree or less. A BAF = 30 (6.88) prism was used. Holgate (1967) showed mathematically that small critical angles were associated with smaller variances of tree count. He confirmed his results with the findings of Husch (1955) . Carow and Rickerd (1969) reported the results of a study of personal bias in point sampling. A cruiser over estimated basal area per acre (ha ) by 6% with the BAF = 10 (2.29) and BAF = 20 (4.59) prisms, came close with the BAF = -40 (9.18) wedge, and made a slight underestimate with the BAF = 75 (17.21) angle gauge. They reported that the between people variance was statistically significant at the 5% level, in an experiment to measure the critical distance of a BAF = 10.25 (2.35) prism. However, individual personal bias tended to be consistent irrespective of the BAF used, though it was possible for an individual to have a positive bias with one BAF and a negative one with another. They re commend checkcruising to be done in terms of basal area, rather than tree count. -8-Laar (1970) compared estimates of basal area per ha obtained with a relascope (BAF =8.7) and a wedge prism (BAF = 10.25), and found that the relascope showed a negative bias of 4.5% and the prism was unbiased (using the fixed plot tree counts as the control). Bias in relascope estimates was attributed to the personal element in the evaluation of "borderline" trees. He reiterated that pro viding more contrast between tree stems and their background did not improve the determination of the status of "border line" trees. This conflicts with Stages' (1962) recommenda tions of providing contrast between tree stems and their back ground. Few investigations have been made on the variation in diameter at breast height and in tree total height measure ments. Some of these investigations are mentioned in the following paragraphs. Myers (1961) found that 94% of the measurements, under taken by forestry students and field foresters on permanent sample plots of ponderosa pine, were within 0.1 inch (0.254 cm) of the true diameter. Modal diameter was taken to be the "true" diameter of a tree. He noted no trend of variation in accuracy with increase in diameter at breast height .Sixty-two percent of the measurements were correct. There were more measurements that were too large than were too small, indicating more sources of positive errors. Most of the large errors were due to incorrect tape readings or recordings; but these greatly incorrect measurements were equally larger and smaller -9-than the true dbh. Ker (1951) reported standard errors of single tree heights of 1.4% for immature Douglas fir and 2.3% for western hemlock. Meyer (195 3) noted that repeated measurements made with the United States Forest Service hypsometer and the Christen hypsometer of six different trees resulted in standard errors of heights varying between 1.1% to 3.5%. Ker and Smith (1957) found a standard error of 2.4% for forestry students working under near optimum conditions using a relescope, and 1.8% using a Bjumleiss hypsometer .They ex pressed the view that tree height measurements were difficult and expensive to obtain and were subject to larger and more sources of error. Schmid et al. (1971) noted that in tree height measure ment, the errors of measurement increased in almost exact proportion to the height of the tree, irrespective of the instrument they used. They also observed that systematic errors resulted when the trees were not exactly vertical or when the top was flattened off (and the person measuring did not stand far enough away from the tree base) . -10-SOURCE OF DATA Data for this study were collected at The University of British Columbia Research Forest, Maple Ridge, British Columbia, during the third year forestry students field school in mensuration. Sixty plot centers, numbered'1 through 60, were system atically set at about 80 m (paced) intervals along the St. Jean's and Lakeshore trails of the research forest. The trails are in a second growth stand about 80 years old, containing mainly Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), western hemlock (Tsuga heterophylla (Raf.) Sarg.), and western redcedar (Thuja plicata Donn.) in intimate mixture. Site index ranges from 15m to 50m at 100 years. 3 Gross volume averaged 650 m /ha. Tree dbh ranged from 5 cm to 190 cm, the average diameter being 53 cm. Topography varied from steep (over 50% slope) to relatively flat (see maps 1 and 2, in the appendix I). Underbrush and the weather did not restrict line-of-sight visibility. The plot centers were divided into six lots of ten. Three to six 3-man student crews were allocated to each lot. Each crew established prism (BAF = 6) and relascope (BAF = 9) plots at the pre-set plot centers designated to them. The students had had a theoretical background of VRP sampling, and many of them had varied field experiences in prism and relascope cruising. Furthermore, the students were given two hours field instruction in prism use and were made aware of the following precautions to be borne in mind -11-during plot establishment with the prism and1 the relascope: i) the prism always to be held directly above the plot center when sighting at any tree in a 360° sweep about the point, ii) the prism to be held so that its base was perpendi cular to the stem of the tree being measured (except when the line of sight to the tree exceeded an angle of 10% from the horizonal when slope correction would be needed), iii) the prism to be held so that its face was perpendi cular to the line of sight, iv) all hidden trees to be examined by moving the observation point provided the exact distance from the plot point to the tree was maintained, v) all "doubtful" trees to be checked by applying the appropriate plot radius factor, vi) the relascope to be held directly above the plot center, vii) ensure that the relascope is corrected for slope, viii) observe the trees at 1.3 m above its average ground level (or germination point). In order to make meaningful comparisons and to establish limits of precision and accuracy, it was necessary that each crew adhere to the same procedures exactly. Plot tally was started from true North and each "in" tree (in cluding stumps greater than 2 m high) was tallied on a clockwise 360° sweep. The dbh of every "in" tree was measured to the nearest 0.1 cm with a linen diameter tape, and recorded in standard forms. The tree species name was -12-noted. The total height of the first "in" tree on each plot was measured to the nearest 0.1 mf. by trigonometric principles using a suunto or relascope and a nylon chain. A prism was used on the even-numbered plots, and a relascope on the odd-numbered plots. The crews made their measurements and recordings independently of each other, and had to complete their study within a period of eight hours. As a control to the study, I made independent observations and measurements, by carefully establishing prism and re lascope plots in all the pre-set sample plot centers and abiding by the same procedures the student crews had been instructed to follow. There was no time limit. The tree count, dbh, and total height measurements by the different crews and the control are given in the appendices II, III, IV and V. -13-METHOD OF ANALYSIS AND RESULTS Lot number 2 of the sample plots had only one set of crew measurements, and so was discarded from the first part of the analysis. Each of the remaining sample plots had 3 to 6 independent sets of crew measurement of (i) tree count, (ii) tree dbh, and (iii) tree total height. Prism plot measurements were separate from those of the relascope. As mentioned in the introduction, the aim of this analysis is to obtain a quantitative estimate of crew variation and bias in estimating the tree count per plot, and dbh and total height per tree. Firstly, the crew variation in tree count estimates was assessed by the method of analysis of variance (ANOVA). Secondly, the crew variation in measuring tree dbh and total height was assessed by calculating the variation for each tree and pooling these to obtain a weighted average variation per tree. In either case, bias was determined by comparing crew measurements with the control values. TREE COUNT The control set of data was treated together with the student crew data, so that for each lot there were up to 7 sets of measurement per plot. There was, therefore, a 25 x 7 tree count matrix, with some observations missing, for each of the prism plots and the relascope plots. First, the crew and plot variation at each plot was determined by the method of ANOVA. -14-The design used was repeated subsampling with unequal numbers in the subclasses. The plots and crews were nested within the lots. Lots varied from i = 1 to p = 5, plots from j = 1 to q = 5, and crews from k = 1 to s (varying between 4 and 7). The lots were considered fixed and the plots and crews random. The model was: X±jk= y + g. + Vj(i) + 6k(i) + ljk(i) (1) where: X. ., = tree count of the k crew on the i 3 th plot in the i lot u = overall tree count mean Der olot th = effect of the i lot vj(i) = effect of the j^11 plot within the lot 6k(i) = e^^ect °^ tne crew within the ith lot $ th th jk(i) = the interaction of the j plot and k crew within the ith lot (ERROR). Reference to this model is made to Ganguli (1941) and to Drs. A. Kozak and S.W. Nash (personal communication) . The general plan of the ANOVA is shown in Table I. TABLE I ANOVA Table For Tfee Count Source of Variation Degrees of Freedom (d..:f.) Between lots P-1 Sum of Squares SS. = p - - 2 qs. I n (x. - x ) 1=1 1.. Expected JMean Squares S? = a2 + qa2 + sa2 + qsa^ l ^5 VP Between plots-witain-lots p(q - 1) SS., = p q _ 2 si I (x.. - x. ) i=l j=l ID. l. S2 v= a2 + sc2 j(i) Between crews-wi thin-lots .T. n. - p i=l l ^ P s - - 2 SS= q I I (x. - x. ) C i=l k=l 1,JC X" i Error Total (q-1) (.1, n.- p)- m 1—X a. q I n. - 1 - m i-i SS = SSm --SS - SS - SS, E SS T L P C p q s _ 2 I I I (x. ., - x...) i=l j=l k=l 13 S2 o X = ^i.. x. . = _i.k x. . = ID-m = *l = cz = V tree count mean per plot of all the crews in all the plots in all the lots, tree count mean per plot of all the crews in all the plots in the i^1 lot. tree count mean per plot of the k^1 crew in all the plots in the i let. tree count mean per plot of all the crews in the j^1 plot in the i^ lot. th the number of crews in the i lot. the total number of missing observations in a tree count matrix. the lot variation* plot variation. crev; variation -I Cn? i=l (l/n.)-(l/ I n.) 1 i=l ~ p-1 (2 -16-The sums of squares SSL, SSp, SSC, SSE, SST in Table I were computed using the general regression method. The UBC GENLIN - a general least squares analysis of variance computer program of the University of British.Columbia -was used. The components of variance and their coefficients were derived with reference to Anderson and Bancroft (1952) . By using the appropriate mean squares from the ANOVA table, the estimates of crew variance, and plot variance, a2 , were made as follows: v S2 - S2 52 = _MiL o (3) 6 q S2 S2 a2 = o (4) v s The crew deviation estimate with the prism was 0.993 trees per plot and with the relascope 0.366 trees per plot. Plot deviation with the prism was 3.134 trees and with the relascope was 2.080 trees. Table II gives a summary of some of the more important statistics in the analysis, and tables III and IV give the ANOVA results of the prism and relascope plots, respectively. Secondly, each plot measurement was treated independently ir order to evaluate the tree count accuracy. Using the total number of plot measurements the variation of the plot measure ments was calculated by variation from control (VFC) classes. The following VFC classes were used in the analysis: 0, ± l,i 2> + 3, ± 4, ± 5, ± 6 trees. The maximum tree count error per plot was ± 6 trees and -17-TABLE II. Important Statistics From The Tree Count ANOVA-Instrument Overall Mean With 95% Conf. Limits (trees/plot) Coeff. of Variation of Crews (%) Stand Coeff. of Variation (%) Prism 9.509 ± 7.280 10.44 32.96 (8.880 ± 6.257) (34.08) Relascope 7.432 ± 5.420 4.93 28 .04 (7.400 ± 4.801 (26.82) Note: The values in brackets are the results of the Control measurements. TABLE III. ANOVA Table for the Prism Plots Source of Sum of Mean Variation d.f. Squares Square F-ratio Notes Between lots 4 408 .43 102.11 1. 513 N.S. Between plots-wi thin-lots 20 1250.90 62.54 28 .820 @ Between crews-wi thin-lots 26 184.74 7.10 3. 27 2 @ Error 100 217.01 2.17 Total 150 2061.08 @ Significant at ct = 0.05 probability level N.S. Not significant TABLE IV: ANOVA Table for the Relascope Plots Source of Variation d.f. Sura of squares Mean square F-ratio Notes Between lots 4 3 38.15 84.53 2. 915 *7 C Between Plots-within-lots 20 566 . 45 28 . 32 17.374 @ Between crews-within-lots 26 59.92 2. 30 1.411 N.S . Error 97 158.47 1.63 Total 147 1122.99 @ Significant at a = o.05 probability level N.S. Not significant. it occurred in 2 out of 130 prism plot measurements and in 2 out of 124 relascope plot measurements. The allowable tree count error in the British Columbia Forest Service is ± 1 tree, UBC Forest Club (1971). Tables V and VI summarize the accuracy in tree count, in the prism and relascope plots, respectively. There were more negative sources of error than positive ones (Fig. la, b). In the prism plots, 56% of the tree counts were over estimated, and in the relascope plots, 4 2% of the tree counts were overestimated. The accuracy of the crew tree counts against the control was tested by fitting a linear regression of the crew tree counts (X-.,) on the control tree counts (X ) , and then using the student t-test, to test whether the slope was statis tically different from unity. If the slope was not different from unity , the interccpi of t lie regression lino was tested whether it was statistically different from */.oro -19-TABLE V- Tree Count Accuracy in the Prism Plots VFC (Trees) No. of Plot •Measurements Percent of total (%). Percent of total (cumulative) (%) ±0 32 24.62 24.62 ±1 38 29 .23 53.85 ±2 27 20.77 74.62 ±3 16 12.30 86.92 ±4 6 4.62 91.54 ±5 9 6.92 98.46 ±6 2 1.54 100.00 Total 130 100.00 -Table VI. Tree Count Accuracy in the Relascope Plots VFC (trees) No. of plot Measurements Percent of total (%) Percent of total (cumulative) (%) + 0 46 37.10 37 .10 * 1 45 36.29 73.39 + 2 13 10.48 83.87 + 3 10 8 .06 91.93 + 4 4 3.23 95.16 + 5 4 3.23 98.39 ± 6 2 1.61 100.00 Total 124 100.00 --20-03 I ID I to I LU UJ OC + + 4 + + COnM + + + + —J I CC a | 4- + + + + + »—i CO o^-l + + 4- 4. 4. 4. 4. 4. 4. 4. 4. 4. 4. 4-4-4- 4-4-4- 4-4-, , , , , , R R— 0.0 2.0 A.a G.O 0.0 m.o 12.0 14.0 ie.o TRUE COUNT (TREES) Figure la A Plot of the Residuals Against True Counts With the Relascope 4-+ + + + + + + •f 4- + + 4- + + + 4- + + + + -f + + + 4- + + + + + -r + -+ + + + i 5.0 i B.O I 7.0 ' i 8.0 1 g.o I 1D.0 i 11.0 i 12. TRUE COUNT [TREES} Figure lb A Plot of the Residuals Against True Counts with the Prism -2 2-a -23-The fol lowing relationships wore obtained (sec? also Fig. 2a, b): Relascope _plots: 1.006 + 0.8 97 X C 0.58 7, SEE = 1.829 Prism plots: 1.065 + 0.998 X'c (6) 0.693, SEE - 2.117 where: 2 r = the coefficient of determination SEE = the standard error of the estimate At the a = 0.05 probability level, the intercept and slope of the regression line CO were not statistically different from zero This indicated that, on average, crews counted trees with only a negligible bias with the relascope, and counted trees with a significant bias with the prism. DIAMETER In total 398 trees were repeatedly measured for diameter by the different crews. The species measured were Douglas-fir, western hemlock and western redicdar „ Firstly, the bias of the crews in measuring dbh was assessed. In order to evaluate the accuracy in measuring the dia meters, each tree measurement was treated independently. There was a total of 1701 measurements, grouper) into 5 cm dbh classes. The control dbh measurements of each tree was accepted as the "true" dbh of the tree. The deviations from true dbh of the crew measurements were calculated for each tree. The following and unity, respectively; but the slope of the line ( 6) was. -25-formula was used for calculating the biases: where th Dg = dbh measurement bias on the i tree by the k.tn crew. th = true dbh of the i tree. th th Dik = dt)n of the *~ tree as measured by the k crew. The biases were then grouped into variation from control (VFC) classes for each dbh class. The following VFC classes were used: VFC Limits Class (cm) 0 0.0-0.09 1 + 0.1-1.0 2 + 1.1-2.0 3 + 2.1-3.0 4 + 3.1-4.0 5 + 4.1-5.0 6 + 5.1-6.0 7 + 6.1-7.0 8 + 7.1-8.0 9 + 8.1-9.0 10 + > 9.0 The variation in accuracy (bias) with increase in dbh was also investigated. This was done by separating the measurement biases by 5 cm diameter classes, on a relative basis. The average stand diameter was 52.67 cm from the crew estimates, and 52.72 cm by the control. Only about 6% of the measurements were correct, and 95% were within 5.0 cm of the true value. Fourty-cight % • Or the measurements were too large and 46% too small. This indicated more negative sources of error (see also fig. 3). A summary of the accuracy of diameter measurements is aiven in Table VII. The allowable dbh measurement error in the British Columbia Forest Service is ±1%, U.B.C. Forest Club (1971). -26-There seems to be a trend of variation in accuracy with increase in dbh. Accuracy tends to decrease at higher dbh trees. Examination of table VIII which shows measurement distribution by diameter classes and deviations from control may reveal this. TABLE VII. Tree Diameter Measurements Accuracy VFC (cm) Number of measurements Percent of total (%) Percent of total ! (cumulative) (%) 0 96 5.64 5 .64 1 943 55.44 61.08 2 271 15.94 77.02 3 142 8.35 85. 37 4 68 4.00 89.37 5 57 3.35 92.72 6 39 2.29 95. 01 7 13 0.76 95.77 8 23 1.35 97 .12 9 14 0.82 97.94 10 35 2.06 100.00 Total 1701 100.00 -Large errors were as much as 30 cm and were equally larger and smaller than the true dbh values. The accuracy of crews in measuring diameters was tested by comparing the crew measurements with the control values using the regression method adopted with the tree count. The following relationship was obtained (see also fig. 4). D_ = -0.527 + 1 .002 D„ (8) r2 = 0.979, SEE = 4.387 where: D,-, = estimated dbh in cm E Dp = control (true) dbh in cm to Figure 3 A Plot of the Residuals Against True Diameter Measurement TRUE DISMETER (CM) Figure 4 Relationship Between True Diameter and the Crew (Estimated) Diameter Measurement -29-TABLE VIII. The Distribution of Diameter Measurement Errors &y Diameter Classes Diameter Class Midpoint 0 1 2 ERROR (VFC) 3 4 5 6 CLASS 7 8 9 10 TOTAL (cm) NUMBER OF MEASUREMENTS 7 3 24 1 1 29 12 8 24 5 5 1 43 17 2 76 3 2 1 1 85 22 6 44 8 3 1 1 63 27 10 86 11 6 4 115 32 1 122 26 16 5 7 2 2 194 37 16 98 21 5 3 4 3 5 1 1 157 42 11 6 7 18 7 2 1 1 107 47 5 76 25 11 8 1 1 3 1 131 52 6 81 24 10 6 5 5 2 139 57 2 67 31 11 4 5 2 1 1 1 123 62 2 56 15 10 8 2 1 5 5 104 67 12 11 10 1 7 6 3 4 1 5 60 72 4 22 9 10 5 4 5 1 1 61 77 22 17 10 1 3 2 1 2 58 82 20 6 5 1 1 .1 1 1 36 87 1 14 7 3 1 1 1 28 92 5 9 4 1 3 1 2 25 97 1 1 2 3 1 2 10 102 1 7 5 1 4 3 1 1 26 107 1 5 2 2 10 112 2 1 3 1 1 8 117 1 1 2 12 2 5 1 1 1 1 1 10 127 1 1 3 2 4 2 13 132 1 6 4 1 3 1 1 1 18 137 1 2 1 2 2 2 10 142 1 4 1 2 8 . 152 1 3 2 6 162 1 1 1 1 1 1 3 9 167 1 1 1 1 2 6 170 1 1 1 4 7 TOTAL 96 94 3 271 14 2 68 57 59 13 23 14 35 1701 -30-The intercept and slope of the regression line (8) were not statistically different from zero and unity, respectively, at the a = 0.05 probability. This indicated that on average, crews measured tree diameters with only a negligible bias. Secondly, variation in dbh measurement from crew to crew was calculated by pooling the squared deviations of the crew measurements on all the trees and dividing the result by the pooled degrees of freedom of all the measured trees. Model 9 was used. n k _ 9 Z Z (D„. .- D^.K Op = 1=1 3=1 J  n I d.f.. i=l (9) where: = diameter measurement crew variation estimate j = crew j measurement on the l tree D_. = mean of the k measurements on the i^ tree £i 1 d.f.. = the degrees of freedom for the i"1"*1 tree (=k.- 1) The average crew deviation in diameter measurement was 4.3 cm per tree. The coefficient of variation = 8.16% (the average tree dbh was 52.67 cm). The variation of precision with increase in dbh was, in general, uniform (see Fig. 4). TOTAL HEIGHT In 31 plots the crews measured the same trees for height. In the rest of the plots the crews did not measure corresponding trees. The data for the 31 trees were analysed to determine the -31-accuracy and precision of crew measurement. The rest of the data were discarded. The procedure of analysis was similar to that used for diameter. Only seven VFC classes, and 1 m height classes were used. Less than 2% of the total measurements were correct, if control values are accepted as the "true" height. About 60% of the measurements were less than and 4 0% more than, the true values. Details of biases in height estimation are shown in Table IX. Tree Total Height Measurement Accuracy VFC Class Limits (ro) Number of measurements Percent of total (%) Percent of total (cumulative) (%) 0 ±0.0-0.09 2 1.82 1.82 1 ±0.1-1.0 19 17.27 19.09 2 ±1.1-2.0 17 15.46 34. 55 3 ±2.1-3.0 19 17.27 51.82 4 ±3.1-4.0 18 16.36 68 .18 '5 ±4.1-5.0 11 10. 00 78 .18 6 ±5.1-6.0 6 5.46 83.64 7 >± 6.0 18 16.36 100.00 Total 110 100.00 -The allowable tree height measurement error is ± 3% in the British Columbia Forest Service, UBC Forest Club (1971) There was no trend of variation in biases of the crews as height increased (see table X). Variability of the residuals around zero was quite high (see Fig. 5) . The tree heights ranged from about 8 m to about 51 m, the crew avaerage being 31.38 m. The control average was 32.33 m. -32-TABLE X. The Distribution of Height Measurement Errors  By Height Classes HEIGHT CLASS E R R 0 R (VFC) C L A S S MID-POINT(m) 0 1 2 3 4 5 6 7 TOTAL NUMBER OF EASUREMENTS 10. 5 1 1 1 1 4 11.5 4 1 4 18.5 2 3 2 2 1 J 10 21.5 1 i •; 2 22. 5 2 1 I 4 24.5 1 1 2 26.5 2 2 3 i j 8 27.5 1 1 2 2 6 29.5 1 1 3 i 5 30. 5 1 1 3 5 31.5 1 1 1 1 i 4 32.5 1 1 2 1 3 ! 8 34.5 1 2 3 35. 5 2 1 1 2 6 37.5 2 4 6 38.5 1 1 1 3 39.5 1 2 3 40. 5 1 1 2 4 42. 5 1 1 3 1 1 7 47.5 1 2 1 1 5 48.5 1 1 2 4 50.5 1 1 2 1 1 1 ! ^ TOTAL: 2 19 17 19 18 11 6 18 110 ..,.„.» -33-+ 4-+ •f 4-4= + + 4-+ 4-+ 4-+ 4-+ + + 4-4-4= 4-t 4-V 4-4-+ t-4-+ 4-+ 4-4-4-+ + + + + 4-4-, , , f 1 , , 1 ,— 10.D 1U.Q 20.0 25.0 53.0 Ifi.O 40.0 4r>.0 r.O 0 TRUE HEIGHT IM) Figure 5 Plot of the Residuals against True Total Heioht MeasuresAnt -24-Figurc 6 Relationship between True Height. and_thc Crew (Estimated) Total 11 e i q h t Me a s u r em e n t -35-A linear regression of the crew measurements (H ) on the control (true) values (H^) gave the following relationship: H„ - 3.358 + 0.8668 1i_ (10) r2 = 0.729, SEE - 5.62 The intercept of the regression line (9) was statistically different from zero at the a = 0.05 probability level. This indicated that on average, crews measured tree heights with a significant bias. The crew deviation in measuring height was 6.86 m the coefficient of variation being 21.86%. There was no trend of variation in precision of height measurements with increase in height (see fig. 6). -36-DISCUSSION It is only until recently that there has been more emphasis on the accuracy of individual tree measurement than on, say, the total number of trees measured per plot or the total number of plots measured per day. The analysis of errors associated with basal area per ha , diameter, and height measurements was, therefore, appropriate. The average tree count per plot is 9.5 trees with a BAF = 6 prism and 7.4 trees with a BAF = 9 relascope. The coefficient of variation of observer tree count is 10.44% with the prism and 4.93% with the relascope. This means that 68% of the tree counts fall wTithin 10.44% of the average prism plots count, or within 4.9 3% of the average relascope plots count. Since basal area per ha is a function of tree count per plot, conclusions reached regarding variability in tree count can be assumed to apply with reasonable consistency to basal area per ha. Therefore, the percentage error in the determination of basal area per ha from the 25 prism plots is ± 4.09% and from the 25 relascope plots is ± 1.93%, at the a = o.05 probability level. Percentage error is half the confidence interval of an estimate expressed as a percentage of the estimate. When the number of sample plots is 1, then coefficient of variation = percentage error. In this case, percentage error means that one is 55% sure that the average basal area per ha is within 4.09% of the true basal area per ha. estimate with the prism, or 1.93% of the true basal area per ha estimate with the relascope. -37-The prism crew coefficient of variation result is lower than, but close to the ± .11.15% with a BAF = 6.8 orism in similar stands found bv Munro (1966), and sliahtlv bigger than ± 6% with a BAF = 4.6 prism found by Carow and Rickerd (1969). Errors of this magnitude warrant checkcruising programs, especially if the number of sample plots established is low. The standard deviation in the crew tree count of 0.99 3 trees per plot using the prism is close to that of 0.805 trees per point obtained by Munro (1966), that of 1 tree per point obtained by Ker et_ al. (1957), and is within the ± 1 stem allowable error of the British Columbia Forest Service. The standard deviation obtained with the relascope is much lower, 0.366 trees per plot. This may be because the BAF used was lower, or because the relascope is a better cruising instrument, or both. It is difficult to pinpoint the exact reason for the low errors obtained using the relascope. The average stand coefficient of variation is 32.9% using the prism and 28.04% using the relascope. The coefficients of variation obtained by the control were 34.08% and 26.82% with the prism and relascope, respectively. This implies that lower tree counts per plot give a lower stand coefficient of varia tion. This is contrary to the findings of Sayn-Wittgenstein (1963). He indicated that the average coefficient of varia tion decreased with an increase in the average number of trees per plot. No plausible explanation can be offered for this difference. -38-The total average tree count per crew of 237.72 trees with the prism is higher than the true count of 222.0 trees. This means that the crews had an overall negative bias of 7.08%. Using the relascope, the total average tree count per crew was 185.8 trees. This is slightly higher than, but close to the true tree count of 185.0 trees. No reason for the high negative bias in prism count is apparent. There was no consistent relationship between the error committed and the number of trees per plot in either of the prism or relascope plots. The variability of the tree counts was due to some of the error sources listed in the introduction. However, one of the more recognisable sources of error is the subjectivity in locating the tree count center when checking the "doubtful" trees. The suggestion by Beers and Miller (1964) that an adjusting factor (related to tree diameter) be made such that distances are measured to the near face of the tree in question, reduces this subjectivity of locating tree heart centers. Another important error occurs when one is cruising for stand table construction or for thinning purposes. Although the tree count per plot of a crew is identical to the true value, it does not necessarily follow that the crews identified the same trees (in terms of dbh). Slope correction with a prism by swinging the prism about a perpendicular axis contributes many personal errors. It is difficult to maintain the angle of incidence within 2 degrees of the correct angle when hand holding an unmounted prism (Miller and Beers 19 75) . Also, slope correction with the relascope is not automatic; it is only done when the brake knob is released. A survey of the student instrument preferences indicates that 17 out of the 32 students asked preferred to use a prism for cruising. The main reason was that they had used a prism previously, but not a relascope. However, 13 out of the 32 students asked preferred the relascope, more because it is easier to correct for slope, and to align with the tree (see also a sample questionnaire in the appendix VI). Effectively, the relascope is a much better cruising instrument. This may in part explain the fact that smaller variabilities in crew tree counts were obtained with the relascope. The average tree diameter was 52.67 cm, and the coefficient of variation in crew diameter measurements was 8.16%. This error is much bigger than the British Columbia Forest Service allowable error of ± 1%. It is also higher than the error found by Myers (1961); but this is mainly because Myers' results were based on measurements on permanently labelled dbh points, and probably on smaller trees. There were more negative sources of error, indicated by a higher percentage of the measurements being too large. Other sources of such errors, other than those listed in the introduction were the following: i) not pulling the measuring tape tight ii) for large trees, there is the tendency to take readings to the nearest 10 cm (or at most, to the nearest 1 cm), and to approximate the diameters when the tree dbh exceeds the measuring tape length. Error source (ii) explains why accuracy is low at larger tree diameters. The average tree total height was '31.38 m, and the coefficient of variation in crew tree height measurements was 21.86%. This error is much higher than the 2.3% error found by Ker (1951), that of 3.5% reported by Meyer (1953), that of 2.4% found by Ker and Smith (1957), that of ± 5,10% noted by Young (1967), and the + 3% allowable error of the British Columbia Forest Service. Distance measurement errors con tributed quite significantly to the high error in tree height measurement. The probable reason for the unreasonably high error in height estimation is that the students had not had enough experience in height measurement. Furthermore, the students could not visualise well in the metric unit terms. Accurate tree height measurement requires more experience and greater patience. Forestry students are often employed to undertake forest cruising for various agencies. If the student inventory results are to be relied upon, errors of this magnitude must be watched out for, and reduced. On the other hand, it is probable that the current measuring procedures are not adequate, as hinted by the following quotation from Young (1967, p. 18) : "If wood were as valuable as gold, we would measure it with the same, accuracy, but as matters stand our methods reflect the quantity and value of standing trees and the primary cut products". -4 1-CONCLUSIOMS The following conclusions were made from the study. The average tree count was 9.5 trees in the prism plots and 7.4 trees in the relascope plots. The coefficient of variation of observer tree count is 10.44% with the prism and 4.93% with the relascope. Errors decrease with increase in BAF. A significant negative bias of 7.08% in tree count with a prism, and an insignificant negative bias of 0.43% with a relascope was observed. About 37% of the tree counts in the relascope plots, and 29% in the prism plots were correct. Ninety-five % of the relascope plots and 91% of the prism plots had errors of ± 4 trees or less. The maximum tree count error per plot was ± 6 trees. The percentage error in the deter mination of basal area per ha from 25 prism plots was ± 4.09%, and from 25 relascope plots, ± 1.93%, at the a = 0.05 pro bability level. These errors show that accuracy of a relascope or prism survey depends on the care with which the work is done, and that the relascope yields more precise results. The average coefficient of variation of the measurer in dbh measurement was 8.16% (the average dbh = 52. 67 cm) . About 95% of all the measurements of dbh were in error by ±5 cm or less. There were errors as large as ± 30 cm. Only 6% of the measurements were correct. Variability of the measurements about their mean was uniform irrespective of the dbh, but accuracy was lower at larger tree diameters. There were more negative sources of error than positive ones. -42-A coefficient of variation of 21.86% of the measurer in height measurement was obtained (average total tree height = 31.38 m). Over 15% of the height measurements were in error by ± 6.0 m. or more. Only 1.8% of the measurements were correct. Height measurements were subject to larger and more sources of errors. Errors of this magnitude in VRP tree counts, dbh and height measurement, clearly demonstrate the need for caution in using untrained crews in forest inventory work. Thex~e is obviously a need for rigorous field training programs and the establishment and implementation of checkcruising guide lines . -43-LITERATURE CITED Anderson, R.L. and T.A. Bancroft. 1952. Statistical Theory in Research. McGraw-Hill Book Company, Inc., New York. p.313-330. Beers, T.W. and C.I. Miller. 1964. Point sampling: Research Results, Theory and Applications. Purdue Univ. Agr. Exp. Sta. Research Bull., No. 786. Lafayette, Indiana. 56 p. Carow, J. 1958. Cruising Pole Timber by Bitterlich's Angle-Count Method. Papers of the Michigan Academy of Science, Arts, and Letters, Vol. XLIII: 151-156". Carow, J. and R. Rickerd. 196 9. Studies of Personal Bias in Bitterlich Cruising. Michigan Academician; 2(2): 67-71. Carron, L.T. 1968. An outline of Forest Mensuration with Special References to Australia. Australian National Univ. Press Canberra, p. 1-4. Cochran, W.G. 196 3. Sampling Techniques. John Wiley and Sons, New York. p. 355-393. F.A.O. 1973. Manual of Forest Inventory with Special Reference to Mixed Tropical Forests. FAO, Rome, Italy, p. 24-25. Ferguson, I.S. 1975. Measurement Bias in Plantation Inventory. Australian Forester, 38(2): 81-86. Ganguli, M. 1941. A Note on Nested Sampling. Sankhya, 5: 449-452. Holgate, P. 1967. The Angle-Count Method. Biometrika, 54(3): 615-623. Husch, B. 1955. Results of an Investigation of the Variable Plot Method of Cruising. J. For., 53: 570-574. Kendall, R.H. and L. Sayn-Wittgenstein 1959. An Evaluation of the Relascope. Canada Department of Forestry. Technical Note No. 77 Ottawa. 26p. Ker, J.W. 1951. A Test of Accuracy of Tree Height Measurement Taken by Abney Level and Chain. British Columbia Lumberman, 35 (1) : 58 . Ker, J.W. and J.H.G. Smith. 19 57. Sampling for Height-Diameter Relationship. J. For., 55: 205-207. Ker, J.W.; J.H.G. Smith; and J. Walters. 1957. Observations on the Accuracy and Utility of Plotless Cruising. British Columbia Lumberman, November Issue: 32-36. -4 4-Kirby, CL. 1965. Accuracy of Point Sampling in VJhite Spruce-Aspen Stands of Saskatchewan. J. For., 63: 924-92 Laar, A. Van. 1962. The Angle-Count Method. S. Afr. For. J., 72: 1-6. Loetsch, F.j 1?. Zohrer/nnd K.K. Hallet. , 1973. Forest Inventory, Vol. 11. BLV Veriagszesellschaft Munchem Bern. Wien. 469 p. Meyer, H.A. 1953. Forest Mensuration. Penns Valley Publishers, Inc., Pennsylvania, p. 101-103. Miller, C.I. and Beers, T.W. 1975. Thin Prisms as Angle Gauge in Forest Inventory. Purdue Univ. Agr. Exp. Sta. Research Bull. No. 929 Lafayette, Indiana. 8p. Munro, D.D. 196 6 . Some Personal Errors in Point Sampling-Forestry Chronicle, 42(4): 407-413. Myers, C.A. 1961. Variation in Measuring Diameter at Breast Height of Mature Ponderosa Pine. U.S. For. Ser. Rocky Mt. For. and Range Exp. Sta. Research Note, No. 6 7 3p. Sayn-Wittgenstein, L. 196 3. An Attempt to find the Best Basal Area Factor for Point-Sampling. Canada Department of Forestry, Inservice Report, Ottawa. 17p. Schmid, P.; Roiko-Jokela, P.; Mingard, P.; and M. Zobeiry. 1971. The Optimal Determination of the Volume of Standing Trees. Mitteilungen Der Forstlichen Bundes-Versuchsanstalt. Osterreichs, 91: 33-48. Stage, A.R. 1962. A Field Test of Point-Sample Cruising. U.S. For. Serv. Intermt. For. Range Exp. Sta. Research Paper No. 67. Ogden, Utah, 17p. U.B.C. Forest Club. 1971. Forestry Handbook for British Columbia 3rd Edition. 815p. Willingham, J.W. 1962. Error in Wedge Prism Calibration. J. For. February: 123-127. Young, H.E. 1967. Forest Measurement Accuracy. IN Wood Measurement Conference Proceedings, edited by F. Buckingham 'University of Toronto Technical Note Mo. 7: 17-23. -45-APPENDIX I : MAP 1. U.B.C Research Fore Pllt Loko AAaple Ridge, B.C. legend: conlours al 50' intorvols ((( main road v branch rood ' trail building ^ powerlinc* ~~* study arecj^ ^jl-. I S\ /Eunice loko / \ « .:• vK-'.V., •'X'X?'-:I'! <• i V'l^S-'v''/' //Gwendoline! , X' ) nor -47-APPENDIX II: TREE COUNTS IN . THE REL A SCOPE (BAF = 9) PLOTS LOT PLOT CREW NO. NC . NO. CONTROL 1 2 3 4 5 6 TREE COUNT 1 1 6 6 7 7 7 7 7 1 2 1 I 1 I 1 1 I I 3 7 7 8 8 8 8 8 1 4 8 6 7 7 7 8 7 1 5 9 9 9 9 9 9 9 2 1 10 11 11 13 12 14 2 2 7 8 8 7 8 a 2 3 6 6 6 6 9 2 4 9 12 11 9 12 2 «; 8 8 . a 9 9 3 I 11 10 6 11 11 11 11 3 2 7 7 4 7 7 7 7 3 3 6 6 a 6 6 6 8 3 4 9 8. 6 7 7 9 3 3 5 8 9 9 8 S 8 s 4 i 9 9 7 6 4 2 10 6 6 5 A 3 12 17 6 11 4 4 9 14 13 15 4 5 8 10 8 10 5 1 6 9 5 6 8 9 5 2 6 5 5 5 5 4 5 5 3 5 6 6 6 5 3 6 5 4 5 6 6 5 5 8 5 5 5 3 4 4 4 4 4 4 APP END TX -4 8-ITT: TREE COUNTS IN THE PRISM . (RAF=6) PLOTS LOT PLOT CRFW NOo NC. NO. CONTROL 12 3 4 5 6 ' TREE COUNT 11 5 4 4 5 54 5 12 4 3 3 54 4 4 1 3 11 13 16 16 12 16 16 1 * 11 11 12 12 12 12 155 366455 2 1 12 8 8 13 13 13 2 2 12 10 11 12 13 14 2 3 8 10 9 12 11 10 2 4 11 12 11 15 11 12 2 5 10 6 6 11 13 9 3 1 9 13 9 9 12 11 3 2 8 11 7 7 10 10 7 3 3 8 8 8 13 11 8 14 3 4 12 12 11 14 14 12 13 3 5 8 12 7 11 . 10 11 11 4 1 9 10 11 11 4 2 6 6 10 9 4 3 14 12 15 16 4 4 12 14 13 17 4 5 12 14 14 15 515 9 12 5585 5 2 8 li 11 11 11 6 11 5 3 14 14 13 14 12 12 15 54 4 9 4 4 4 3 4 55 4 6 6 3 53 4 - 4'9-APPENDIX IV: TREE DIAMETER MEASUREMENTS TREE NO. CREW NC. CONTROL 1 2 3 4 5 6 MEA SUREMENTS(CM) 1 5.9 12.9 6.0 6.0 5.9 6.3 6.0 2 6.1 5.6 5.0 3 7.5 7.5 7.6 4 8.3 8.6 8.7 8.6 9.0 8.8 8.7 5 9.3 9.2 9.2 9.4 9.2 9. 1 9.2 6 9.7 9.6 9.9 9. 8 9.7 9.8 7 9.8 10.8 10.8 8 10. 1 9.9 10.0 10. 1 9.9 10.0 9.9 9 10.5 io.e 1C.5 10.6 10.5 10.7 10 10.6 10.8 11 10. 8 10.8 10.9 10.7 10.9 12 11.4 15.3 11.4 11.2 11.0 13 13.0 13.9 13.5 13.5 13«5 13.3 14 13.5 14.9 14.9 14. 8 14.7 15.0 15 14.2 14.5 14.1 14.2 14.2 14.1 14.2 16 14.5 14.7 14.7 17 14.8 12.2 12.4 12.2 12.0 12.1 18 15.2 19.5 18.5 19 15.4 15.9 15.8 15. 8 15.7 15.2 15.7 2C 15.7 16.1 15.8 15.6 21 16.0 15.7 15.8 15.8 15.4 15.2 15.7 22 16.8 16.9 16.6 16.6 17.0 16.6 16.5 23 17.1 16 o 9 17.2 17.2 16.8 24 17.8 18.5 18.3 18.4 17.3 17.7 25 18. 1 17.9 18.0 £8.0 17.9 17.9 26 18.2 18*7 19.2 19.2 18.6 • 17.0 27 18.4 21.0 18.0 18.3 18.3 17.8 18.2 28 18 o & 18.1 18.1 18.5 17.9 18.5 18.1 29 18.8 18.3 18.4 18.3 18«4 18.3 30 18.9 19.0 19.0 19.0 19.0 18.8 18-9 31 19.2 22.0 19.0 19.2 18.0 17.9 32 19.2 19.1 19.1 33 19.6 19.7 19.4 20.6 19.8 19.3 19.4 34 19.9 20.4 19.2 20.5 19.8 19.7 20.6 35 20.4 20.1 19.7 19.4 20o3 20.0 20.7 36 20.5 20.3 20.5 21. 1 20.8 20.5 20.6 37 20.8 21.0 21.0 21.5 20.8 38 20.9 20.5 20.8 20. 5 39 21.2 21.8 .21.1 21.2 21.0 21.0 40 21.5 20.7 41 21.7 22.6 21.8 21.5 42 22.0 21.8 20.5 20.5 21.1 43 22.1 23. 1 21.9 22. 1 22.1 22.2 44 22,5 21.1 45 22.5 23.9 23.0 23.0 22.0 46 22.5 20.8 21.C 21. 1 22.0 47 23.8 27.4 32.0 27. 1 48 24.3 27.9 24.5 49 24.4 23.7 23.2 24. 0 23.9 32.3 50 24.6 24.8 25.2 24.0 24.0 -50-APPENDIX IV: TREE CI A METER MEASUREMENTS TREE NO. CREW NG. CONTROL 1 2 3 4 5 6 ME4SUREMENTSCCM) 51 24.9 24.6 24.8 24.6 52 25.0 25.0 24.8 24.9 24.0 25.0 25.1 53 25.2 24.8 24.8 25. 1 25.5 25.0 24.8 54 25.4 26.3 26.4 26.3 25.4 25.9 55 25.6 24.9 25. 1 24.9 25.0 56 26.0 26.2 27.5 26.0 25.4 57 26.0 26.3 27.4 26.1 26.0 27.1 26. C 58 26. 5 26.5 27.5 27.0 26.2 59 26.7 26.5 26.4 27.0 26.2 26.5 26.2 60 26.7 24.4 24.2 24.8 25.0 61 27.1 28.3 26.8 27.2 26.9 27.7 62 27.2 27.8 27.8 27.6 27.7 63 27.5 27.6 27.4 27.6 27.7 27.3 27.6 64 27.6 27.5 27.3 27.1 65 28.0 28.2 27.8 28.4 27.9 2 7.8 28 .3 66 28.1 27.6 67 28.4 27.0 26 22.4 68 28.5 28.0 28.8 69 28.6 28.2 29.0 29.2 28.5 28.4 28.6 70 28.7 31.7 31.5 30.7 71 29.0 29.0 29.1 29.5 72 29.0 27.6 28.3 28.2 21.6 29.1 73 29. 1 29.6 29.0 30.0 29.5 74 29.2 31.9 30.0 29.9 29. 1 29.9 75 29.4 30.7 29.1 29.1 30.1 29.1 29.7 76 29.6 29.7 29.5 29.6 77 29«,6 28.8 26.6 28.6 29.5 28o8 78 30. 0 28.6 30 = 3 30.4 30c0 29.6 30-.! 79 30.0 30.0 31.7 32.2 80 30. 1 34.0 35.3 35.2 34.7 81 30.1 29.2 29.1 29.6 29.4 30.0 82 30. 1 26. 1 32.7 83 30.3 30.9 84 30.4 30.5 30.5 29.5 -85 30.4 30.3 31.2 30.0 86 30.5 30.1 31.4 31.3 87 30.7 30.8 31.2 30.5 30.8 88 30.7 28. 1 30.0 89 31.2 31*3 29.0 32.7 90 31.2 31.2 30.8 35.9 31.3 91 31.3 31.a 31.5 31.9 92 31.4 31.7 31.8 31.8 31.0 31.7 93 31.4 31.7 31.8 31. 8 31.0 31.7 31.3 94 31.5 33.3 32.4 33.6 33.3 31.3 34.2 95 31.5 32.S 32.0 32. 1 32.8 32.0 32.7 96 31.5 30.9 31.3 31.5 31.5 97 31.5 32.9 32.9 31.8 98 31.5 31.2 31.5 30. 3 99 31.5 31.8 31.5 31.4 100 31.5 31.5 APPENDIX IV: TREE NO. CCNTRCl 101 32.0 102 32o0 103 32o0 104 32* 1 105 32. 1 106 32o2 107 3.2 © 2 1C8 32.3 109 32.5 no 32. 5 111 32.7 112 33.2 113 33.2 114 33.2 115 33.2 116 33.3 117 33.4 118 33.7 119 33. 8 120 33.9 121 34.0 122 34.0 123 34.3 124 34.5 125 34.8 126 35.0 127 35.3 128 35.9 129 36.0 130 36.0 131 36.0 132 36.0 133 36.2 134 36.2 135 36.2 . 136 36.2 137 36.3 138 36.5 139 36„5 140 37. 1 141 37.1 142 37. 1 143 37.2 144 37.2 14 5 37.3 146 37.3 147 37.3 148 37.4 149 37.5 150 37.5 -5 1-TREE CI AM ET ER MEASUREMENTS CREW NO. 1 2 3 4 5 6 ME A S LREMENTS( CM) 27.8 27.9 31.6 27.6 27.8 32.3 33.2 29.9 30.4 32.9 33.5 33.0 31.5 31.4 31.2 29.4 31.6 31.3 32.8 3 1 .9 32. 1 32.3 31.8 32 .2 31.1 31.4 31. 3 34.0 35.9 34 .7 32.2 20.7 32.5 32.8 33.2 33 .0 32.9 32.8 32.8 32.7 32. 5 32.3 33.3 32.2 32.3 32 .2 33.2 33.1 32.7 34.0 32.5 33.3 32.5 31.8 32.0 33.6 32.8 32.0 32.2 32.7 32.0 33.0 36.2 34.2 34.0 32.3 32.8 33.6 33.2 33.2 32.5 32.2 32.0 34.1 34.5 32.6 33. a 33.5 33.2 35.0 31.0 34.0 34.0 34.4 34.3 33.9 34.0 37.3 36.4 33.5 36.4 38.7 36.7 34.2 34.2 37.5 34.2 35.4 34.5 35.3 34.2 35.0 35 .3 32.7 33.8 35.4 33.9 34.0 34. 8 34.6 34.9 35.0 34.6 34 .7 35.5 35.5 35.9 36.8 35.9 35.5 35ol 35.5 35 .2 35.4 36.6 35.1 35.4 36.0 35.4 35.2 36.2 36.4 37.4 36.4 36.7 36.4 35.5 36.0 36.6 35.7 36 .2 36.3 35.9 36c3 35.0 35.1 36.0 36.2 36. 3 44.2 44.1 35.8 35.9 35.1 36.2 35.1 37.5 36.8 36.4 36.G 33.0 35.9 36„8 34.4 37.7 34.5 37. 2 36.8 36.9 36.9 36.4 36 o 8 32.8 32.1 31.8 42.9 42.3 37.2 38.0 37.5 37.6 37.7 37.1 36.8 37.2 38.8 37.2 37.9 36.8 3 8.7 39.7 37.7 38.0 37.5 37.1 37.3 39*0 37.0 38.8 37.0 37.0 35.8 37.0 38.8 38.5 37.2 37. 1 37.5 37.2 37.0 37.0 APPENDIX IV: TREE CIAMETER MEASUREMENTS TREE NO. CREW NC. CCNTROL 1 2 3 5 6 MEASUREMENTS(CM 1 151 37.5 42.3 37.5 47. 0 152 37.6 38.8 37.1 37.7 153 37.7 37.6 30.0 37.7 154 37.7 38.0 37.7 38.6 37.7 155 37.7 41„5 38.1 37.4 156 37.9 40.0 38.2 40. C 38.7 157 38. 1 46.G 4>5<,8 45.5 38.5 158 38.2 38.2 38.0 38.4 38.6 3 8.4 159 38.2 38.1 38.8 39„9 38.1 35.0 38ol 160 38. 3 39. 1 38o3 37. C 161 38.4 38.5 38.5 39.3 162 38o7 37.0 36.9 36.4 163 38.9 37o9 39.6 164 39.0 38.1 38.2 38.1 16 5 39.2 31.2 27.4 166 39.9 38.5 4Q.0 38.4 167 40.0 42.0 42.2 44.2 39.0 39.5 168 40.0 40.3 169 40.0 40.1 40.0 40.2 40.0 40.4 170 40.2 38.2 41.8 38.0 40o2 42.2 38.7 171 40.3 40.5 40.4 43.5 172 40.8 42.8 42o2 40.8 40.8 173 41.0 41.2 174 41.0 38.7 38.9 39.0 40.8 175 41.1 40.7 40.9 39.3 176 41. 2 41.5 41.2 40. 8 41.7. 41.0 41.1 177 41.2 43.7 41.0 41.9 42.2 41.9 41.9 178 41.3 41.4 41 • 6 4-0.5 179 41.5 41.7 42.5 42.5 41.2 41.3 44.2 180 42.0 38.7 42«5 42.0 181 42. 1 42.4 42.8 42*6 42*9 43.0 42.0 182 42.6 43.3 43.8 42.6 43.0 42.0 42.9 183 43.6 43.7 43.8 43.7 43„2 45.0 184 43.7 43.3 44.1 42.4 185 43. 8 46.5 43.0 44.0 56. 1 44.0 186 43.9 42.9 187 44.1 44.0 42.5 42.3 188 44.2 44.3 44.0 44.3 44.2 44.0 44.0 189 44.3 43.9 44.1 45. 1 44.3 44.3 45«8 190 44.5 44.6 191 44.8 44.5 44.6 45.4 192 44o9 43.6 44.0 43.0 46.8 193 45.0 A4.5 45.2 45.0 45.0 45.4 45.5 194 45.1 45.0 45.3 45.3 195 45.7 50.7 51.0 53. 2 48. i 5 0.7 48.1 196 45.7 44.6 44.3 44.3 45.1 44.3 45.2 197 45.8 45.5 19 8 45.9 47.7 46.0 46.7 45.2 46.5 46.4 199 46.0 46.3 46.5 46.1 200 46.2 45.0 45.7 47.6 47.2 47.2 49.9 -53-APPENDIX IV: TREE DIAMETER MEASUREMENTS TREE NO CREW NC. CCNTRCL 1 2 3 4 C 6 MEA SUREMENTS(CM> 201 46,4 46.2 45.0 46.0 k 202 46.9 46.7 48.9 49.0 48.3 203 46o9 46.6 47.2 46.8 204 47.1 48.2 45.5 46. 5 47.8 46.3 46 .9 205 47.1 47. 1 47.5 47.5 46.6 47.0 206 47.8 47.9 48.3 48.2 48.0 4 7.9 207 47.9 48. 1 48.3 49. C 49.7 208 47.9 55.2 56.1 49.0 49.0 51.0 50 .3 209 48.0 48.0 47.9 47.6 48-4 48.7 210 48.0 47.7 48.8 47.4 46.6 . 47.8 47 .6 211 48.1 44.8 212 48.4 48.8 47.9 48.0 48.5 213 48.5 48.0 48.2 47.9 48.7 49.0 214 48.5 50.2 48.7 48.9 48.5 215 48.8 45.7 46.3 45.8 216 48.9 50.8 50.4 50.7 51 .0 42.0 49 .2 217 48.9 49.8 49.5 49.6 49.6 44.2 218 49. 1 49.0 48.0 219 49.3 49.5 46 .9 47.0 48.5 46.0 49 .0 220 49.9 52.5 51.8 46. 0 49.7 221 49.9 50.1 50.2 51.1 222 49.9 53.7 53.0 223 50. 0 48. 5 51.0 50. 0 50.7 50.8 50 .4 22 4 50.0 49.0 50.6 51.0 50.7 50.0 49 .7 225 50.0 50.8 226 50.1 49.0 49.0 49.6 50.4 22 7 50.1 49,8 46 .8 51.0 50.9 50.9 228 50.4 50.2 50.9 52. 1 51.0 51.0 52 .8 229 50.5 51.1 51.3 51.0 50.3 51.3 55 .5 230 50.8 51.5 49.9 50.6 231 51.0 50.2 51.0 51. 3 50.4 50.8 232 51.5 52.5 54.3 53.4 53.4 52.5 51 .8 233 51.5 51.0 51.6 234 51.7 ' 45.9 44.3 50. 8 235 51.7 44.0 47.6 49.6 236 51.8 52.3 53.2 51.3 237 51.9 50,3 49.7 51.2 50.2 51.2 238 52.1 46.3 47.1 50 o 9 53.0 47.4 46 .5 239 52.5 52.0 52.8 52.5 240 52.9 53.5 52.5 53.0 52.8 52.7 241 52.9 53.4 53.0 53.4 53.6 242 53. 1 52.5 53.6 53.7 52.6 54.1 24 3 53.1 53.7 57 .I 54.4 54.5 54.5 244 53.1 53.5 53.1 55. 2 53.5 54.0 54 .1 245 53.1 50.9 51.6 49.4 246 53.4 53.4 53.3 57.2 53.0 52.1 53 .3 247 53.6 53.5 53.7 55.8 24 8 54.0 52.7 52.7 55.2 53.7 50.9 249 54.0 54.5 53.0 52.2 250 54.2 53.0 50.1 52. I 54.3 50.3 51 .3 i6„5 -5 4-APPENDIX IV: TREE DIAMETER MEASUREMENTS TREE NO, CREW NC CONTROL 1 2 3 4 . 5 6 MEASUREMENTS(CM) 251 54.5 52.2 53.1 53.4 52.8 53.0 252 54„6 54.8 55.6 55.2 55.0 54.8 253 54.8 54.0 60.1 54.0 254 55.0 54.3 53.9 55.0 54.6 255 55o4 55.2 54.9 55..1 55.2 55.3 55.2 256 55o4 56.5 56.6 54. 3 257 55o5 64.5 56.1 54.9 258 55.5 54.3 54.5 55.5 55.7 54.2 259 55.6 55.5 56.7 56. 8 56.6 56. 1 260 55o6 55.7 55.1 55.1 55.8 55.7 56.2 261 56o0 58.5 58.2 58.5 58.7 56.5 262 56. 1 52. 1 57.7 56. C 57.0 56.5 53.7 263 56.5 53.1 51.2 54.6 55,5 264 56.6 56.1 55.8 55.6 60.9 60.0 59.5 26 5 56.9 55.6 56.1 57.C 57.2 56.1 57.6 266 57.0 50.8 58.7 59.8 55.8 57.2 267 57.1 51.5 50.0 57.8 52.7 53 = 2 268 57o6 55.4 57 o 2 57.4 57.9 59.0 56.4 269 57o9 63.0 43.6 60.8 58.7 59.7 58.7 270 57.9 57.5 57.5 56.4 57.2 271 58. 0 56*8 57.5 57.8 57.8 272 58.2 58.5 58.8 50.1 57.2 59.3 273 58.6 58.1 58.0 59.3 58.3 56.2 58.3 274 58.8 56.9 59.2 58.0 60.1 59.2 58.4 275 59.0 58.4 58.5 57.9 276 59.0 58.8 60.7 60.2 277 59.9 58.4 60.1 61.7 60.9 61.3 278 59.9 61.4 58.8 59.0 61.2 279 60.0 59.3 5^.2 60. 1 60.8 58.0 59.3 280 60.0 59.9 60.3 59.5 281 60. 1 62.5 77.3 78.5 282 60.4 60.2 60.5 64.2 61.5 59.3 61.8 283 61.0 69.7 60.4 60.0 61.7 61.4 284 61.3 ' 61.G 62.0 61.2 61.5 63.2 235 61.3 61.3 60.9 62.1 60.5 60.7 286 61.3 58.0 58.4 62.3 61.2 61,2 287 61.6 60.6 62.3 61.1 288 63.1 62.8 66.7 62.8 289 63.4 63.5 6 3.1 62. 0 64. 1 61.3 55.2 290 63*4 64.0 '63.5 291 63.5 59.1 59.6 62.0 60.8 61.0 60.1 292 64.0 67.7 63.9 64.2 63.8 63.9 64.8 293 64.0 66.2 66.5 66.7 64.5 294 64. 1 65.9 60.2 64.8 63.5 66.8 295 64.3 63. 1 64.0 64. 1 296 64.4 65.6 65.7 65.7 65.7 64.5 63.4 29 7 64.4 64.7 64.0 64. 5 64.9 59.8 65.1 298 64.5 58.0 64.5 64. 1 299 64.5 56.4 54.4 52.5 55.3 56.4 300 64.6 60.9 61.7 66. 6 64.5 64.5 63.2 55.2 57.2 73.0 - 5 5-APPENDIX IV: TREE CIAMETER MEASUREMENTS TREE NO CREW NG. CONTROL I 2 3 4 5 6 MEASUREMENTS(CM) 301 65.0 57.3 52.4 57.0 54.8 302 65.6 72.2 70. 1 70.0 71.1 70.1 62.7 303 65.8 60.5 60.3 58.1 53.5 61.0 65.S 304 67.0 65.6 65.3 68. 1 76.3 61.3 67.2 305 67.2 65.0 60.3 66.2 65.4 64.3 65 .4 306 67.6 68.7 67.3 72.0 307 67.7 67.8 68.2 67.0 66. 5 68.2 67.0 308 67.9 64.2 309 68.0 72.3 65.0 66.9 66.9 310 68. 1 75.0 69.9 68. 5 69.9 72.6 311 68.4 67.4 62.5 66.0 65 .9 66.1 312 69.2 66.2 313 69.5 72.0 62.2 71.9 314 69.9 75.4 78.4 80.8 69 = 2 315 70.0 67.8 69.0 70. 0 69. 1 316 71.0 73.3 75.3 71.0 73.2 317 71.1 66. 1 67.9 66.8 70.0 6 8.6 318 71.5 73.5 73.6 71.9 319 72.0 73.0 73.0 73.8 71.0 72.5 320 72. 1 78.0 75. 1 71.9 74.8 72.5 321 72.6 74.7 76.5 74. 1 74.5 76.2 322 72.6 73.3 72.9 73. 1 73.5 323 72.7 66.4 66.7 68.8 67.4 72.8 324 73. 1 73.0 73.5 74,0 73.4 325 74.0 79.7 77.5 79.2 73.7 326 74. 1 82.2 77.0 77.0 '74.1 77.0 72.9 32? 74.5 74.0 76.2 76.0 74.5 76.0 73.0 328 75.1 74.6 76.1 7 5. 5 75.3 76.1 74.5 329 75.5 74.® 72.3 72.5 82.7 72.5 70.5 33 0 75.8 80.0 75.3 80.4 331 76.0 73.3 78.5 74.2 332 76.1 75.2 75.1 76.0 76.7 70.0 333 77.0 . 75.0 75.0 334 77.1 75.4 69.5 335 77.1 75.1 76.0 75. 1 76.8 76.6 336 77.1 75.2 75.0 75.2 75.9 76.2 337 77.4 72.0 77.8 76.2 76.8 76.6 338 77.4 76.0 76.2 75.7 339 77.7 80.2 78.5 72.5 340 79.3 81.6 81.7 81.5 81.4 79.0 341 79.5 80.7 81.5 78.4 79.0 78.6 342 82.1 79.8 89.6 76.4 88.9 81.0 34 3 82.2 78*7 80*7 S2*-0 344 82.6 82.0 83.2 82*7 82.7 82.5 82*4 345 83. 0 85.4 84.3 85.3 83.2 346 83.2 83.4 84.2 86.1 86.0 82.5 98.4 347 84.2 84.8 84.0 84. 1 85.5 85.5 34 8 84. 7 84. 5 85.0 85.2 86.4 85.5 85.1 349 85.2 84. 1 84.8 86.0 86.3 84.8 86.6 350 85.5 82. 1 82.5 86. C 83.4 73.9 85.6 APPENDIX IV: -56-TREE CIAMETER MEASUREMENTS TREE NO. CREW NO. CONTROL 12 3 4 5 6 7 MEASUREMENTS(CM > 351 87,0 85.2 85.2 86.7 87.0 352 89.0 90.6 88.0 353 89.4 90.5 89.2 89. 1 89.9 354 89.7 90.0 92.0 89.8 355 89.8 90.4 98.4 90.3 90.6 95.0 356 90.1 91.0 92.9 92. 8 88.4 357 92. 1 98.7 91.0 90. 8 96.2 358 92.4 93.5 93.5 105.0 359 92.7 91.1 91.0 91. 5 92.0 360 93.5 93.0 92.7 95.6 98.2 92.1 96.6 361 94.0 127.2 94.5 96.2 93.6 362 95.4 92. 6 90.0 90. 1 93.3 92.3 363 97. 1 94.0 97.3 364 98.0 93.2 96.8 94.5 365 100.0 99.7 110.9 366 100.1 107.1 107.8 103.2 96.7 367 100.1 96.5 100.2 96. 0 368 100.7 100.7 101.0 103 . 0 101.1 101,0 101 .1 369 102.9 112.2 104.5 113.9 104.0 370 103.0 107.3 107.7 101.8 105.6 102.0 371 103.1 101.7 100.0 103.5 102.5 372 103.3 104.5 106.5 105.3 373 105. 0 103. 1 103.8 102.0 101.7 374 106.9 104.0 108.0 105.5 106.8 110.2 105.1 37 5 110.0 108. 8 112.9 114.0 110.9 112.9 114.5 376 113o0 110.0 112.0 377 115.0 116.8 111.0 378 120.5 123.5 103.0 122.0 121.0 379 121.8 121.5 120.8 129.0 120.8 120.8 115.6 380 125.0 125.0 120.8 121.2 126.8 120.6 126,6 381 125. 1 129.0 127.6 127.5 124.0 382 128.2 125.0 129.0 124.5 383 130. 1 127.0 132.0 131.0 129.3 132o0 384 130.9 124,2 134.7 38 5 131. 1 131.5 127.0 131.1 132.0 133.3 126.0 386 132.5 131. 1 128.7 132.0 130.5 132.8 387 135.0 134.7 130.3 388 135.2 141.0 137.4 137P4 141.0 13 7.2 131,0 389 135.3 135.3 135.0 390 140.1 135.2 150.0 391 140.2 150.0 146.2 146.2 149.0 146.2 134.5 392 154.4 150.0 150.0 152. 8 159.2 151.0 154.2 393 160.0 166.0 164.8 160.0 394 160.0 133.5 132.9 126. 0 395 160.5 159.0 163.0 •»> *r o n. ii. J O v V 396 167.7 135.6 140.0 169.0 168.7 165.0 163.0 397 170.0 131.5 195.0 39 8 190.0 180.0 179.0 302.0 207.0 214.0 - D /-APPENDIX V: TR F E TOTAL HEIGHT MEASUREMENTS TREE NO. CREW NO. CONTROL 1. ? 3 4 5 6 MEASUREMENTS(IN M) 1 8.1 10.8 9.0 9.5 2 11.8 10.8 11.5 11.0 11.4 3 18.6 21.3 22.0 22.3 22.9 4 18.8 20.1 i8.2 17.9 17.5 17.6 28.0 5 21.5 13.2 19.9 6 22.1 22.0 27.0 22.2 7 24.6 30.3 32.1 21.0 8 26.5 23.8 28.0 24.2 26.3 24.0 28.4 9 27.4 26.0 24.2 25.0 30.7 10 27.7 34.8 23.6 32.4 11 29.8 29.4 26.8 33.0 33.3 12 30.0 30.5 27.3 23.2 13 30o5 28.0 28.4 14 31.2 31.2 32.5 26.7 28.7 15 32.3 26.6 21.0 30.5 16 32.4 31.1 30.4 28.4 28.5 17 32.5 48.5 53.8 31.5 18 34.8 49.0 19 35.0 49.0 30.3 30.2 26.2 20 35.7 29.5 41.0 29.0 21 35.8 31.2 32-0 32.2 22 37.4 37.3 38.0 23 3S.0 33.6 35.3 35.G 41.0 40.5 24 39.0 40.2 .35.0 25 40.0 35.0 37.5 -58-APPENOIX V: TR F E TOTAL HEIGHT MEASUREMENTS TREE NG. CREW NO 1 CONTROL 1 2 3 4 5 6 MEASUREMENTS (IN M). 26 40.2 3 3.8 24.0 3 5.5 27 42.5 44.1 39.8 42.9 40.2 28 42.6 33.0 45.0 38.0 2 9 4 0.0 4 1.5 42.0 5L.5 44.0 49.5 30 48.8 29.0 47.0 42.8 27.0 31 50.5 52.0 49.0 49.5 54.0 56.0 50.5 APPENDIX VI: INSTRUMENT USER PREFERE NCE SURVEY Faced with a choice between a prism and a relascope for determining fhe basal area per hectare of a forest, cruisers (for some reason) usually prefer to use one more than the other. This survey aims at establishing the preference of the final year forestry students, for the two instruments. Please mark with a " V^" your choice: 1. In the UBC Research Forest, which instrument would you rather use for determining the basal area per ha.? ( ) PRISM ( ) RELASCOPE ( ) BOTH. 2. What is the reason for your choice in (1) above? ( ) I AM USED TO THE INSTRUMENT ( ) THE INSTRUMENT IS EASIER TO CORRECT FOR SLOPE ( ) THERE ARE FEWER CASES OF "BORDERLINE" TREES ( ) THE INSTRUMENT IS EASIER TO ALIGN WITH THE TREE ( ) OTHER If OTHER please state: 3. For how many summers have you been cruising? ( ) ZERO ( ) ONE ( ) TWO ( ) THREE ( ) FOUR OR MORE If you checked other than zero in (3) above, did you use: i ( ) PRISM ( ) RELASCOi'E ( ') BOTH PLEASE RETURN THE COMPLETED FORM TO: STEPHEN OMULE, MacMillan Room 192. THANK YOU. SO/DDM/mpl Sept 1977 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0075436/manifest

Comment

Related Items