forestry bulletin no 14 practical point sampling
DESCRIPTION
ÂTRANSCRIPT
BULLETIN NO. 14 APRIL, 1967
PRACTICAL POINT-SAMPLING
ELLIS V. HUNT, JR.
and
ROBERT D. BAKER
School of Forestry
STEPHEN F. AUSTIN STATE COLLEGE
Nacogdoches, Texas
STEPHEN F. AUSTIN STATE COLLEGE
SCHOOL OF FORESTRY FACULTY
LAURENCE C. WALKER, Ph.D Dean and Professor
ARTHUR VERRALL, Ph.D Professor
NELSON T. SAMSON, Ph.D. Professor
M. VICTOR BILAN, D.F Professor
ROBERT D. BAKER, Ph.D Associate Professor
HARRY V. WIANT, JR., Ph.D Associate Professor
LEONARD BURKART, Ph.D. Associate Professor
ELLIS V. HUNT, JR., M.S Assistant Professor
EUGENE HASTINGS, Ph.D Assistant Professor
KENNETH G. WATTERSTON, Ph.D.. . Assistant Professor
J. ROBERT SINGER, Sc.D Assistant Professor
BILL H. WILPORD Visiting Professor
ROBERT S. MAXWELL, Ph.D Professor of ForestHistory
LOWELL K. HALLS, M.S. Part-Time Instructor,Forest Game Management
HENRY L. SHORT, Ph.D Part-Time Instructor,Forest Game Management
MARY Jo LINTHICUM, B.S Forestry Librarian
BULLETIN NO. 14 APRIL, 1967
118354
BULLETIN NO. 14 APRIL, 1967
PRACTICAL POINT-SAMPLING
ELLIS V. HUNT, JR.
and
ROBERT D. BAKER
School of Forestry
STEPHEN F. AUSTIN STATE COLLEGE
Nacogdoches, Texas
$1.00
PRACTICAL POINT-SAMPLING
FOREWORD
Point-sampling is a valuable tool in the kit of the practicing forester.It is employed for permanent and temporary sampling and for growthstudies. Since the concept of point-sampling is new in American forestry,different approaches have been employed to explain its application. Thisbulletin consists of a synthesis of point-sampling lectures and lessonplans developed at Stephen F. Austin State College and presented as aforestry short course in 1963. It explains these sampling techniques andsuggests choices in methodology and computational procedures. As a seriesof lectures, it was written in outline form, and this form was retainedbecause the organization and use of the ideas seemed to be facilitated inthis way. It is a revision of the first part of a two-part bulletin on point-sampling published by the School in 1964.
TABLE OF CONTENTS
Point-sampling theory 5
Calibrating an angle-gauge or prism 9
Constructing a stand-table 9
Computing board-foot and cubic-foot volumes 11
Sampling techniques 13
Constructing point-sample volume factors 19
Miscellaneous volume factors 28
The stand-table blow-up factor 29
Constructing a table of limiting distances 31
Slope corrections in point-sampling 34
Growth-projection with point samples 37
Number of points to achieve stated precision 37
Comparing methods of per acre point-sample volume estimates 39
Literature cited 42
Abbreviations 43
LIBRARYP. AUSTIN STATE COLLEGE
NACOGDOCHES., TEXAS
PRACTICAL POINT-SAMPLING
Ellis V. Hunt, Jr., and Robert D. Baker
POINT SAMPLING—THEORY
Mr. L. R. Grosenbaugh (1952) introduced point-sampling to Americanforesters, citing Mr. W. Bitterlich's earlier work in Europe. Primarily it is,he said, "A rapid method of sampling tree basal area per acre . . ." Itis rapid and in 1952 it was revolutionary! Since then, it has been explained,re-explained, embellished, tested and discussed. Here is how it is done andwhy it works.
I. Tally the selected trees at the point.A. With the angle-gauge, tally all trees appearing larger than the
angle.B. With the prism, tally all trees that are not wholly offset.C. Be sure to look at trees at dbh (diameter breast height, 4.5 feet
above ground line).D. Tally every borderline tree (those exactly coinciding with the
angle or those just exactly offset by the prism) as Vz tree.
OUT BORDERLINE
II. Multiply the number of trees tallied per point by the BAP (basalarea factor) that fits the instrument used, to obtain basal area peracre in square feet.A. For an angle-gauge describing 104.18 minutes, the BAP = 10.
1. A gauge 33 inches long with a cross arm 1 inch wide describesan angle of 104.18 minutes.
2. A 3.03 diopter prism does the same job.B. This is why it works—using a BAF of 10 as an example:
1. The area of a circle is 0.7854 D2.7rD2 3.1416 D2
a. Area of a circle = = = 0.785 D=.4 4
i Stephen F. Austin State College
2. The area of the "tree circle" is 662 X BA (basal area) of thetree.
a. The "tree circle," with the tree at its center, has a diameter66 X the diameter of the tree; it describes the boundarieswithin which one mu"t stand if he is to tally the tree withan angle gauge of 104.18 minutes or a 3.03 diopter prism.
b. These are examples of "tree circles" showing maximumdistance one may stand from a tree and still tally it (ra-dius of "tree circle").(1) For a tree of 1 foot dbh, one must stand not more
than 33 feet away.(2) For a tree of 2 feet dbh, stand not over 66 feet away.(3) For a tree of 3 feet dbh, stand not over 99 feet away.
III. Now consider a tree which is 0.5 foot dbh:
A. BA (basal area) of the tree is 0.7854 D° = 0.196 square feet.
B. One could count such a tree if he stood not farther than 16.5 feetaway from it.
C. The diameter of the "tree circle" for this tree is therefore 2 Xthe radius, or 33 feet.
D. The area of this "tree circle" is 0.7854 D2 or 0.7854 X 332 or0.7854 X 1089 or 855.3 square feet.
E. Statement II.B.2. said the area of the "tree circle" was exactlythe BA of the tree X 662. If this is true for this 6-inch dbh tree,then:1. BA of the tree = 0.196 square feet,
area of the "tree circle" = 855.3 square feet,66 squared = 4356.
2. 855.3/4356 = 0.19630.
IV. Then, for a tree that has a BA of 1 square foot (approximately13.55 inches dbh):
A. The "tree circle" has 1 X 662 square feet of area (as stated inII.B.2. and proved in IIT.E.) or 4356 X 1 square foot or 4356square feet.
B. Visualize this tree in the center of its "tree circle," the latterbeing in the center of an acre of land (Figure 1).1. Tree BA is 1 square foot.2. "Tree circle" area is 4356 square feet.3. Acre area is 43,560 square feet.
V. If a cruiser stands at random inside this acre, what chance wouldhe have of being inside the "tree circle" (i.e., close enough to thetree to count it with an angle gauge or prism defining an angle of104.18 minutes)?
Practical Point-Sampling
Figure 1. A one-square foot BA tree in its 4356-square-foot "tree circle"inside a circular acre of 43560 square feet.
A. The chances are equal to the proportion of the areas of the "treecircle" and the acre, or 4356 to 43560, or 1 to 10.
VI. If there were a single tree on each of 100 such acres and each treecontained 1 square foot of BA, and if the cruiser were on each acreat random, he would tally 10 trees in 100 chances.
VII. Change tally to total BA of trees by multiplying the tally by 10,because 10 X 10 = 100, and on 100 acres there would be 100 squarefeet of BA.
VIII. Compute BA per acre from such a tally by multiplying the tallyby 10 and dividing by the number of "points" the cruiser visited, or10 X 10 -f- 100 = 1 BA per acre.
IX. Conclusion: The BAF of an angle gauge or prism denning an angleof 104.18 minutes is 10, and the formula for computing BA in squarefeet per acre is:
Stephen F. Austin State College
BA per acre =SX
N(BAF)
where X = number of trees tallied,N = number of points, and
BAF = basal area factor of the instrument.
Figure 2. 10-BAF—Illustration of 3.03 diopter count. A 9.55-inch treeof BA = 0.5 square feet has "tree circle of (0.5) (662) = 2178 square feet
2178 1and a random tally chance in 1 acre of = or 1 in 20. A 13.55-
43560 20inch dbh tree has a BA = 1.0 square foot, a "tree circle" of (1) (662) =4356 square feet and a random chance of 1 in 10. The "tree circle" overlaparea, because of the position of trees, is 871.5 square feet with a randomchance of 1 in 50. In 100 random counts, one would get:
2 overlaps of count 2 each total 48 other big tree counts of 1 total 83 other little tree counts of 1 total 3
total 15 in 100 random-chances.SX 15
BA per acre is (BAF) or X 10 or 1.5 square feet per acre.N 100
Practical Point-Sampling 9
CALIBRATING AN ANGLE-GAUGE OR PRISM
I. As explained earlier, the BAP is the reciprocal of the random chanceone has of being in the "tree circle" of a single tree containing 1 squarefoot BA, if this tree is in a circular acre, and one is taking a point-sample in this acre.
II. One can find the radius of the "tree circle" by sighting at the treewith the instrument from such a distance that the angle just coincideswith the tree's diameter at breast height (with a prism, just soopposite sides of the tree coincide). High accuracy prisms, which aremore expensive, are calibrated at the factory. Uncalibrated prisms,available at about $1.00 each, must be calibrated by the user.
III. Procedure
A. Set up a rectangular target of a convenient width, such as 1 foot.
B. Lay a steel tape along the center line, perpendicular to the target.
C. Move back along this line until one side of the target image seenthrough the prism is precisely aligned with the other side seen overthe prism. Make sure the axis of the prism is perpendicular to theline of sight.
D. Measure accurately the distance from prism to target.
E. Compute the BAF using the following formula:43,560
BAF =l+4(d/w)2
where w = target width in feet,d = distance from prism to target in feet.
IV. As an alternate procedure, Dixon (1958) suggests using the sameformula but measuring "the amount of deflection at a fixed distance:
A. Fix prism in position at one end of a long table.
B. At the other end, at a right angle to face of the prism, set up atarget of two vertical pins that can be moved.
C. Measure accurately (d) the distance between the target and prism.D. Move pins until the displaced image of one pin coincides exactly
with the image of the other. Measure the distance between thepins (w) precisely."
E. BAF "calculated using formula above. It is essential that d r.nd wbe measured in the same units."
CONSTRUCTING A STAND TABLE
I. The probability of a tree dbh class being sampled is in direct ratio toits BA.
A. Consider a tree with BA of 0.5 square foot in the center of a cir-cular acre.
10 Stephen F. Austin State College
1. Its "tree circle" contains 662 times as much area, or 2172 squarefeet, or 1/20 acre if an angle of 104.18 minutes is used.
2. Its chance of being tallied in any one sample is therefore 1 in 20.
B. It was stated earlier that under the same circumstances a tree of1 square foot BA had 1 in 10 chances of being tallied. Thus thechance of a tree being in a sample is exactly proportional toits BA.
II. To construct a stand table, it is necessary to measure the dbh oftallied trees.
A. Formulating a conventional stand table most simply requires twosteps.1. Convert the tree count to BA in each dbh class per acre.2. Then divide each of these BA's by the BA of the tree at nominal
class center.
B. As an example, if an angle-gauge having 104.18 minutes is used,the following trees are tallied at 5 points:
1. Point Tally by DBH-—Inches1 6, 12, 16, 202 10, 18, 14, 6, 16, 10, 8, 8, 12, 143 10, 2, 20, 18, 12, 184 16, 10, 10, 12, 12, 10, 14, 14, 14, 85 14, 16, 18, 16, 8, 10, 12, 12, 10, 4
2. The total tally is 40 trees at 5 points.
C. In our example, from Grosenbaugh (1952), we have two 6-inch,one 4-inch and one 2-inch trees.1. To convert these to BA per acre, multiply by BAF and divide by
the number of sample points.2. To convert to trees per acre, divide this (Item 2) by BA of the
the tree of nominal class midpoint (Item 3).
Item 2-Inch Class 4-Inch Class 6-Inch Class1 . Count 1 1 22 . S q . f t . B A p e r acre 2 2 43. Sq. ft. BA per tree .022 .087 .1964. Trees per acre 90.90 22.98 20.40
D. The entire stand table could be constructed in this manner fromtree dbh's recorded for the trees counted as samples at the points.
III. In summary, to get frequency per acre (f), multiply the tree count bythe BAF, divide the product by N (number of points), and divide this
quotient by BA of a tree at nominal midpoint diameter (0.00545415D2). Putting this in equation form, we get:
BAFCount X
Nf = , or
0.00545415 D2
Practical Point-Sampling 11
BAP
NX Count,
0.00545415 D2
BAF
Nso factor is
0.00545415 D2
COMPUTING BOARD FOOT AND CUBIC FOOT VOLUMES
Note: The first step (as for any volume estimating) is to secure an appro-priate volume table.
I. A suitable standard volume table in cubic feet or board feet may bemodified (for instance, a Girard Form Class Volume Table).
A. To modify the table, divide each tree volume in the table by its treeBA. Enter these quotients, called bff [board foot factor] or cff[cubic foot factor], in a new table by tree dbh and height class.1. This table may be used as follows:
a. Tally trees in each point sample by dbh and height.b. Secure the sum of the bff or cff for the trees in the sample.c. Divide this sum by the number of trees.d. Multiply this quotient by BA per acre when:
no. treesBA per acre = - - X BAF.
no. points
B. To further modify the tables of factors (described in I.A.) divideeach value by 10 to reduce the size of the numbers for easy mentalarithmetic, and round off all values to whole numbers.1. This will not introduce any serious errors where samples of
reasonable size are to be collected to compute volume per acre.2. This table may be used exactly as outlined before except that
the quotient (I.A.l.d.) will need to be multiplied by 10 beforemultiplying by BA per acre.
II. A different kind of table may be similarly produced giving a bff orcff for height classes only, instead of dbh and height classes.
A. This is true because the height-tree-volume relationship is prac-tically a. straight line.1. Thus for Mesavage & Girard Form Class 80, Grosenbaugh (1952)
showed these bff's divided by 10 (International %-inch log rule):
DBH ILog 2 Log 3 Log 4 Log10 7 12 — —12 8 12 16 —14 8 13 17 2016 8 14 18 2218 8 14 19 2320 8 14 20 23
12 Stephen F. Austin State College
B. From the table, it can be seen, for instance, that if 2-log treesaverage 14 inches dbh, the Bff is 13.
C. Using the table, the procedure for determining volume per acre is:
1. Tally point sample trees by height classes.
2. Total the trees in each height class.3. Multiply each sum by the appropriate Bff.
4. Total these products.5. Divide this total by the number of points in the sample.6. Multiply by BAF to get board feet per acre.
(Note that since Bff was obtained by dividing tree volume byBA X 10, it will be necessary as a final step to multiply finalanswer by 10.)(Note, too, that the procedure above is simplified, but likeI.A.I., where:
2 bff no. treesVolume = X X BAF,
no. trees no. pointshere:
2 BffVolume = - - X BAF X 10.
no. points
III. If no suitable volume table is available the bff or cff may be determinedby felling timber in each height class or, if preferred, each dbh-heightclass, and measuring it for volume.
A. The following procedure is used:
1. Fell the sample trees.
2. Secure suitable tree measures (usually d.i.b. at some regularlength interval, and dbh).
3. Apply a suitable log rule to obtain log volume.
4. Total to obtain tree volume.5. Calculate the average tree volume and basal area for each dbh-
height class, or each height class.
6. Divide tree volume by tree BA, or usually in case of Bff divideby 10 times BA.
IV. Appropriate factors for southern pine were suggested by Grosenbaugh(1962ft).
A. Bff's based on Mesavage and Girard Form Class 80 (about averagefor reasonably well-stocked second-growth southern pine) :
BffInternational
Merch. Lengthzero log
12345
B. The cff s for reasonably stocked southern pine:
International% Inch0713182328
Scribner0611162025
Doyle048121521
Practical Point-Sampling 13
Merch. Length, cff Merch. Length, cfffeet feet
0 0 40 2610 7 50 3120 14 60 3630 20 70 39
V. A formula suggested by Grosenbaugh (1955) for International boardfoot volume per acre in southern pine is:
(SL) (60) (BAF)Board foot volume per acre = ,
Nwhere L is merch. height of trees in 16-foot logs,
N is number of sample points.A. The factor 60 is based on average southern pine and allows for an
increase in tree dbh and form class with tree heights.
SAMPLING TECHNIQUES
I. Point-sampling is tree sampling proportional to tree BA. This wasshown in the section concerning construction of a stand table, wherethe equation for computing the stand table blow-up factor was:
BAF
NFactor =
0.00545415 D2
A. The factors for three common values of BAF are:1
DBH ClassInches
468
1012141618202224
Stand5-BAF
57.3025.4714.329.166.374.683.582.832.291.891.59
Table Blow-up10-BAF
114.5950.9328.6618.3312.739.357.165.664.583.793.18
Factors20-BAF229.20101.86
57.3036.6725.4718.7114.3211.329.167.586.37
1Some values in the tabulation are slightly different from those whichwould be computed from values in Table 6, due to rounding off.
B. These are easily computed using the equation shown as, for in-stance, the 10-BAF and dbh class of 4 inches:
Factor =
10
1
(0.00545415) (42)
14 Stephen F, Austin State College
10
(0.00545415) (16)
10
0.0872664
= 114.5916.
C. Hence, the smaller the dbh class, the smaller the percent of sample,and therefore the larger the multiplier must be to convert toper-acre values. This is an advantage over some other samplingmethods since the bulk of counting or measuring labor is usedon the more valuable, larger dbh classes.
II. Several other kinds of sampling are proportional to area, as for in-stance plot-sampling.
A. In plot-sampling, if one wanted to change count of sample trees ona 1/5-acre plot to trees per acre, the same multiplier (5) would beused for all dbh classes.
B. Generally, the same principle applies to line sampling and tran-sect-sampling. Counts along a line or transect are related to thecount for an area, and all dbh class counts are multiplied by aconstant to obtain per acre counts.
C. Actually, as the typical uneven-aged stand table clearly shows,there are far more small trees than large trees in the forest, soa cruiser spends more time than warranted, by volume or value,in measuring or counting them. (For instance, on 110 circular1/5-acre CFI plots established in Sabine County by the Schoolof Forestry, over half of the stems came from the 5- and 6-inchdiameter classes.)
D. The same considerations would be valid for a mixture of smalleven-aged stands. One would expect to stratify and sample dif-ferently in even-aged stands large enough to be treated separately.
III. Some problems perhaps best solved by a sampling method otherthan point-sampling are:
A. To determine survival of seedlings in plantings, sample-row count-ing is practicable.
B. In natural reproduction inventories, random quadrat counts ortransect or line counts may be more reasonable.
C. For forage volume inventories, quadrat clippings may be desired.
D. When a count is not directly related to area, transect or line countsmay suffice.
IV. There are many considerations involved when trying to devise a point-sampling scheme. Some of the most obvious are discussed, but manyothers are encountered in the field. One must be alert to note themand to make field time as productive as possible.
Practical Point-Sampling 15
A. Random sampling- principles should be used whenever reasonablebecause any measure which is averaged will be of limited valueif not accompanied by a measure of variability or accuracy.
B. Systematic samples may be subjected to analyses suitable for ran-dom samples but the results must be accepted with reservation.Nevertheless, the systematic forest sample will generally be moreprecise than other kinds of samples of the same size, so systematicsampling does have some advantage.
C. Two-prism sampling may be advantageous when two arbitrarilydefined populations occupy the same area.1. A sawtimber population and a pulpwood population are present.
To count the sawtimber trees with a 5-BAF prism would be tootime consuming, but to count pulpwood size trees with a 10-BAFprism might admit too few trees for an adequate sample.
2. The obvious solution is to sample each arbitrarily defined popu-lation with the instrument best suited to measure it.
D. Combined plot and point-sampling may be suitable for some situa-tions similar to that discussed above, or perhaps combined pointand transect or line-sampling would be better.1. For timber volume and reproduction counts, one might use
point-samples for volume, and stocked quadrats for reproduc-tion.
2. For timber volume and thinning prescription, one could userow tree counts for prescribing, and point-sampling for timbervolume estimating.
E. Subsampling (to save time) may be desirable in situations whensome attribute with small variance is needed in combination withsome other of larger variance, or where economics may dictatefewer measurements of a difficult nature.1. For estimating volume in pulpwood size plantations, use point-
sampling to count trees, and randomly select trees out of thepoint-sample count for height measures (height is a time-con-suming measure, and all the merchantable trees may be nearlythe same height).
2. The same kind of solution might apply if one were sampling toconstruct a stand table or to obtain volume when volume fac-tors are based on dbh.
3. Particularly, in IV.E.2., the random, selection of individualsample trees for the measured subsample would be important(avoid the idea of measuring trees you select on the premisethat a set number should be measured in each class).
4. Because some size class may be limited by the nature of thestand (40-inch dbh trees for example), perhaps all individualsof some classes should be measured in the subsample. Whenthis is done, computational procedures must take this into ac-count.
F. Proportional sampling might be used in areas where stands can bestratified.
16 Stephen F. Austin State College
1. Separate hardwood from pine stands or young timber from oldtimber, etc.; then sample each stand.
2. This may involve not only variance of the populations, but alsototal stand area, value per unit, etc.
G. Cluster sampling may be better than the selection of a single sam-pling unit at each location.
1. In 1957-58, an area was sampled using the average of two 10-BAF samples as a measure of BA per acre with the followingresults, compared to 1/5-acre plots also measured in 1957-58and a subsequent single 10-BAF sample in 1962 at the samelocation:
Sampling Unit
Avg. of 2 prism samples1957-58
1/5-acre plot 1957-58One prism sample 19621/5-acre plot 1962
BA per A(Sq.Ft.)
70.1870.7262.8770.99
am.(Sq.Ft.)
2.172.202.802.44
C(Per Cent)
30.9231.1144.5336.00
2. The sampling scheme referred to above contained 110 samplingunits, so a am. of about ±10% could have been obtained in the1957-58 inventory with only about 10.52 or 11 points, but usingonly one prism count, as the sampling unit in 1962, would haverequired about 21 points for similar precision. (Note: Theprobability is 68 in 100 chances.)
3. Also, the 1957-58 point-sampling was as precise as 1/5-acreplot-sampling, showing that on this area it is just as efficientfor BA data as the larger plot-sample.
4. Point-sampling is now employed in the Southern Forest Surveywhere 10 points, each a chain apart on varying azimuths, areused with limiting distances for a 37.5-BAF prism.
H. Prescription sampling with the prism may be done advantageously.
1. Grosenbaugh (1955) described a method of prescribing andmarking thinnings in southern forests. With a 50-BAF prism,the count should never be 3 (over 2) when the "leave" BAtarget is about 83 square feet.
2. Similarly, where one-fourth or more points contain zero poten-tial crop trees with 50-BAF prism, the area needs special atten-tion for restocking or increasing stocking.
3. On the Texas National Forests, a 37.5-BAF prism is similarlyused in sampling during pre-sale planning.
4. Seed tree marking, or checking of marked stands, may wellbe accomplished after marking standards have been formulated.For instance, if the goal is 7 large (average 19-20-inch dbh)trees per acre to be left as seed trees, then BA per acre of seedtrees would be about 15 square feet; so with a 5-BAF prism,seed tree counts should average 3.
Practical Point-Sampling 17
V. The choice of a point-sampling instrument (choice of a BAF) mightmake point-sampling somewhat comparable to plot-sampling in theconcept of percent of sample.
A. This concept is disclaimed by authorities contacted by the authorsand by published material on point-sampling or forest samplingbecause sampling varies with BA.
B. Nevertheless, a case example from a Texas industrial forest ispresented here because it may have value for some situations:
1. Trees on company land average slightly over 12 inches dbh, andwith a 10-BAF prism, this size tree has a tree circle (or plot)of about 0.08 acre; thus plot size averages 0.08 acre.
2. 500 points will be located on 200,000 acres.3. Equivalent acres tallied: (500) (0.08) = 40.00 acres.4. Percentage cruise: 100 X 40.00/200,000 = 0.2 percent.
VI. Bias in point-sampling is as important a consideration as in anysampling. It is to be avoided.
A. Systematic instead of random samples may cause bias.
1. As a result of the coincidence of the natural population dis-tribution and sample point locations, all points could fall inhigh site-class areas, thus omitting ridge tops and dry sites.
2. Bias could happen as a result of man's activities. By randomselection of a starting point for a systematic cruise, one mightselect a spot 2 chains southeast of the northwest corner ofeach 160-acre square block of forest land, not knowing that 50years ago a steam skidder yard was located in each of thesespots, making each such spot non-representative of the wholeforest.
B. Improper use of tools causes bias.1. If the prism is hand-held and by some quirk always 4-5 degrees
out of horizontal (which would not be compensating), the countwill be in error.
2. If the cruiser swings the prism around him instead of viceversa, the count will be wrong.
3. If one measures distance from the point to the nearest edgeof a borderline tree instead of to center of tree, the count willbe incorrect. (Note if one-half dbh is subtracted, then nearestedge of tree would be correct.)
C. Ignoring slope correction may be acceptable for some southern for-ests; for many in East Texas it would not.
D. Borderline trees, if not properly handled, may be a source of bias.1. As a result of erroneous reading or writing of a rule, bias
may be introduced.a. The rule should be: "Include border trees on the west side,
exclude those on the east side" or, rephrased, "Count everyborderline tree as % tree."
18 Stephen F. Austin State College
b. The rule should not be: "At each point, count the first bor-derline tree, omit the second, count the third, etc." (Becauseborderline trees will be about 1 in 20, and because ordinarilyonly one would be found on 2 or 3 plots, all would getcounted.)
2. If borderline trees are not measured to accurately determinewhether in or out, an unconscious bias may develop whereinmany "out" trees are counted, or vice versa. This may not becompensating.
3. Looking at trees always at eye level or some place other than4.5 feet above average ground line would not be compensating.
E. "Brush bias" (trees behind brush or other trees) may cause countto be low, but never high.1. Proper use of the prism and proper selection of BAP can
usually overcome this in southern forests (a problem in "BigThicket" country).
2. In the Pacific Northwest this bias is controlled by looking ateach tree at top of first 16-foot log instead of breast height(Bruce 1961).
VII. Point-sampling statistical analysis is similar to that for any sam-pling, particularly where one is working with relatively large samplesand a continuous variable. Actually, tree counts, as in BA estimationwith point-samples, are discrete; but this should not be a barrier tothe use of normal statistical analyses for continuous variables, sincebasal area per acre is a continuous variable.
Statistical analysis uses variance as measured in a sample. If X isthe measured variable, then its value must be ascertained for eachsample point. The variance is computed with the equation:
NV =
N — 1where V = variance of X,
X = measurement at a point,N = number of points.
Other statistical values normally needed from point sample cruisesare:
A. Standard deviation = a = V V.
B. Standard error of the meanir(100)
C. Coefficient of variation = C =
Jl= am = \.
M2 am (100)
D. Proportional limit of error = Pie. —M
Practical Point-Sampling 19
E. Number of observations for 68 percent confidence in <rm = N =V a2
Number of observations for 95 percent confidence in am = N =4V
cF. Sampling error of mean in percent = S = —
NG. Combined sampling error in percent = Se%, where
Se% = VlSTof bff (or'cff))2 + (S of tree count)2'.
CONSTRUCTING POINT-SAMPLE VOLUME FACTORS
I. There are many ways of computing timber volumes from point-sam-ples. Five are illustrated:
A. The first is similar to the standard volume table approach in plotcruising, sample trees being classified by height and diameter.
B. The second is based on local volume tables as used in conventionalplot cruising, sample trees being classified by diameter only.
C. The next is like the local volume table approach except that sampletrees are classified by height rather than dbh, with a volume factorfor each height class.
D. The fourth method is like I.C. except that a single average factoris used for all height classes.
E. The last method involves converting trees tallied in the field intovolumes per acre in a cumulative tally.
II. The system wherein sample trees are classified by height and dbh issimple and direct.
A. A good standard volume table is needed. To illustrate the proce-dure, a volume table for an area on the Sabine National Forestfor loblolly pine is shown in Table 1.
B. Each tree volume in the table is divided by the BA of the dbh classto which it belongs.1. For 10-inch dbh, 1 log, the volume is 23 bd. ft., and
23bff =
(3.1416) (10)V144(4)— 23/0.545,= 42.
2. As a further refinement, each of the basal area factors thusderived is commonly divided by 10 to reduce the size of thenumbers involved; thus
bff
10= 42/10,= 4.2 which is frequently rounded off to 4.
20 Stephen F. Austin State College
C. Table 2 was made in that manner from Table 1.
TABLE 1. LOBLOLLY PINE GROSS BOARD FOOT VOLUMES(SCRIBNER LOG RULE), SABINE NATIONAL FOREST.
Form Class
DBH Class10111213141516171819202122232425
74
12330384756667688101114128142157174191208
76Height in
1.5314152657793110127144163182203225250274302
77Logs
239526583101119138162186212238266293327362396
79
2.5506888112136163190222254291327368409456503550
80
3537493119144174205239273312351394437488537592
82
3.5
106137166201236279322365410461511570630693
TABLE 2. LOBLOLLY PINE BOARD FOOT FACTORS BY DBH ANDHEIGHT CLASSES, SABINE NATIONAL FOREST.
DBH Class10111213141516171819202122232425
14.24.54.85.15.25.45.45.65.75.85.95.95.96.06.16.1
1.55.76.26.67.07.27.67.98.18.18.38.38.48.58.78.78.9
Height in2
7.27.98.39.09.49.79.9
10.310.510.810.911.111.111.311.511.6
Logs2.59.2
10.311.212.112.713.313.614.114.414.815.015.315.515.816.016.1
39.7
11.211.812.913.514.214.715.215.415.816.116.416.616.917.117.4
3.5
13.514.915.516.416.917.718.218.518.819.219.819.820.120.3
D. A three-step procedure then determines volume per acre frompoint-samples.
1. Tally sample trees by dbh and height at the sample points.
2. For each tree tallied, extract the appropriate Bff from thetable.
3. Sum the Bff's, divide by the number of points, and multiplyby 10 X BAF to obtain board foot volume per acre.
Practical Point-Sampling
E. A suitable tally sheet is shown in Figure 3.
Stephen F. Austin State College
Point-Sample Tally Sheet
21
School of Forestry
DBH
Class
10
II
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Lmft
No. of
1
Points
T
1.5
TYF
ree Height
2
10
in 16 Foot L
2.5
Estimator _
ogs
3
Date
3.5
Figure 3. A tally sheet for use when all point-samples in a line or unit aretallied together as a single unit.
III. The system wherein the sample trees are tallied by dbh class onlysimplifies field work.
A. A local volume table is used without modification.
B. Stand blowup factors, one for each dbh class, are computed asfollows:
BAFStand blowup factor = • ,
BA per treewhere: BA per tree is in square feet,
and BAF is determined by the prism.C. For the 5-inch class and for a 3.03 diopter prism, for instance:
Stand blowup factor = 10/0.1364,= 73.3
22 Stephen F. Austin State College
D. With the local volume table and the stand blowup factor, to getvolume per acre in each dbh class, multiply sample tree count bystand blowup factor, then multiply that product by volume pertree (Figure 4).
S. F. Austin State Col lege
TALLY SHEET
School of Forestry
FOR POINT SAMPLES ONLY
Llr.e
Poi nt
Type
Est ima tor
DiameterClass
F ie ldTa 1 1 y
5&7S91C1?516182022•2i42628JO
Treesper
Point
StandTab le
BlowupFactor
Treesper
Acre
CubicFoot
Vo 1 LrneB ! owu p
Short leaf Pir.e.
73.370.9
37.1*S3. 622,6
Hi2t ^ o V
9. It7.15.71*.6•;.e3.2?./gi3?.G
l.irfl&^7u.286,09s,te
BeardFoot
Vo ! umeB lowup
S38li.5
11*0.7?0%l281.5353.6Ii8.7
5UB.O66ii.8790.8922.0
CubicFoot
Volumeper
BoardFoot
Vo lumeper
Loblo l ly Pins
5o789
1012m161820222u26
£S^0
Total
73.350.937. i*?8.622.6
17.312.79.1*7.15.71*.63.83.22.72.32.0
l.t*32.6?1*.366.208.69
1*9.382.0
I30jt720^ Ji280.138£*.91*57.7559.0678.1808.691*0.1*
Sawtimber Volume by Scribner Log Rule
Figure 4. A tally sheet for use when point-sample tree dbh only will bemeasured and local volume table employed.
IV. When sample trees are classified by height only, field time is saved.
Practical Point-Sampling 23
A. To develop the factors needed for this point-sampling, start witha curve of dbh over height and a standard volume table as, forinstance, Table 1 from which Table 2 could be developed.
B. Then, for each height class from the curve, determine averagedbh, and from Table 2 extract the proper Bff for that height class.
C. If a suitable volume table is not available, derive the factors frombasic data.
D. The necessary tree measurements by height classes will be:
1. Tree volume, and2. Tree dbh.
Note: To show the procedure, Table 3 presents some data fromthe Sabine National Forest.
E. Compute a bff for each height class by dividing average volumeby average BA in each height class (Table 3).
TABLE 3. LOBLOLLY PINE SAWTIMBER TREE MEASUREMENTSFROM SABINE NATIONAL FOREST, 1958 (300 trees)
Ht. Class(feet)12161820222426283032343638404244
Trees( number)
1
3654
1299
192320132313
710
AVK. dbh(inches)
10.39.8
10.710.911.111.411.011.511.911.513.213.512.814.614.215.1
AVK. Vol.(bd. ft. Scrib.)
36.437.946.651.355.861.060.369,478.174.9
107.2115.8110.0149.1143.0176.8
Ht. Class(feet)
464850525456586062646668707276
Total
Trees( number )
211612138
12984455231
300
Avg. dbh Avg. Vol.(inches) (bd. ft. Scrib.)
16.316.816.417.316.716.717.517.720.417.618.117.618.120.119.2
220.0240.1231.3273.0258.9270.6308.4323.8456.5335.9377.7354.6384.6506.8477.9
F. Make a curve of bff over height. This may be done graphically ormathematically. From Table 3 the equation for this line wasy = 15.2 + 2.9x (When y was bff and x was height class).
G. For each height class, take from the curve (or compute by theequation of the curve) the bff.1. The equation from Table 3 gave these:
Ht. Logs~0
0.511.522.5
bff
386285
108131
Ht. Logs33.544.555.5
bff154178201224247270
24 Stephen F. Austin State College
H. To use these in point-sampling, tally sample trees by height classesand determine the sum of bff by multiplying the number of treesin each class by the appropriate bff.
I. Then compute volume per acre with the equation:
2 bffVol., bd.ft. per acre = (BAF).
no. points
J. Table 4 shows similar basic data and Table 5 the cff's from cubicfoot volumes from the Sabine National Forest.
V. The system where one average factor is used to compute volume inboard feet when sample trees are classified by height only is explainedby Grosenbaugh (1955). Hunt (1961) said that the bff (a numbernormally near 60 for average southern pine timber volumes based onInternational J/4-inch log rule) could be derived from tree height andstock and stand tables usually available.
A. When the bff has been obtained, board foot volume may be computedby the equation:
(ZL)(BAF)(a number near 60)Vol., bd. ft. per acre =
Nwhere N is the number of points, and
2L is the sum of sample tree heights in 16-foot logs atthose points.
B. A suitable field form when using a factor like this is shown inFigure 5, where the bff is 55 for hardwood and 60 for pine, andthe local cord volume factors are given.
Note: Throughout this discussion, bff and board foot volumes generallyhave been used, but cff's are similarly computed.
VI. A cumulative tally sheet using adapted values obtained from a standardvolume table may be preferred in converting point-sample measure-ments into per acre volumes rather than to compute special volumefactors as in II-V, above.
A. The procedure to convert volume per tree into volume per acre:
1. Determine the volume of each tree from a standard volume table.For instance, a 12-inch, 3-log loblolly pine tree in the SabineNational Forest has a gross volume of 93 board feet, ScribnerLog Rule (Table 1).
2. Multiply volume per tree by stand table blowup factor for thediameter class. A 12-inch tree has a factor of 12.7 (Figure 4):93 board feet X 12.7 = 1181.1 board feet per acre.
3. Round this value to the nearest hundred board feet per acre andenter in the cumulative tally sheet (Figure 6). The entries are12, 24, and 35, respectively under 12 inches, 3 logs.
4. In like manner, cubic-foot volume per acre can be employed inthe cumulative tally (Figure 6).
Practical Point-Sampling 25
TABLE 4. LOBLOLLY PINE PULPWOOD TREE MEASUREMENTSFROM SABINE NATIONAL FOREST (437 trees).
Ht. Class(feet)
8101214161820222426
Trees( number )
15275453544630342328
Avg. dbh(inches)
4.95.15.45.95.96.56.87.17.57.3
Avg. Vol(cu. ft.)
.951.181.552.062.393.163.794.405.385.55
Ht. Clai(feet)
28303234363840424448
is Trees(number)
721208733211
Avg. dbh(inches)
8.27.98.28.28.59.29.38.49.19.3
Avg. Vol.(cu. ft.)
7.337.318.368.799.95
11.9812.8411.4913.6215.32
TABLE 5. LOBLOLLY PINE CUBIC FOOT VOLUME FACTORS FORPOINT-SAMPLING, FROM THE EQUATION—CFF = 1.98 + 0.65 H.
Merch. Ht.(feet)
08
12162024
Cu. Ft. Factor
7.189.78
12.3814.9817.58
Merch. Ht.(feet)
283236404448
Cu. Ft. Factor
20.1822.7825.3827.9830.5833.18
26 Stephen F. Austin State College
3, Fo Aus t i n S ta te Co l lege SchooI of Forestry
POINT SAMPLE TALLY SHEET
SIO ID
Figure 5. A tally sheet for use when point-sample tree heights are talliedand a single factor for each product is applied in computingvolume per acre.
Practical Point-Sampling 27
CUMULATIVE TAL
Line Point
R u n n i n q o n
S. F. A. FORESTRY CAMP.Y — 10-BAF POINTS — HUNDRED BOARD FEET PER ACRE (GROSS), SCRIBNER RULE
-Type
Location
Date
E st imator
O Jcr •
0 0LJ! Ol
to to0 —
to N:C?i OCD NJ
28\jt oJ
C^ 0
-J -.
ol^01 CD
Ol OOJ toto OJ
Oi N;jj Vj.
Ol OCO -v,VO 0
— cto 01
o —
&crx
Ol
Oi
CO
VO
o
ffl
I"?
<t>COft)Q.
^ X=T O— Q.
O OOl
0 PT
/\: -o— 0O (ft
o~I O
ye
CD CO
Qi 0>O 0)^~ -f
o enc c3 3
n n> B
ee
ch
Iozr0)-1
o~1
Z3a.—ioQ)
Su
bto
tal
j
OJo
Oi
OJ
vOtoOi
VO
VO
toCO
-j
to
£Oi
-F*
toVO
OJ
-J
SOito
VO
CO
toro
Oi
OJ
vO
N)Ol
VO
CO
to~g
• j
toto•F*
Ci
-pi.
CO
V)
•-J
OJ-pi
Ol
VO
-J
Hardwoods
toOl
ON
OJ
^O
toOl
vO
CO
^j
Oi
to-£•
-F*
CO
OJ
^J
OJ
Oio
CO
-J
£c\
VO
toOl
lO
CO
0\
Ol
£
03
\
^*J
UJ
010
CD
>f
toto
01
to
O
MOl
VO
-g
Ol
toV.M
V.TI
^
--J
-
O\
•F*.
CO
VJt
8
Ol
NJ
•X
\
10
^-J
un
oNJ
fr
LfJ
0\U
Ol
-p>.
^J
$•
CO
to
CO
"O
CO
•vj
4^-
•o
^J
W
Ol
CO
Ul
o*»
Cn
LH
Oi
\
*^J
\
CO
Ol
N5fi
to
otoo
to
•f*
Oi
^
CD
^
IJ1
S
-fi
Ol
•vj
to
CO
IJ1
toLr4
0
0
too
£>
—
to
*UJ
OJ
c^Jt
f*\)
CO
M
O
Ul
to0
^J
OJ
8•-J
*o
CO
too\
ID
o
-
^w
to
^"i0
°TOo
IT
o
crn
CrIB
Su
bto
tal
L ob lo l ly Pine
wo
Cf\
CO
to
«o
CO
to-~j
en
X
-^J
LHOJ
0
CD
cnOi
\
-J*CO
to
fh
toj
\jtro*O
OlCO
i j
0.K)
MCO
Oi
totoCO
M
\
CO
to-^j
01
toLH
Ol
U4OJ
\
CO
Ol
Ol
to
•t*
to
.
toOl
NJtoCO
OlCh
L^O
S
K>c\i
to
CO
to
\Q
CD
tosj
Ol
a
01
UJto
CO
--J
Ol
Ol
too*h
to
fk
toOi
-uCO
OlOl
K
Oi
to
CO
to
o
CO
toOl
01
toL»
Oi
L"lto
CO
-J
LJ1
8f*oo
^
toOi
---J
01A
toto
Oi
to
CO
to
10
^J
toOi
j£
toto
Oi
o^
-vj
^J
>l
Ul0
10UJ^D
N)
4^
£
"£>
Oi
Ol
8Ol
to
CO
to
00
•*jtoOl
IrJ
toto
Ol
OJo
Ol
Oi
to-P»CO
wCO
to
#»
toL»J
•vj
VJOi
^>VO
CO
Ol
to
^jto
CO
Oi
toO!
LH
to
f"
^UJ
en
^>
CO
o*^J
K£*
Oi
m
—
Ol
to
CO
0>
toL>J
O
8
UJ
to• j
0
8•f*-
^jOJps.
O
VjJ
.
Ol
—
Oi
to
•^j
•f*
toto
vD
VO
Ol
toO>
VjJCO
--J.£•
o\J
-^J
vx
to
Ol
o
f».
•o
UJ
8Oi
Oi
toto
s
-F>-
OJ
OJ
to^J
Ol
OJ
o
t*
CO
ro
Ol
Oi
^
>o
^j\Oi
CO
NJ
Short 1 ea f
Su
bto
tal
ujo
Ch
>otoOi
O
CO
to~j
Oi
to^
toCO
-^oj
Ul
to0
VO
to
toOi
to
CO
OlOl
toCO
o>
otoOi
<o
CO
toa\o
^t
•oCD
to
^
>o
^o
i;CD
Ol
Ol
too\l
•0toOi
vO
CO
to-J
toL*
S
toCO
-£-to
£
vO
10
roUJ
•j
U1
Ul
KOi
^0
toOi
VO
CD
to-j
Oi
totoOJ
Ol
toCO
*>to
•Pi
VD
CD
tow
-J
UlOlo
toto
o\
toOi
VO
CO
toOi
toUJ
S
\•*j•f*.
UJ
VO
CO
V)^
Ol
**
CO
Pi ne
8
o-
VO
toOi
"O
^J
toOi
toto
to•w4*0
,\
CO
Oi
S
M\
UJ
\>_*J
>Ol
CO
--J
8tooj
vO
-Jtoo>
toto
toOi
UlVO
to
CD
U-JOl
VO
Oi
o>
UD
S
^
-J
to01
\
M
toOi
LMCO
0
•J
Jj-fs.
CO
4^-
-J
tooto^J
VO
-vj
8
.\
\
to
^UJ• 1
VO
Oi
wM
-»J
Y
to
L^
b
CO
toOl
0
MO
OJ0
toN->>*>i
VjJIJ
,\
q
oV>Jo
o
o
v>
M0
N)
CO
Ol
X
>jJNJ
VO
CO
MCO
o
0
L*
NJ-P>
L»J\T
0COX
Ol
to
toOl
L*
W
Ol
-F*
-P>
Oi
Ol
'Ol
Oi
— 10
o>
Figure 6. Cumulative tally sheet for point-samples.
28 Stephen F. Austin State College
MISCELLANEOUS VOLUME FACTORS
I. Using shortleaf pine tree measurement averages from the Sabine Na-tional Forest, and the procedure illustrated for board feet by Hunt(1961), a cff of 0.7 was computed for use as follows:
(sum tree merch. ht. in feet) (0.7) (BAF)Cubic feet per acre =
number of sample pointsII. Grosenbaugh (1955) offers the following equation for computing south-
ern pine cord volume per acre, using an instrument with BAF = 10:
Volume per acrein roughstacked sum of total tree height in feet of all sample treescords =
(20) (number of sample points)A. If a local factor other than 20 is desired, it can be calculated and
is usually between 18 and 21.
B. The procedure:1. Select several small representative areas.2. On each, measure to determine individual tree BA and total
height in feet.3. Determine sum of (BA X height) for each area.4. Cut the trees into pulpwood; stack and measure cords for each
area.5. Compute the factor for each area with the equation:
sum of [(BA) (height)]Factor =
sum of volume (cords)
III. Grosenbaugh (1955) rewrites his equation so that it may be used forany instrument with a known BAF in sampling southern pines:
Volume per acrein rough (sum total tree hts.) (0.005) (BAF)stacked cords —
number of sample pointsA. This is the same equation as in II, with the BAF inserted and the
values rearranged.
B. If a locally more accurate factor than 0.005 is desired, it may becomputed in the same manner as described for the 20-factor above,except that the right side of the equation is inverted to read:
sum of volume (cords)Factor =
sum of [(BA)(hts.)]IV. Using the procedure in III., and values from yield tables, classes at
Stephen F. Austin State College devised factors for three southernpine species in various localities:
Virginia pine in east Tennessee 0.0047,Longleaf pine in Georgia 0.00524,Shortleaf pine plantation of low form
class in east Tennessee 0.0045.
Practical Point-Sampling 29
A. Note that some of these are appreciably different from 0.005, andtherefore a locally derived factor may be of some importance.
V. If we rewrite Grosenbaugh's equation in II., to include the BAF, itbecomes:
Volume per acrein rough (no. trees) (BAF) (avg. total tree ht.)stacked cords = -
(10) (20) (no. sample points)
A. Rearranged, this equation could be shown thus:
Volume per acrein rough (no. trees) (BAF) avg. total tree ht.stacked cords = X .
100(no. sample points) 2i
no. treesB. But (BAF) = BA per acre,
no. pointsso, BA per acre in 100 square feet units X half the average totaltree height = volume per acre in rough stacked cords.
THE STAND TABLE BLOW-UP FACTOR
I. The stand table blow-up factor (S.T.F.) gives the stand table valueswhen multiplied by point-sample tree counts tallied by dbh class at onepoint.
A. Therefore, for each dbh class:Trees per acre = (S.T.F.) (point-sample tree count).
B. A stand table may be made in this manner.
C. In addition, if the count came from N points, the product is dividedby N.
II. The S.T.F. for each dbh class and BAF can be computed by the equa-tion:
BAFS.T.F. =
BA per tree
A. It would be easier to compute for any particular BAF if multiplica-tion by the constant BAF could be performed.
B. This can be accomplished if the equation is changed to:S.T.F. = (BAF) (reciprocal of BA per tree)
C. Table 6 is a list of these reciprocals; the values may be used in theabove equation to produce the appropriate S.T.F.'s for any BAF.
30 Stephen F. Austin State College
&HoEH
H2
KO
Wrj
^-AHH
.
.
tnpq^-v
HK
£"
HK
O"w•s« '
•W
j02
P3
o021-5
<^OoPH
HH
OaPH
50
HJpq
EH
.
C/2HKo£-\H
OS:CO
OH
s^HHH
rH
sPife
HffiOHKH02
KPQ
PHHHWa
OS
00
t-;
CO
1O
T*
CO
cxj
1 — 1
0
ffiC
Tf LO CO CD OSiH CO CO CO LOCO 00 CO COC~ t- O CO COO rH CO t- LO10 CO rH
CO LO Tt* Tf COc- •** o LO osOS CO OS LO ^fTH CO CO OS -^CO CO CO t- LOLO CO rH
rH CO OS CO COrH LO CO CO COOS CO CO OS -^CO rH CO CO CD
CO LO CO OO LOCO CO rH
CO CO 00 LO O00 O -^ LO COCXI O Tt CO -^• rH rH CD COrH t- T^ OO LOt- CO rH
CO LO iH O COO LO O 00 OO CO t- LO COCO CO Oi O O
-H O Oi CO00 CXi iH
OS rH OS t~ COL— t- t- OS OSLO •<* -* to co^ CO 00 rt COCO rH LO OS COOS CO iH
CD -H O CO LO CO LO O t>os o co co coco to oo os LO00 ^F CO OS COO CO THTH
CO CO rH O t~CX] 00 C75 LO OS00 t- 00 OS t-LO CO 00 CO t-
CO C- C- O COCO CO rH rH
1—1
CO CO O rH COLO CO "" LO t—rH O1 OO O -^LO O OS OiH rH OS O C-LO ^f TH rHiH
co to co co -^00 rH CO rHrH t- CO LO COOO CO CO -^ CO
rH LO O iH t-00 C<1 rHrH
TH CO CO -^ LO
CO t- 00 CO COO t- Tt O COLO CO rH t- TfCO OS CO CO LOCO C<1 CXI rH rH
rH OS rH COLO CO t- OS rHCO rH CO O t-OS O CO OS LOCO CO CO rH rH
O rH LO CO LOLO CXi CO CO rH00 OS CO OO O Tt OS CO• CO CO. rH iH
00 CO OS CO COOO -^ GO OS rHO t- t- CO COCO rH OS CO"tf CO CO rH TH
CO LO H/ t- THO OS t- rH CO-^ LO CO CO COCO CO LO O CD
•^ CO CO CO TH
CO 00 00 rH COCO L- CO LO LOL- -^ OS t- OS•^ CO LO O CD• CO CO CO rH
OS O t- O COCO O rH O COrH -^ CO CO COCO "tf CO rH t-Tt1 CO CO CXI rH
t- CO O Tf rHCO t- L— CO COCD co cxi CD CDtr- LO L^ rH t—
^ CO CXI CO rH
LO t- OS OS COCO L— CO t-CO CO OS iH OSOS CO t- CO t-• CO CO CO rH
CO rH LO LO LOTf T-H -rt CO COOS ^F CO CO COO C~ CO CO 00
LO CO CO CO rH
CO t- CO OS O
L- CO OS 00 COTt rH CO LO LOOS O T^ CO COCO r-i OS CO t-
TH TH O O O
OO rH t- O •<*CO OS CO t- TjirH TH CO CO COCO rH OS 00 0
rH TH O O O
TH CO CO LO COOS CO CD 00 COCO CO t- T}H TfCO rH OS 00 t-TH iH O O O
CO OS CO rH ^HCv] -^ rH O COCD LO OS CO LOCO iH OS OO t-rH TH O O O
-^ T^ O TH THCO CO tO CO CO00 t- O t- COCO TH O CO t-rH rH rH O O
CO LO rH CO rHO CO rH -^ COrH OS CO OO t--^ rH O 00 C-
rH rH rH O O
O CO LO CD COCO rH CD CO COCO rH CO Ol CO-^ CO O CO t-tH iH rH O O
CO OO CO CO COiH rH CO OS COCO CO LO O OSrf CO O O1 t-
rH rH rH O O
rH CO CO rH00 CO CO CO - tCO LO CO CO O"tf CO O OS CO
iH iH iH O O
CO <M CO LO OSLO CO -^ LO -^rH t- CO CO THLO CX] O OS CO
TH rH rH O O
TH CO CO -^ LO
OS CO CO O t-rH CO CO OO OS• t- iH CO THCD LO LO ^t
O O O O O
CO t- 00 t- 00OS OO CO O CO•<* t- rH CO COCO LO LO Tf ^t
o o o o o
•^ Cvl CO OSt- LO ttf CO t-LO CO CO t- COCD lO LO -^ -^
o o o o o
• OS O CO rHLO rH O t> COCD OS CO t- COCO LO LO H^ -^
o o o o o
• t- t- Cvl COCO CO LO CO CDt- OS CO CO COCO LO LO ^f
o o o o o
t- CD LO CO COTH LO TH t- G00 O -^ OO TfCD CD LO
O O O O O
rH CO LO CO OSO CO t- CX] OS i—t Tt< OS -^CD CD LO "* *rf
O O O O O
CO t- LO rf CO00 OS CO t- OSOS rH LO OS -^co CD LO -^ -^o o o o o
CO O t> CD 00t- t- OS CXI COO CO LO O LOt- CO LO LO -^
o o o o o
CO T? OS OS fCO -^ LO L- COrH CO CD O LOL^ CO LO LO
00000
CO t- CO Ci OCO
00 CO O t- COCO OS rH LO COCO -^ CO OS L—CO CO CO CO CO
o o o o o
GO t- tr- rH ^J*LO CO CO CO LOOO LO CO OS t-CO CO CO CO COo o o o o
^ 00 * LO CDOS LO CO O t-00 LO CO O L--CO CO CO CO CO
o o o o o
O O CO O COCO OS OS CO OSOS LO CO O t-CO CO CO CO COo o o o o
CD CO O ^H OCO Cvl CO LO COCTS CO CO O 00CO CO CO CO CO
o o o o o
• -^ oo o coO 1C TJH CO ^hO CD CO O COTf CO CO CO COo o o o o
rH t- t- lO -^T^ 00 t- O COO CO CO rH OO•sf CO CO CO COo o o o o
OS O CO rH I>t~ CO O CO COO t- rH 00Tf CO CO CO CM
o o o o o
00 CO t- O,H LO CO LO rHrH t— rH OS•^ CO CO CO CO
o o o o o
t- OO CO CO TflO OO CO CO COrH C^ rH OS• t CO CO CO Cxi
00000
rH CXI CO Tj< 10CX|
Practical Point-Sampling 31
CONSTRUCTING A TABLE OF "LIMITING DISTANCES"
I. A diopter is 1 centimeter of displacement of a light ray at a distanceof 100 centimeters from the optical lens—in our case a prism.
II. A wedge or triangular prism is a piece of optical glass which refractsa light beam the refraction (or deflection) being a factor of the rateof convergence of its non-parallel sides.
A. Therefore, with a 10-BAF prism, which is 3.03 diopters, a 12-inchdbh tree would be on the borderline at 33 feet:
12 X 100= 12 X 33 inches = 33 feet.
3.03
B. By like reasoning, a 12-inch dbh tree viewed through a 20-BAFprism, having- 4.29 diopters, would be borderline at 23.3 feet:
12 X 100= 12 X 23.3 inches = 23.3 feet.
4.29
C. For ease in computation, derive a multiplying factor for inches ofdbh to get the borderline, or limiting distance. For a 10-BAF prism,12 inches of dbh had 33 feet to its borderline and, therefore, the"tree circle" factor is 2.75:
33— = 2.75 the "tree circle" factor.12
D. The "tree circle" factor for a 20-BAF prism is 1.944:
23.3= 1.94
12
E. In like manner, by knowing the BAF of the prism and its diop-ter power, we can calculate the multiplying factor for dbh toconvert to feet for the borderline distance (Hovind and Rieck,1961).
III. Tables 7 through 10 give the borderline distances by tenth-inch dbhclasses for 5-, 10-, 20-, and 37.5-BAF prisms. Actually, by chainingto the center of each tree and measuring its diameter, point-samplingis performed without a prism.
Stephen F. Austin State College
TABLE 7. LIMITING DISTANCES FOR 5-PACTOR PRISMS-HORIZONTAL.
DBH .0 1 .2 .3 .4 .5 .6 .7 .8 .9FEET
567
89
10
1112131415
1617181920
2122232425
2627282930
19.4423.3327.2231.1135.0038.89
42.7846.6750.5654.4558.34
62.2266.1170.0073.8977.78
81.6785.5689.4593.3497.22
101.11105.00108.89112.78116.67
19.8323.27,3135,39,
43,47,5054.58,
62,66.70.7478.
82.85,
,72,61.50.39,28
.17
.06
.95
.83
.72
,61.50.39.28.17
.06,95
89.8493,97,
101,105,
,72.61
50,39
109.28113.117,
17.06
20.2224.1128.0031.8935.7839.67
43.5647.4551.3355.2259.11
63.0066.8970.7874.6778.56
82.4586.3490.2294.1198.00
101.89105.78109.67113.56117.45
20.6124.5028.3932.2836.1740.06
43.9547.8351.7255.6159.50
63.3967.2871.1775.0678.95
82.8486.7290.6194.5098.39
102.28106.17110.06113.95117.84
21.0024.8928.7832.6736.5640.45
44.3348.2252.1156.0059.89
63.7867.6771.5675.4579.34
83.2287.1191.0094.8998.78
102.67106.56110.45114.34118.23
21.3925.2829.1733.0636.9540.83
44.7248.6152.5056.3960.28
64.1768.0671.9575.8479.72
83.6187.5091.3995.2899.17
103.06106.95110.84114.73118.61
21.7825.6729.5633.4537.3341.22
45.1149.0052.8956.7860.67
64.5668.4572.3476.2280.11
84.0087.8991.7895.6799.56
103.45107.34111.23115.11119.00
22.1726.0629.9533.8337.7241.61
45.5049.3953.2857.1761.06
64.9568.8472.7276.6180.50
84.3988.2892.1796.0699.95
103.84107.73111.61115.50119.39
22.5626.4530.3334.2238.1142.00
45.8949.7853.6757.5661.45
65.3569.2273.1177.0080.89
84.7888.6792.5696.45
100.34
104.23108.11112.00115.89119.78
22.9526.8330.7234.6138.5042.39
46.2850.1754.0657.9561.84
65.7269.6173.5077.3981.28
85.1789.0692.9496.84
100.73
104.61108.50112.39116.28120.17
TABLE 8. LIMITING DISTANCES FOR 10-FACTOR PRISMS-HORIZONTAL.
DBH .0 .1 .2 .3 .4 .5 .6 .7 .8 .9FEET
56789
10
1112131415
1617181920
13.7516.5019.2522.0024.7527.50
30.2533.0035.7538.5041.25
44.0046.7549.5052.2555.00
14.0216.7719.5222.2725.0227.77
30.5233.2736.0238.7741.52
44.2747.0249.7752.5255.27
14.3017.0519.8022.5525.3028.05
30.8033.5536.3039.0541.80
44.5547.3050.0552.8055.55
14.5717.3220.0722.8225.5728.32
31.0733.8236.5739.3242.07
44.8247.5750.3253.0755.82
14.8517.6020.3523.1025.8528.60
31.3534.1036.8539.6042.35
45.1047.8550.6053.3556.10
15.1217.8720.6223.3726.1228.87
31.6234.3737.1239.8742.62
45.3748.1250.8753.6256.37
15.4018.1520.9023.6526.4029.15
31.9034.6537.4040.1542.90
45.6548.4051.1553.9056.65
15.6718.4221.1723.9226.6729.42
32.1734.9237.6740.4243.17
45.9248.6751.4254.1756.92
15.9518.7021.4524.2026.9529.70
32.4535.2037.9540.7043.65
46.2048.9551.7054.4557.20
16.2218.9721.7224.4727.2229.97
32.7235.4738.2240.9743.72
46.4749.2251.9754.7257.47
Practical Point-Sampling 33
2122232425
2627282930
57.7560.5063.2566.0068.75
71.5074.2577.0079.7582.50
58.0260.7763.5266.2769.02
71.7774.5277.2780.0282.77
58.3061.0563.8066.5569.30
72.0574.8077.5580.3083.05
58.5761.3264.0766.8269.57
72.3275.0777.8280.5783.32
58.8561.6064.3567.1069.85
72.6075.3578.1080.8583.60
59.1261.8764.6267.3770.12
72.8775.6278.3781.1283.87
59.4062.1564.9067.6570.40
73.1575.9078.6581.4084.15
59.6762.4265.1767.9270.67
73.4276.1778.9281.6784.42
59.9562.7065.4568.2070.95
73.7076.4579.2081.9584.70
60.2262.9765.7268.4771.22
73.9776.7279.4782.2284.97
TABLE 9. LIMITING DISTANCES FOR 20-FACTOR PRISMS-HORIZONTAL.
DBH .0 .1 .2 .3 .4 .5 .6 .7 .8 .9FEET
56789
10
1112131415
1617181920
2122232425
2627282930
9.7211.6613.6115.5617.5019.44
21.3923.3425.2827.2229.17
31.1133.0635.0036.9538.89
40.8342.7844.7246.6748.61
50.5652.5054.4556.3958.34
9.9211.8613.8115.7517.6919.64
21.5823.5325.4727.4229.36
31.3133.2535.2037.1439.08
41.0342.9744.9246.8648.81
50.7552.7054.6456.5858.53
10.1112.0614.0015.9417.8919.83
21.7823.7225.6727.6129.56
31.5033.4535.3937.3339.28
41.2243.1745.1147.0649.00
50.9552.8954.8356.7858.72
10.3012.2514.1916.1418.0820.03
21.9723.9225.8627.8129.75
31.7033.6435.5837.5239.47
41.4243.3645.3147.2849.20
51.1453.0855.0356.9758.92
10.5012.4414.3916.3318.2820.22
22.1724.1126.0628.0029.95
31.8933.8335.7837.7239.67
41.6143.5645.5047.4549.39
51.3353.2855.2257.1759.11
10.6912.6414.5816.5318.4720.42
22.3624.3126.2528.2030.14
32.0834.0335.9737.9239.86
41.8043.7545.7047.6449.58
51.5353.4755.4257.3659.31
10.8812.8314.7816.7218.6720.61
22.5624.5026.4528.3930.33
32.2834.2236.1738.1140.06
42.0043.9545.8947.8349.78
51.7253.6755.6157.5659.50
11.0813.0314.9716.9118.8620.81
22.7524.7026.6428.5830.53
32.4734.4236.3638.3140.25
42.2044.1446.0848.0349.97
51.9253.8655.8157.7559.70
11.2813.2215.1717.1119.0621.00
22.9524.8926.8328.7830.72
32.6734.6136.5638.5040.45
42.3944.3346.2848.2250.17
52.1154.0656.0057.9559.89
11.4713.4215.3617.3119.2521.19
23.1425.0827.0328.9730.92
32.8634.8136.7538.7040.64
42.5844.5346.4748.4250.36
52.3154.2556.2058.1460.09
34 Stephen F. Austin State College
TABLE 10. LIMITING DISTANCES FOR 37.5-FACTOR PRISMS-HORIZONTAL.
DBH .0 .1 .2 .3 .4 .5 .6 .7 .8 .9FEET
56789
10
1112131415
1617181920
2122232425
2627282930
7.138.519.97
11.3512.8014.19
15.6417.0318.4819.8721.32
22.7024.1625.5426.9928.38
29.8331.2232.6734.0635.51
36.8938.3539.7341.1842.57
7.268.65
10.1011.4812.9414.32
15.7717.1618.6120.0021.45
22.8424.2925.6727.1328.51
29.9631.4232.8034.2535.64
37.0938.4839.9341.3242.77
7.398.78
10.2311.6213.0714.45
15.9117.3618.7420.2021.58
23.0324.4225.8727.2628.71
30.1031.5532.9334.3935.77
37.2238.6140.0641.4542.90
7.528.98
10.3611.8113.2014.65
16.0417.4918.8820.3321.71
23.1724.5526.0027.3928.84
30.2331.6833.0734.5235.90
37.3638.7440.1941.5843.03
7.669.11
10.4911.9513.3314.78
16.1717.6219.0120.4621.85
23.3024.6826.1327.5228.94
30.3631.8133.2034.6536.10
37.4938.9440.3341.7843.16
7.799.24
10.6312.0813.4614.92
16.3017.7519.1420.5922.04
23.4324.8826.2727.7229.11
30.5631.9433.4034.7836.23
37.6239.0740.4641.9143.30
7.929.37
10.8212.2113.6615.05
16.5017.8919.3420.7222.18
23.5625.0126.4027.8529.24
30.6932.0833.5334.9136.37
37.7539.2040.5942.0443.43
8.129.50
10.9612.3413.7915.18
16.6318.0219.4720.8622.31
23.6925.1526.5327.9829.37
30.8232.2133.6635.0536.50
37.8839.3440.7942.1743.63
8.259.64
11.0912.4813.9315.31
16.7618.1519.6020.9922.44
23.8325.2826.7328.1229.57
30.9532.4133.7935.2436.63
38.0839.4740.9242.3143.76
8.389.77
11.2212.6114.0615.51
16.9018.3519.7321.1922.57
24.0225.4126.8628.2529.70
31.0932.5433.9235.3836.76
38.2139.6041.0542.4443.89
SLOPE CORRECTIONS IN POINT-SAMPLING
I. Sloping land complicates point-sampling just as it complicates plot-sampling. If slope correction is neglected, the errors will be of sig-nificant magnitude when slope is more than about 10 percent.
A. A slope factor (s.f.) may be easily determined using the averageslope at each point, because:
s.f. = secant of angle of slope.
B. With this factor, point-sample tree counts may be corrected bymerely multiplying:
Corrected count = (tree count) (s.f.)
C. Using this method, the correction can be made for any slope situa-tion although, in some western states, cruisers have encountered agreat deal of difficulty and noted some significant inaccuracies inactual application.
D. Grosenbaugh (1955) gives correction factors for most slopes (Table11). Note the equation at the foot of the table.
Practical Point-Sampling 35
II. Slope may be corrected also by rotating the prism through a verticalangle equal to the angle of the slope.
A. Certain prisms are manufactured with a rounded top to facilitateuse in the rotated position.
B. This is perhaps the best solution of the slope problem.
C. It has been used throughout the United States and, in some westernpoint-sampling—wherein the tree is always viewed at the top ofthe first 16-foot log as a hedge against brush-bias—almost everytree is on a slope (Bruce 1961).
TABLE 11. APPROPRIATE CORRECTION FACTORS FOR BA ORVOLUME PER ACRE CALCULATED FROM UNADJUSTED ANGLE-GAUGE TALLIES TAKEN ON A SLOPE, WHERE SLOPE PERCENTIS MEASURED AT RIGHT ANGLES TO CONTOUR.
Limitsof
PercentSlope10.0
17.4
22.5
26.7
30.4
33.6
36.6
39.5
42.1
44.6
47.0
49.3
51.5
53.7
55.8
SlopeCorrection
Factor
1.01
1.02
1.03
1.04
1.05
1.06
1.07
1.08
1.09
1.10
1.11
1.12
1.13
1.14
Limits Slopeof Correction
Percent FactorSlope55.8
57.8
59.8
61.7
63.6
65.4
67.2
69.0
70.8
72.5
74.2
75.8
77.5
79.1
80.7Correction factor for
1 /Slope
\5
1.16
1.17
1.18
1.19
1.20
1.21
1.22
1.23
1.24
1.25
1.26
1.27
1.28
Limitsof
PercentSlope80.7
82.3
83.9
85.4
86.9
88.4
89.9
91.4
92.9
94.3
95.8
97.2
98.7
100.1
101.5
SlopeCorrection
Factor
1.29
1.30
1.31
1.32
1.33
1.34
1.35
1.36
1.37
1.38
1.39
1.40
1.41
1.42
steeper slopes is:
percent >
100 /1
2
Source: Grosenbaugh, L. R. "Better diagnosis and prescription in southernforest management." U. S. Forest Service, Southern Forest Ex-periment Station, Occasional Paper 145, 1955.
36 Stephen F. Austin State College
CONTINUOUS FOREST INVENTORY WITH POINT-SAMPLING
I. Point-sampling techniques are just as suitable for continuous forestinventory (CFI) as are plot sampling techniques.
A. Point-sampling has the special advantage of taking a smaller per-cent sample of the many small stems in uneven-aged forests, whilesampling the fewer large stems in much the same proportion as infifth-acre plot-sampling.
B. This is illustrated from experience with CFI in the Sabine NationalForest:
DBH Class Number of Trees Tallied—110 Locations1/5-Acre Plots 10-BAF Point Samples
(1957-58) (1962)5 - 9 1782 23910 -14 601 21115 -19 248 18420 - 24 55 5025+ 2 5
II. The techniques of point-sampling, and the resulting calculations, areas convenient for electronic data processing as are the mechanics ofplot-sampling.
A. U. S. Steel in Alabama (University of Georgia, 1959) used point-sampling in its CFI plan and measured six to eight stems at eachpoint. They measured all trees due to the small number at each loca-tion, and thus have a permanent record of trees from seedling-sizeup.
B. The Southern Forest Survey, a large-scale CFI system, now em-ploys point-sampling as its inventory technique. Samples are ona 3 x 3-mile grid. Clusters of ten 37.5-BAF samples are collectedat each survey location.
III. Several problems not attributed to plot-sampling are present whenpoint-samples are used in continuous forest inventory.A. Trees grow into the sample according to their size and distance
from the point, not just according to their size. In other words, a20.0-inch tree 79.34 feet from the point will come into the samplewhen it grows to 20.4 inches dbh, if the sample is being taken witha 5-BAF prism. Since its total volume is not really ingrowth, thevolume per acre and the number of stems per acre it representswill be used for future growth determination, but will not neces-sarily figure in past growth calculations.
B. Unless dbh and distance are both remeasured, omitted or extratrees will severely affect stand and stock table values. In brushyareas, obscured trees may be omitted accidentally in the point-sample.
C. Since a CFI sample is exceedingly small and blowups are large,extreme care must be exercised in taking the CFI point-sample.
Practical Point-Sampling 37
GROWTH PROJECTION WITH POINT-SAMPLES
I. Point-sampling can be applied to growth projection studies (Fenderand Brock, 1963) for future volume or basal area as well as to deter-mine actual current volume or basal area.
II. To show the principle involved, in an exceedingly simplified method,trees in a southern pine plantation were bored to determine averageprojected 5-year diameter growth. Using a table of limiting distancesfor the "new" diameters, the projected stand 5 years hence was "tal-lied," giving the following results:
Present FutureCount BA/Acre Count BA/Acre
8 160 9 1809 180 10 2009 180 13 2609 180 10 2006 120 8 160
Average 164 Average 200
NUMBER OF POINTS TO ACHIEVE STATED PRECISION
I. The forester, contemplating a point-sample cruise, must know howits accuracy compares with that of the conventional plot cruise he haspreviously employed. He needs to know the total number of points toobserve on a property of given acreage to yield the required precision.
A. Bruce (1961) says that in western timber, point-sampling requiresabout the same number of points as 1/4- or 1/5-acre plots to achievea stated precision of volume estimate, if the prism is properlymatched to tree size in the stands sampled. This means that countsper point must average about 5 trees.
B. For southern forests, Grosenbaugh (1955) wrote, "The reconnais-sance would require a minimum of about 20 unbiased sample pointswell distributed throughout the record unit for a standard error ofabout ± 25 percent in volume per record unit. A volume total of25 such record-units would have a standard sampling error of onlyabout ± 5 percent for the ownership or working circle as a whole."
C. Bell and Alexander (1957) said, "present experience shows that thesame intensity as that used for one-quarter-acre plots will resultin a stastically better sample than given by the one-quarter acreplot method."
D. Putnam et al. (1960) wrote, "The reconnaissance preceding thefirst managed cut should ordinarily be an inexpensive affair. Aninspection and sight estimate may suffice for experienced men,though a more formal approach will be necessary for most foresters.The point-sampling technique makes it possible to get approximateestimates of volume with very little effort. If a prism or anglegauge with a basal area factor of 10 is employed to select sampletrees, 25 sample points distributed systematically over manage-ment unit will usually give a sampling error for board-foot volume
118854
38 Stephen F. Austin State College
of less than 23 percent. The aggregate volume on four managementunits will ordinarily be estimated within twelve percent. The goalis not extreme accuracy but rather rough estimates obtained asquickly and cheaply as possible.
"If a more intensive survey is needed, the following tabulation maybe used to obtain approximate and usually conservative samplingerrors as a percent of the total volume.
Points Error(number) (percent)
25 2350 17
100 12200 8400 6
1600 3"For cruises of the specified intensity, differences between samplevolume and volume by 100-percent tally will be exceeded, on theaverage, in only one cruise in three. Five percent of the time, thedifference might be twice the chosen limit, but 50 percent of thetime the difference will be less than 68 percent of the tabulatederror.
"This means, for example, that a cruiser choosing 12 percent as hislimit of sampling error may be wrong by more than this amount onone cruise out of three. In fact, one time in twenty his actual errormay be more than 24 percent, but half the time his error will beless than 8 percent. In the rare cases where a sample of low in-tensity might deviate from the true mean considerably, an ex-perienced cruiser will suspect the lack of representativeness andintensify the sampling accordingly.
"Whether sample points are located systematically or at random,the area of the tract has little effect on the number of points re-quired for a given sampling error. The important thing is thevariability among the aggregate volumes sampled at various points.The tabulation is based on average variation over a large portion ofan entire southern state. Errors may be larger in extremely ir-regular stands and smaller where stands are relatively uniform."
E. Kendall and Sayn-Wittgenstein (1959) said, ". . . The number ofpoint samples required to give a precision equal to that from con-ventional 1/5-acre-plots varied with the type of stand and averaged2 0.o.
NUMBER OF POINT SAMPLES (N) EQUIVALENT TO 1/5-ACREPLOT
(basal area factor = 10)Location Stocking Diameter Modal N
Type of (trees Range Diameter (by randomStand per acre) (inches) (inches) selection)
Eagle Depot even-aged pine 340 1-25 5 3.1Forbes Depot pine-spruce 760 1-14 4 2.3
Practical Point-Sampling 39
"Although no detailed time study was undertaken, it is estimatedthat with the relascope, BA was sampled to a given precision inone-half or one-third the time required when 1/5-acre plots wereused. In the tests carried out in the Forbes Depot and Eagle Depotareas in 1957, the number of trees tallied for volume calculationsin the point samples numbered, 1,290, whereas 12,301 were talliedin the corresponding 1/5-acre plots. This is partly due to the factthat with the relascope each tree size is sampled on a land areawhich is proportional to its BA. A greater proportion of largetrees is therefore sampled, and the time consuming measurementof small trees is largely eliminated."
F. For 110 double sample points, the error of BA would be onlyabout ± 3.1 percent and for 110 single points the error would beabout ± 4.5 percent on a pine-hardwood area in the Sabine NationalForest (see page 16).
COMPARING METHODS OF PER ACRE POINT-SAMPLEVOLUME ESTIMATES
I. The School of Forestry, Stephen F. Austin State College, maintains110 permanent continuous forest inventory plots on the Sabine NationalForest, Texas. These were measured in 1957-58 and remeasured in1962. At the time of remeasurement, a point-sample using a 10-BAFprism was made in addition to the fifth-acre plot-sample. As an indi-cation of the various per acre volumes, using various methods andfactors in computation, the following data are presented from ten ofthe 110 point samples. The range of per acre volumes may be less ifdata from the entire array of 110 points were presented (Table 12).
TABLE 12. SUMMARY OF THE FACTORS BY POINTS USED TOCOMPUTE VOLUMES IN VARIOUS WAYS
Point X Bff
212223242526372829210
Saw-timber
(1)27.942.510.820.019.184.030.460.217.338.8
351.0
SbflSaw-
timber(2)
193.0416.8131.2239.2239.2980.2308.8594.4216.0393.6
3712.4
2 LogsSaw-
timber(3)3.58.02.54.54.5
18.56.0
11.54.07.5
70.5
2 Bd. Ft.Saw-
timber(4)
2677.704326.711442.012395.732515.169972.113209.106299.681927.774100.79
38866.32
2 Cu. Ft.PulpTrees
(5)105.552366.048701.001893.298120.813494.318705.422
0541.352450.957
4388.761
2 cffPulpTrees
(6)12.3842.9672.9295.0814.9855.3477.50
049.5247.54
468.02
2 FeetPulpTrees
(7)15.957.1
102.2128.0
18.973.4
100.20
71.162.9
629.7
(1). Using the standard volume table type of approach:
2 bff 351.0Vol. bd. ft. per acre = (BAF) (10) = (10) (10) = 3510.0
N I D
40 Stephen F. Austin State College
(2). Using the board foot factor per height class method:
2 bff 3712.4Vol. bd. ft. per acre — (BAP) = (10) = 3712.4
N 10
(3). Using the 60 factor and sum of logs method:
(2 L) (60) (BAF) (70.5) (60) (10)Vol. bd. ft. per acre = = = 4320.0
N 10
(4). Using the stand table approach with volume for each d.b.h. class:
2[(T) (S.T.F.) (Vol)] 38,866.32Vol. bd. ft. per acre = = = 3886.6
N 10
(5). Using the stand table method for cubic foot volume of pulpwood:
2[(T) (S.T.F.) (Vol)] 4388.761Vol. cu. ft. per acre = = = 438.9
N 10
(6). Using the c.f.f. for each height class in pulpwood trees:
2 cff 468.02Vol. cu. ft. per acre = (BAF) = (10) = 468.02
N 10
(7). Using the cubic foot factor 0.7052 and merchantable pulp tree height(in feet) method:Vol. cu. ft. per acre =
( 2 H ) (.7052) (BAF) (629.7) (.7052) (10)= = 444.06
10 10
For comparison: 1958 volumes of 10 fifth-acre plots at same location:
410.02 cu. ft. per acre and 4806.8 bd. ft. per acre.
A TEST OF POINT-SAMPLING IN A PLANTATION
I. Whether point-sampling gives adequate results in a plantation is notreadily resolved. The Nacogdoches City Plantation was inventoried in1966 by several methods. The results indicate that point-sampling doesadequately inventory plantations, although the 1/20-acre circular plotsyield basal area and cubic foot volumes which were closer to the 100%tallv than were the Doint-saniDles (Table 13).
TABLE 13. METHODS OF SAMPLING, NACOGDOCHES CITY PLANTATION, 1966.
Inventory Method
100% Stand TableSystematic Point-Sample2
Number of Basal AreaSamples
_20
Systematic 1/20-acre Circular Plot 20Random Point-SampleRandom 1/20-acre CircularRandom 1/20 Row PlotSwinford's Progressive
Point-Sampling1
Systematic 1/20-acre RowSwinford's Progressive
Point-Sampling1
20Plot 20
20
12Plot 20
8
per Acre(sq.ft.)
91.643103.00088.488
108.00091.08299.698
106.66094.356
100.000
ProportionalLimit ofError
(Percent)
none13.37.8
16.06.7
14.9
15.612.4
21.4
crMofBasalArea
none6.8603.4568.6323.0626.807
8.3105.831
10.690
Cubic FeetPer Acre
1960.982229.471835.342359.821899.212073.74
2343.222094.89
2251.18
ProportionalLimit ofError
(Percent)
none13.99.9
16.77.8
14.4
15.713.8
22.8
aMofCubic
Volume
none156.280
90.620196.62074.320
149.54
183.74141.660
256.160
TreesPer
Aero
236254241238243267
264226
236
-sa
£.h_J
0
ftCos
"8.H
Adapted from Swinford (1966).
2A11 systematic samples taken at same locations.
42 Stephen F. Austin State College
LITERATURE CITED
Bell, J. F. and L. B. Alexander. Application of the variable plot method ofsampling forest stands. Oregon. State Board of Forestry. ResearchNote. No. 30. 1957.
Bruce, Donald. Prism cruising in the western United States and volumetables for use therewith. Mason, Bruce and Girard, Consulting For-esters, Portland, Oregon. 1961.
Dixon, R. M. Point sampling, wedge prisms and their application in forestinventories. Ontario Department of Lands and Forests. 1958.
Fender, Darwin E. and Gerald A. Brock. Point center extension: a tech-nique for measuring current economic growth and yield of merchantableforest stands. Journal of Forestry 61: 109-114. 1963.
Grosenbaugh, L. R. Plotless timber estimates . . . new, fast, easy. Journalof Forestry 50: 32-37. 1952.
Grosenbaugh, L. R. Shortcuts for cruisers and sealers. U. S. Forest Serv-ice. Southern Forest Experiment Station. Occasional Paper 126. 1952a.
Grosenbaugh, L. R. Better diagnosis and prescription in southern forestmanagement. U. S. Forest Service. Southern Forest Experiment Sta-tion. Occasional Paper 145. 1955.
Hovind, H. J. and C. E. Rieck. Basal area and point sampling WisconsinConservation Department. Technical Bulletin 23. 1961.
Hunt, Ellis V., Jr. Need a reliable point sampling board foot factor?—Lookin your files. Journal of Forestry 59: 512-513. 1961.
Kendall, R. H. and L. Sayn-Wittgenstein. An evaluation of the relascope.Canada Dept. of Northern Affairs and Natural Resources. Forest Re-search Division. Technical Note 77. 1959.
Putnam, J. A., et al. Management and inventory of southern hardwoods.U. S. Department of Agriculture. Agriculture Handbook No. 181: p. 83.1960.
Swinford, Kenneth R. Plantation measurements. Proceedings of 15th annualForestry Symposium, Louisiana State University, p. 136-162. 1966.
University of Georgia. Continuous inventory control in forest manage-ment, short course proceedings. Athens, Georgia. May 4-8, 1959.
Practical Point-Sampling 43
ABBREVIATIONS
BA basal areaBAP basal area factorbff board foot factorBff modified board foot factorC coefficient of variationCFI continuous forest inventorycff cubic foot factord distance between target and prismD tree diameter in inchesdbh or DBH diameter at breast heightf frequencyP tree total height in 10 foot unitsH total height of tree in feetht. heightL merchantable height of trees in 16-foot logsM meanN number of points or observationsPie. proportional limit of errorS sampling error of a mean in percentSe. combined sampling error in percents.f. slope factorS.T.F. stand table blowup factorT tree count in a diameter classV variancew target width, or distance between pinsS sum ofa standard deviationo-m standard error of the mean
FORESTRY BULLETINS, Stephen F. Austin State College
1. References of Value in Studies of Insects Affecting the Southern Pines,an Annotated List, by R. C. Thatcher. 1957.
2. Directory of Wood-using and Related Industries in East Texas, by N. T.Samson. 1957. (Out of print.)
3. Bibliography of Forestry Films and Filmstrips, by N. T. Samson. 1958.(Out of print.)
4. Root Development of Loblolly Pine Seedlings in Modified Environments,by M. V. Bilan. 1960.
5. Continuous Forest Inventory with Punched Card Machines for a SmallProperty, by R. D. Baker and E. V. Hunt, Jr. 1960.
6. Point-sampling from Two Angles, by E. V. Hunt, Jr., R. D. Baker andL. A. Biskamp. 1964. (Out of print.)
7. Films and Filmstrips on Forestry, by N. T. Samson. 1965.8. Soil Moisture and Soil Temperature under a Post Oak-Shortleaf Pine
Stand, by G. Schneider and J. J. Stransky. 1966.9. Silviculture of Shortleaf Pine, by L. C. Walker and H. V. Wiant, Jr.
1966. Price $1.10. Texas Pulpwood Production, by N. T. Samson. 1966.11. Silviculture of Longleaf Pine, by L. C. Walker and H. V. Wiant, Jr.
1966. Price $1.12. Pine Seedling Survival and Growth Response to Soils of the Texas
Post-oak Belt, by M. V. Bilan and J. J. Stransky. 1966.13. Directory of Wood-using and Related Industries in East Texas, by
N. T. Samson. 1966. Price $2.14. Practical Point Sampling, by E. V. Hunt, Jr. and R. D. Baker. 1967.
Price $1.