Historic, archived document

Do not assume content reflects current
scientific knowledge, policies, or practices.

s

@ untedse, Effects of Leader Topping and

Agriculture

SZ coerscee Branch Pruning on Efficiency
peucserees Of Douglas-Fir Cone Harvesting

Experiment Station

reseacnnote With a Tree Shaker

PNW-431
October 1985
Pace PSW FOREST AND RANGE

UAS Donald L. Copes EXPERIMENT STATION

EMT OF MICS

JAN 2 GIGS

STATION LIBRARY COPY

Abstract In 1983, a study was conducted to evaluate the effects of leader topping and
branch pruning on the efficiency of tree shaking to remove Douglas-fir
(Pseudotsuga menziesii (Mirb.) Franco) cones. Removal efficiency for three topping
and pruning treatments averaged 69 percent, whereas for the uncut control treat-
ment it was 62 percent. The treatment combination that resulted in the greatest
cone removal (70.8 percent) was cutting the last 2 years of growth from the leader
but not pruning lateral branches. The pruning-topping effects were not additive,
possibly because branch pruning several weeks prior to cone harvest resulted in
premature removal of a portion of the most easily shaken cones.

Keywords: Cone collection, pruning, top pruning, pruning tools, Douglas-fir, tree
shaker.

Harvesting conifer cones with mechanical tree shakers usually results in lower effi-
ciency than is achieved with identical machines in harvesting deciduous nuts and
fruits. Conifer cones are often attached to the branches with tough, woody stalks
which necessitate a strong shake force to remove the cones. An attempt to induce
abscission zone formation in the woody stalks by spraying abscission-inducing
chemicals was tested in southern pine species, but the procedure was not suc-
cessful (5). Similar abscission tests have not been reported for Douglas-fir
(Pseudotsuga menziesii (Mirb.) Franco).

Results from several field trials of harvesting cones of Douglas-fir with tree shakers
have been reported (1, 2, 4, 11). In addition, two carefully controlled laboratory
studies have been done to gather basic information on shaking characteristics of
Douglas-fir. In the first study, detached cone-bearing limbs were shaken on a
laboratory device that could vary both stroke length and frequency of shake (8); the
second laboratory test simulated shaking two stems with different shapes (2).

DONALD L. COPES is a principal plant geneticist at the Forestry
Sciences Laboratory, Pacific Northwest Forest and Range Experi-
ment Station, 3200 Jefferson Way, Corvallis, Oregon 97331.

Methods

Results from tree shaking trials with Douglas-fir (2, 4, 11) and with other conifers (1,
10, 11) have been encouraging for use of this technique, but orchardists want more
complete removal of cones. An average of 55 percent of Douglas-fir cones were
removed when 7 to 9-m-tall seed orchard trees were shaken (4), and 52 to 70 per-
cent were removed from larger forest-grown Douglas-fir (11). Attempts to evaluate
the variables that influence the proportion of cones that can be removed indicate
interactions of many factors: species (1, 2, 8, 11), clone (4, 10), stem diameter (3,
10, 11), tree height (1, 3, 4, 11), stem taper (1, 11), height to live crown (1, 3, 4, 11),
physical characteristics of the wood (9), foliage density (1, 11), crown shape (11),
branch size (2), branch angle (4), damage susceptibility (1, 10), cone distribution
and size (1, 11), and cone maturity (1, 2). Sources of variation related to the tree-
shaking machines themselves include factors such as length of shake stroke (7, 8,
11), number of shake cycles per minute (1, 4, 7, 8), amount of shake force or thrust
applied to the bole (4, 11), type of thrust applied (4, 11), shake pattern (4, 7, 8),
shaker head attachment height (1, 4, 10, 11), duration of shake (4, 7, 10, 11), and
number of shakes (4, 8). The fact that many variables interact to influence cone
removal is indicated by correlation coefficients (r2) being less than 0.44 when cone
removal is correlated with single variables, such as tree height, stem diameter, or
height to live crown (1).

Results from the 1982 shaking test with Douglas-fir (4) suggest that crown shape
and branch length can influence cone removal and crown damage. Cones could
not be easily shaken from trees with long, pendent branches. Cones were most
readily removed from short branches in the upper third of the crown. More shake
energy appeared to be conducted to the ends of short, stiff branches. Cone re-
moval was poorest from the lower third of the tree. The damping action of dense
crowns on machine-induced vibrations and the propensity for excessive top
breakage during shaking are both documented for Douglas-fir trees (1).

In 1983, | decided to test the effects of minor modifications of crown shape or
structure of Douglas-fir on the shaking characteristics. The hypothesis was that
trees with shorter, stiffer branches and less flexible tops would yield a greater
percentage of cones and have less shaker-induced damage than untreated trees.

Objectives of the study were to determine (1) whether branch pruning and leader
topping prior to tree shaking improves cone removal, and (2) whether topping

results in less crown damage. A future objective is to evaluate harvest efficiency
and tree vigor or health after the trees have been shaken in different crop years.

Sixty-eight trees growing in a Forest Service experimental plantation 24 kilometers
north of Corvallis, Oregon, were chosen for study in 1983 because the planting
closely approximated conditions in seed orchards. The trees all had normal crown
form and foliage density and had good potential for cone production. Most trees
had produced one or two cone crops during the preceding 4 years and 65 of the
68 trees were producing cones in 1983. The trees were in an area protected for
research. Of the 68 study trees, 48 had been propagated as rooted cuttings in
1968 and had been planted in 1971; the other 20 trees were grafts made in 1973 on
large, well-established rootstocks. Grafted trees were somewhat smaller in diameter
and height than the cuttings (12.0 cm and 8.0 m vs. 15.8 cm and 89 m).

Leader topping and branch pruning were studied for their effect on cone removal
efficiency and degree of top damage. Topping consisted of cutting off the top of
each tree approximately 20 cm above the 1982 branch whorl. Pruning consisted of
cutting off the distal end of major branches in each annual growth whorl. The
amount cut from a branch varied according to branch age: all of the 1982 and
1983 increment was cut from the tip of branches 3 years and older; only the 1983
increment was cut from branches in the 1982 whorl.

The experiment was a 2 x 2 factorial of pruned vs. unpruned branches, and topped
vs. untopped leaders in a randomized complete block design with 17 replications.
Within each replication, each treatment was applied to one of four adjacent trees
selected at random. Twelve replications consisted entirely of cuttings and five con-
sisted of grafts. Pruning and topping were done in early August 1983; one worker
on the ground used a pole pruner and another worker in the bucket of a lift truck
used hand pruning shears.

Tree shaking was done the last week of August 1983 when the cones were mature.
Cones had already begun to open on several trees. The Kilby Co.’ boom-type
tree shaker described in the 1982 shaking study (4) was used. The same low
energy, gentle shaking procedure developed in 1982 (4) was also used, except that
clamping pressure of the shaker head was reduced from 750 to 500 Ib/in? and the
trees were shaken with the shaker in only one position rather than in two positions
as in 1982. All 68 trees were shaken, including 3 trees that did not have cones.
The 3 barren trees were given the same shaking treatment to collect crown
damage information and to ensure that cumulative effects of shaking can be
evaluated in future years on all 68 trees.

Each tree was shaken until it appeared that no more cones could be removed
without increasing the shake energy to levels that would cause unacceptable upper
crown breakage. The cones fell onto plastic tarps placed under each tree. The
length of time each tree was shaken was recorded to the closest 5 seconds. Cones
remaining attached to the trees after shaking were hand picked. Both the shaken
and hand-picked cones were weighed with a spring scale immediately after collec-
tion. Tree measurements recorded included stem diameter, height before and after
leader pruning, and height after shaking. Branch damage was recorded by noting
the age of the oldest branch whorl on which broken tips occurred. Any external
trunk damage to the bole was noted.

Data were analyzed for the 2 x 2 factorial by analysis of variance techniques.
Percentage data were subjected to arcsin transformation prior to analysis. Cone
removal data were analyzed for the 65 cone-bearing trees; and tree size, pruning
length, and shaking data were evaluated for all 68 trees.

The use of company or brand names is for the convenience of
the reader and does not constitute an endorsement by the U.S.
Department of Agriculture.

Frequency of trees (percent)

60-69 70-79 80-89 90-99

Cone removal classes (percent of cones removed)

Figure 1.—Frequency distribution of study trees in various cone
removal classes.

Results Cone removal by shaking was successful with most trees: Only 14 percent of trees
failed to drop at least 50 percent of their cones (fig. 1). Cone removal averaged 67
percent for all trees in the study (table 1). Trees subjected to topping or pruning
treatments yielded more cones than did control trees. Controls averaged 61.5 per-
cent, whereas pruned and/or topped trees averaged 68.7 percent. Analysis of
variance tests showed neither pruning nor topping treatments to be significant at
the 5-percent level, but the pruning-topping interaction was significant. The most
effective treatment was topping without pruning; that treatment averaged
70.8-percent cone removal.

Effects of replication were significant at the 1-percent level for diameter, initial tree
height, and total amount of height reduction because of the height and diameter
size difference between grafts and cuttings (table 2). Randomization of treatments
within replications was effective in that treatments were not significantly different for
initial tree heights or stem diameters.

The two leader-topping treatments resulted in the mean removal of 1.4 and 1.7
meters from the upper crown of each tree. Total tree height reduction caused by
topping and by shaker breakage averaged 0.5 meter more in topped trees than in
untopped trees; the difference was significant at the 5-percent level. Leader
breakage occurred in one-third of all leader-topped trees; that loss averaged 0.2
meter for all topped trees. The length of time the trees were shaken was not
significant. -

Damage to branches after shaking was confined to 1983 branch tips in the top one
or two whorls. Fewer branches broke on pruned trees because the most fragile tips —
had already been removed. Leader topping did not appear to affect shaker damage
to branches.

Table 1—Effect of leader topping and branch pruning treatments on weight
and percentage of cones removed from Douglas-fir by shaking (means and
standard errors), and analysis of variance

Average weight of
cones per tree

Branch and Number
leader of Removed by Cones removed
treatment trees Total shaking by shaking
- - - - Kilograms - - - - Percent
Cone removal:
Leader topping--
Branch pruning 15 BiG) sp OA! B59. ae (55 66.6 + 3.5
No branch pruning 17 14.4 * .9 Oo Fore 10580 +53).0
Submeans Us S2e ete Sc i 685 /—
No leader topping--
Brancn pruning 17 S80 vt. Gatiete Al 68.8 + 3.1
No branch pruning 7 ex =
(control ) 16 Wile2 ae WAw 6.6 + <6 61.5 + 4.5
Submeans HWS 7 OAT 65scum
Means Will lens 67.0
Degrees
of freedom Probability
Analysis of variance: ce
Replications 16 .02
Leader topping (T) 1 31
Branch pruning (B) | 73
TxB ] .05
Experiment error 45

Table 2—Tree parameters (means and standard errors) before and after shaking
Douglas-fir, and analysis of variance

Length
Branch and Number Leader of top Total
leader of Time Original length broken by reduction
treatment trees shaken Diameter height cut off shaking in height
Seconds Centimeters ----------- Meters -----------
Tree parameters:
Leader topping--
Branch pruning 7 44 + 2.6 15.0 + 0.7 8.6 + 0.3 1.4 + 0.1 0.2 + 0.1 1.6 + 0.1
No branch pruning 17 54 + 3.8 15.0 + .8 Boi 22 Sion al, a mea Veco Al
Submeans 49 ~ 1St0' a 87 160m 2m Jaen
No leader topping--
Brancn pruning 17 AT F255 14.0 + .8 B25! CG} ew) es} 1 Nathan all
No branch pruning 7% ir € a - "
(control ) 7 49 + 4.3 tS de GP Bro ae O00 12 1 Nez ail
Submeans 48 14.4 — 55 os Oma Vesa Vso
Means 48.5 14.7 8.6 -- -- 125
Degrees of
freedom ---------------- Probability -----------------
Analysis of variance:
Replications 16 ~45 0 0 -- et/ 0
Topping (T) 1 .59 18 -49 -- 0 0
Pruning (B) ] .14 -45 ~52 -- 46 34
TexaB ] .28 52 -61 -- -46 21

Experimental error 48

Discussion and
Conclusions

Little bole damage was detected where the shaker head was attached to each tree.
Three trees experienced slight bole damage (bark slippage) when the shaker was
accidentally operated with less than 400 Ib/in2 clamping pressure. Inadequate clam-
ping pressure allowed undesirable movement between the pads of the shaker head
and the bole. A clamping pressure of 500 Ib/in2 was adequate for proper transfer of
shake energy to trees that averaged 8.6 m in height but was not so strong that
clamping pressure damaged the cambial tissues beneath the pads.

Cone harvest by shaking was more effective in 1983 than in 1982. Average cone
removal increased from 55 percent in 1982 (4) to 67 percent in 1983. The most ef-
fective 1983 treatnfent was to leader-top the trees without pruning branch tips. The
6.5-percent increase in yield of cones of control trees in 1983 over yield in 1982 is
attributed primarily to increased skill of the machine operator. The additional 5- to
10-percent increase in cone yield in 1983 probably resulted from the branch-
pruning and leader-topping treatments.

The pruning and topping treatments were not done until 2 weeks before shaking;
consequently, some cones were removed when the branch tips were pruned.
Cones located on the tips of branches in the top one-third of the crown could have
been easily removed by the shaker. The cone loss prior to shaking is evidenced by
the lower average total weight of cones on branch-pruned trees than on trees that
were not branch pruned. A more practical seed orchard procedure would be to
limit pruning to the branches that were more difficult to shake in the lower two-
thirds of the crown. Another technique that would enhance yield would be to prune
only when there is no cone production; that would eliminate cone loss, yet still
shorten and stiffen the most difficult-to-shake lower branches.

Both leader topping and branch pruning had greater influence on cone removal
when done separately than when combined, but neither treatment effect was
significant at 5 percent. The significant pruning-topping treatment interaction in-
dicates that although the two treatments did affect cone removal, their effects were
not additive when used together on the same trees. Removal of cones by clipping
branch tips in the upper crown prior to shaking may be the most likely cause of
the significant topping-pruning interaction. Large standard errors suggest that other
uncontrolled variables also had considerable influence. As previously stated, many
variables—both plant and machine—interact to determine the proportion of cones
that can be removed by shaking (11).

Predicting the actual point where untopped trees will break during shaking is dif-
ficult. The top two or three internodes of Douglas-fir trees are slender and very
Subject to breakage when energy applied is sufficient to remove a majority of
cones. Leader pruning prior to shaking allows orchardists to determine final tree
height more accurately. Topping removes the most break-prone portion prior to
shaking and may allow the machine operator to safely shake leader-pruned trees
with slightly more force than trees whose leaders are not topped. On the average,
topped trees had 0.5 m greater reduction in total height than did untopped trees,
but this will not result in a meaningful difference in the size over the productive life
of a seed orchard. Shaker-induced breakage of the main leader averaged 0.6 m
and occurred in only about one-third of the leader-topped trees, whereas breakage
in untopped trees averaged 1.25 m and occurred in all untopped trees.

English Equivalents

Literature Cited

Shaking these same trees when the next cone crop is produced will likely yield
results somewhat different from those in 1983 because branch pruning is not
planned for at least 2 years and then only during a noncrop year. Leader-topping
may not be necessary if cone crops are frequent enough that top breakage induc-
ed by tree shaking keeps the upper crown in proper shape.

Observations and growth measurements will be made in future years on each of
the 68 study trees. Trees will be examined for unseen or delayed symptoms of
lower bole damage. | do not expect adverse effects from shaking on tree vigor or
cone production since none were found in slash pine 4 years after shaking (6).
Cumulative effects on crown structure will also be evaluated to determine if
repeated shaking will gradually improve cone removal efficiency by allowing the
shaker operator to use greater shake energy without causing unacceptable
damage.

1 centimeter (cm) = 0.3937 inch (in)
1 meter (m) = 3.2808 feet (ft)

1 kilogram (kg) = 2.2046 pounds (Ib)

1 kilometer (km) = 0.6214 mile (mi)

0.0703 kilogram per square centimeter (kg/cm?) = 1 pound per square inch (Ib/in?)

1. Anon. An evaluation of two shakers for harvesting cones. Equipment develop-
ment and test report. Missoula, MT: U.S. Department of Agriculture, Forest Ser-
vice, Missoula Equipment Development Center; 1972. 27 p.

2. Baily, Michael Paul. A vibration simulation analysis of two Douglas-fir stems
under a single harmonic forcing function for cone harvesting purposes. Seattle,
WA: University of Washington; 1982. 85 p. M.S. thesis.

3. Chappell, Thomas W. Harvesting pine cones with mechanical tree shakers.
Forest Engineering. 368: 65, 68; 1968.

4. Copes, Donald L.; Randall, William K. Tree shaking machine aids cone col-
lection in a Douglas-fir seed orchard. Res. Note PNW-406. Portland, OR: U.S.
Department of Agriculture, Forest Service, Pacific Northwest Forest and Range
Experiment Station; 1983. 8 p.

5. McLemore, B.F. Chemicals fail to induce abscission of loblolly and slash pine
cones. Res. Note SO-155. New Orleans, LA: U.S. Department of Agriculture,
Forest Service, Southern Forest Experiment Station; 1973. 3 p.

6. McLemore, B.F.; Chappell, T.W. Mechanical shaking for cones harmless to
slash pine. Journal of Forestry. 71(2): 96-97; 1973.

7. Miles, J.A.; Mehischau, S.; Moini, S. Factors affecting harvest of loblolly pine
cones. Transactions of the American Society of Agricultural Engineers. 24(2):
323-325; 1981.

8. Moini, S.; Miles, J.A. Conifer cone harvesting: an experimental approach.
Transactions of the American Society of Agricultural Engineers. 24(3): 549-551;
1981.

9. Moini, S.; Miles, J.A.; Chen, P. Physical properties of ponderosa pine tree.
American Society of Agricultural Engineers. Paper 79-3545; 1979. 27 p.

10. Shea, G.M.; Bowyer, P.C.; Frodsham, T.M. An investigation into tree shaker
harvesting of slash pine cones in south Queensland. Tech. Pap. 16. Brisbane,
Queensland: Department of Forestry; 1979. 11 p.

11. Tietz, John. Equipment for cone and seed harvesting—project record report.
Missoula, MT: U.S. Department of Agriculture, Forest Service, Missoula Equip-
ment Development Center; 1968. 44 p.

Pacific Northwest Forest and Range
Experiment Station

319 SW. Pine St.

PO. Box 3890

Portland, Oregon 97208