Long-Term Growth Records of a Loblolly Pine Plantation

Boris Zeide and John Stephens

Abstract: Long-term observations conducted on permanent plots are indispensable for forestry research. Permanent plots serve as a touchstone upon which we test our hypotheses, models, and methods. Long-term observations in forestry are probably among the longest projects in all biology. The longest active study of the effects of thinning and pruning on growth of loblolly pine (Pinus taeda L.) has been conducted for 36 years on 45 permanent plots located in southeastern Arkansas. This study was initiated by the Forest Service researcher Dr. James D. Burton, who in 1970 established 40 permanent plots in a typical 12-year-old loblolly pine plantation. During this period the Georgia-Pacific company, presently Plum Creek, provided, in addition to the land, much help in maintaining the plots. The original design included 40 plots regularly thinned to four levels of basal areas of 21, 16, 12, and 7 m²/ha (90, 70 ,50, 30 ft²/ac). Within each level the trees were pruned in two stages at 12 and 15 years, finally clearing the bole to the height of 10, 8, and 7 m ( 33, 26, and 22 ft) and reducing the original crown length to 25, 40, and 50% of the total tree height, respectively. Each thinning-pruning combination has three replications within a randomized complete block design. Four plots were thinned (one for each of the four thinning treatments) but not pruned. Since 1981, the study has been continued by the School of Forest Resources of the University of Arkansas at Monticello. Five control plots (without thinning or pruning) were established at the age of 27 (in the fall of 1984) on the adjacent untreated part of the plantation. The variation in stand density was further enhanced by three severe ice storms that hit the plots at age 16, 21, and 36 years. The height to even-digit upper stem diameters was measured by a Zeiss Teletop (and later by a Criterion 400). As a result, reliable estimates of total and merchantable volumes were obtained using Grosenbaugh’s height accumulation method. Tree diameters have been measured 13 times, the last time in the fall of 2005 at age 48. This report summarizes growth trends of main stand variables (diameter, height, and volumes) by thinning and pruning levels and presents results of other projects (mill study, evaluation of ice storm damage) conducted on the plots. Too often long-term observations are locked in file cabinets or in growth databases with restricted access. In contrast, the complete set of our data and related projects are freely available at: http://www.afrc.uamont.edu/growthyield/montthinprun/index.html

HISTORY

In 1970 Dr. James Burton of the U.S. Forest Service established a thinning and pruning study in a typical pine plantation planted by Georgia-Pacific Corp. The stand had been initiated during the dormant season of 1958-1959. The trees were loblolly 1-0 stock, of unknown seed source, and machine-planted on a site five miles south of the University of Arkansas at Monticello. Forty square plots were initially laid out and thinning and pruning treatments applied.

The study area is located 5 miles south of Monticello, Arkansas on the coastal plain physiographic region near the Mississippi delta. The area receives 107 cm (42 in) of rainfall a year. The mean high temperature for January is 11° C (52° F) and the mean low is -1° C (30° F), and the high for July is 34° C (93° F) with a mean low of 21° C (70° F). Since 1981, the study has been continued by the School of Forest Resources of the University of Arkansas at Monticello under the supervision of Dr. Boris Zeide. A succession of research specialists-Kevin Dobelbower, Daniel Leduc, Suzanne Wiley, Vadim Shvets, Yujia Zhang, Curtis VanderSchaaf, Ralph Meeker, and presently John Stephens have measured and maintained the plots. During this period the Georgia Pacific Corp, presently Plum Creek, provided, in addition to the land, much help in maintaining the plots.

Soil analysis was initially done on each of the forty plots and concluded that 7 of the plots were Boswell silt loams with the remaining 33 being Sawyer silt loams. Basically, they are silt loam surface layers ‘Ap 5-20 cm (2-4")’ overlaying a clayey substrata ‘Bt to 107 cm (42")’ with the only difference between the two being the depth to the clay substrata. Burton stated that the land had been row cropped, and some of the A horizon had been removed by accelerated erosion. The land had been terraced not long before retirement. Slopes range from 0 to 4 percent, mainly southward. The soils are moderately well drained, with slow to very slow permeability.

Over the course of decades there were numerous events and changes which occurred on the site. In 1972 a series of 13 inventories began, the last being in 2005. At the same time the first of 6 bush-hoggings on the site occurred. A severe ice storm occurred a year after thinning at age 15 and caused heavy damage at age 16. Plots 4, 15, and 17 suffered most and required many years to recover to their target basal areas. A salvage cutting of 211 trees was conducted in response to this ice storm. There were less severe ice storms in 1979, 1994, and 2000. In 1981 the first of five prescribed burns was implemented. Five control plots were established in 1984, which included the herbicide injection of hardwood undergrowth. In 1986 the construction of a new road destroyed one of the control plots, and as a result a new control plot was laid out. After the creation of this last control plot, the study plot design has remained unchanged. In 1996 there was a salvage cutting of 6 trees due to southern pine beetle damage. At the age 40 thinning, a mill study was conducted of 54 felled trees. Measurements and grades of all dimension lumber were recorded by plot and log position within the tree.

OBJECTIVES

1) As the name thinning and pruning study indicates our first objective is to determine the optimal basal area that maximizes merchantable volume by age 30.
2) And to determine the pruning level which maximizes quality without affecting quantity.

METHODS

Layout

Plots were laid out on a 16 ha (40 ac) site using a randomized complete block design. Forty square treatment plots and 5 square control plots were laid out, each with 20 m by 20 m (66' x 66') inner measurement plots within identically treated 40 m by 40 m (132' x 132') square borders. The inner measurement plot corners were marked with red posts and the outer borders were marked with blue posts. All posts were tagged to indicate the associated plot number or, in the case of the outer border plots, the associated plot numbers. Four thinning treatments of 9, 14, 18, and 23 m²/ha (40, 60, 80, and 100 ft²/ac) residual basal area were initially established, and changed in 1972 to 7, 12, 16, and 21 m²/ha (30, 50, 70, and 90 ft²/ac) residual basal area. Pruning treatments in 1972 shortened crowns to 25, 40, and 55 percent of total height followed in 1973 by a second (and last) pruning treatment leaving crown ratios of 25, 40, and 50 percent, which cleared the bole to the height of 10, 8, and 7 m (33, 26, and 22 ft). There are 3 replications of thinning and pruning treatment combinations and 4 plots were thinned but not pruned. A diagram of plot layouts is shown in Figure
1 directly below.

Figure 1. Plot layout.

Density treatments in square meters per hectare
T1 = 7
T2 = 12
T3 = 16
T4 = 21

Pruning treatments as percent of live crown to total tree height
P1 = 25
P2 = 40
P3 = 50






 

Tree numbering

Various methods have been used to number the inner study plot trees. Instead of metal tags used by Dr. Burton, Dan Leduc suggested using a caulking gun to write tree numbers on a smoothed surface. In 2005, John Stephens etched the numbers into the tree bark with a tree scribe and then applied oil-based tree marking paint to the resulting grooves.



Measurements

In addition to diameter measurements at the marked breast height level of 1.4 m (4.5'), the height of even-digit upper stem diameters was determined using a Zeiss Teletop (and later by a Criterion 400). As a result, we have highly accurate estimations of total and merchantable volumes obtained using Grosenbaugh’s height accumulation method.

RESULTS

Basal area dynamics

Table 1. Basal level (before and after thinning) by thinning level (m²/ha).

Basal area dynamics reflect the changes in basal area for each target basal area treatment over time. In Table 1 the before basal level by thinning level is the basal area of the actual trees measured for any given target basal area and age on a per hectare basis. The after basal level by thinning level is the same basal area less trees that were missing by the next inventory. For instance, for age 27 and target basal area 7 m²/ha, the before basal level by thinning level was 8.8 m²/ha (38 ft²/ac). However trees were missing by the next inventory and when those trees are removed from the calculations, 7.4 m²/ha (32 ft²/ac) is the diminished basal area or after basal level by thinning level. However, by age 30 ingrowth has increased 7.4 to 9.2 m²/ha (40 ft²/ac), or the next measurement cycle’s before basal level by thinning level. At age 43 of target basal level 7 m²/ha, both before and after basal level by thinning level are 7.2 m²/ha (31 ft²/ac). This indicates that no trees were missing for that target basal level by the next measurement cycle of year 45.

It is important to note that the after basal level by thinning level (or what basal level actually exists on the treatment) has closely approximated the target basal area (or the basal level that we desire) for most treatments and years. As can be seen in Table 1, in the first measurement cycle or year 12, the after basal level by thinning levels were above the target basal area due to initial higher target basal areas of 9, 14, 18, and 23 m²/ha. The lower than desired after basal level by thinning level in year 16 can be explained by extreme ice storm damage. In later years after basal level by thinning level begin to rise above target basal area. This is due to the fact as plots begin to contain only a few large trees, it is exceeding difficult to thin back to the desired target basal level. However for the most part, the continuity of the treatments has been maintained over the life of the experiment. Also low standard errors areas are associated with the after basal level by thinning levels. For instance, for year 43 and target basal level 12 m²/ha, there is a standard error of 1 based on a basal level by thinning level of 11.9 m²/ha (52 ft²/ac). While there is a range in the proportion between standard error and after basal level by thinning level, in no case is the proportion excessive and in most cases is relatively small. The low standard errors in comparison with the basal area by thinning level indicate a statistical difference between the target basal level treatments for any given age for the overwhelming majority of treatments. This also shows that the divergence of basal level between repetitions or plots within a given target basal level treatment is minimal and is further evidence that the integrity of the target basal level treatment has been maintained over time.

Height growth

Height growth characteristics can be examined in Table 2. One important fact to note is the low standard errors, or values in the column titled StdErr, which are associated with the mean values. Almost all standard errors are equal to or less than .5. For instance, for age 45 and target basal area 7 m²/ha, a mean height of 26.4 m (87 ft) has only a .5 standard error associated with it. This trend is true across all years and target basal areas. This indicates that the mean heights of plots (or replications) within each target basal area treatments were very similar. This can be explained by three reasons. First, the loblolly pine planted in the original plantation were most likely of local provenance and genetically similar. The growth characteristics of all trees were equivalent. Second, all plots fall on a very uniform site. The site is flat, and with very similar soils across all plots. And third, plots have a uniform application of target basal level treatment. As was seen earlier in Table 1, basal level by thinning level closely corresponded with target basal level.

A second fact to note is that the tallest trees are in the target basal area of 12 m²/ha. For age 45 the tallest average height of 27.8 m (91 ft) is in this target basal area treatment, not the least dense target basal area of 7 m²/ha which had an average tree height of 26.4 m (87 ft). A model of this relationship between height and target basal area for age 30 is presented in Figure 2. With increasing basal area, heights increase until approximately 14 m²/ha target basal area and then begin to diminish. At lower basal areas trees had no need to compete for sunlight by growing upward and had the luxury of expanding their crowns outward, until 14 m²/ha basal area is reached. At this point height is maximized, and with increasing densities competition becomes limiting to height growth. This is in contrast to tree diameters, in which the largest diameters are found in the least dense treatments. Diameters progress downward as stands become more dense.

The third fact to note is the uniformity of heights. By year 45 there is only a 3.8 m (12 ft) average height difference of 27.8 vs. 24 m ( 91 vs. 79 ft) between a low density target basal area of 12 m²/ha with an actual 12.5 m²/ha (54 ft²/ac) after basal level by thinning level and the control treatment which had an after basal level by thinning level of 33.5 m²/ha (146 ft²/ac) (Table 1). And among target basal area treatments (excluding the control treatment), differences in heights are negligible. This can easily explained by loblolly pine’s shade intolerance. In dense control plots, dominant trees self pruned their lower branches and invested their resources in their terminal leader. Suppressed trees soon died and, hence, no longer lowered average heights. However, when we look at diameters there will substantial differences among treatments.

Figure 2. Heights produced at given target basal area for age 30.



Table 2. Average height after thinning (m) with associated standard error.

Diameter growth

As was true with average heights in the previous section, average diameters also have low standard errors as can be seen in the mean columns in Table 3. And the reasons for these low standard error are the same, the genetic similarity of the study trees, site uniformity, and target basal area treatment uniformity. However, while average heights were relatively uniform, average diameters diverge greatly. For instance, at age 43 target basal area of 7 m²/ha had an average diameter after thinning of 61.0 cm (24") while in the same year the control section had an average diameter of 34.3 cm (13"). The least dense target basal level treatment had nearly double the tree diameter of the dense, unthinned control section. This is due to the shade intolerant nature of loblolly pine. Having a thin stem will not necessarily result in immediate death, being over-topped by other trees will. Any tree which invests growth into diameter instead of height will soon be eliminated from the stand. However, once the competition is eliminated diameter growth is no longer a risk to survival, and is, in fact, a plus as far as wind-throw is concerned.



Table 3. Average diameter after thinning (cm) with associated standard error.

Volume growth

Stand volume is an important consideration for commercial plantation forestry. Perhaps the first thing one might notice is that the greatest volumes, located in the mean columns of Table 4, are in the control section. For example, for age 30 (a typical Arkansas pine plantation rotation age for small sawtimber) the control section has a volume of 302.1 m³/ha, while the least dense target basal area treatment of 7 m²/ha has a volume of 76.8 m³/ha. In terms of sheer volume, unthinned plots had the greatest return.

However, at age 30 the control plots had an average diameter of 27.4 cm (11"), as can be seen in Table 3. The cut-off for small sawtimber diameters is 25.4 cm (10"). Only 49% of the control plot

Table 4. Stand volume after thinning (m³/ha) with associated standard error.

Figure 3. Merchantable volumes produced at given target basal areas for age 30.

trees would be of sawtimber size at age 30. When the control section volumes are calculated with only merchantable stems, 302.1 m³/ha drops to 148 m³/ha. The target basal treatment of 21 m²/ha at year 30 had an average diameter of 34.7 cm (14"). Considering the homogenous qualities of loblolly pine growth, the great majority (93% actual) of trees in this target basal area treatment (21 m²/ha) would be of sawtimber size. Also, for year 30 this target basal area treatment (21 m²/ha) also yields the greatest volumes when only merchantable stems are included. In fact, for merchantable volumes, the target basal treatment of 21 m²/ha outdoes the 16 m²/ha treatment by a volume of 203 m³/ha to 177.8 m³/ha. Considering the much greater economic value from dimension lumber, the 21 m²/ha target basal area is an indicator of the optimal density for loblolly pine plantations.

In Figure 3 the relationship between merchantable volume and target basal area for age 30 is displayed in a graph. The points of the parabolic curve represents a model of the merchantable volume produced at a given target basal area. The stars represent the actual merchantable volumes produced at the basal level by thinning levels for age 30. The area under the curve which is highlighted represents the recommended range of 16 to 28 m²/ha (70 to 122 ft²/ac) in target basal area in order to maximize merchantable volume production. This range in target basal area represents a 3% difference in volume produced (using average from 174 to 187 m³/ha) between the target basal area at the center of the range and those on the periphery. In other words, there is some latitude in maintaining target basal area without substantially reducing merchantable volume production.

Pruning levels

Perhaps the most note worthy comment that can be made on pruning level and volume has nothing to do with pruning levels at all. If one examines the volumes at age 15 for the different pruning level treatments and the control (Table 5), all are about 93 m³/ha. At age 16 the 25 percent pruning level treatment had dropped to 55.5 m³/ha. At age 16 an ice storm caused severe damage to 3 of the plots or replications within the 25 percent pruning level treatment. Inspection of the table indicates that not until age 35 did the 25 percent pruning treatment recover a volume on par with the other pruning level treatments and control. Otherwise volumes are virtually identical, with the exception of a slight surge in the control section for ages 37 through 45.

Table 5. Stand volume after thinning with associated standard error.

There are comments that can be made about pruning level and grades of lumber yielded. At the age 40 thinning, a mill study by Dr. David Patterson was conducted on 54 felled trees, and volumes and grades of dimension lumber were recorded. The data for this study can be found at the Monticello Thinning and Pruning Study web page listed in the abstract of this paper. In Table 6, for all butt logs, if any pruning treatment had been applied then 45% of the yield was of the higher grades. For the control 34% of the material was of higher grades. This would indicate an improvement in quality, at least for the portion of the tree receiving pruning treatment. According to Dr. Patterson the pruning substantially increased the quality of lumber below the pruning level of 1.5 logs, but reduced quality above the pruning level due to increased branching.


Table 6. Percent yield by grade by pruning level for butt logs

The results can be further broken down by looking at each individual pruning level treatment. The 25 percent pruning level indicates that a 25 percent live crown was left after pruning and a 40 or 50 pruning level indicates a 40 or 50 percent residual live crown. In other words, the smaller percent of live crown, the greater the height of pruning. A pattern of increasing percent of high grades of lumber with higher levels of pruning can be seen in Table 7. With an increasing height of pruning, the percent yield of higher grades of lumber goes from the control value of 16% , to 21% for a 50% live crown, to 28% for a 40% live crown, and finally to 32% high grade lumber for the 25% live crown. When current average price prices for high and low grade sawtimber are applied to an ideal volume of 175 m³/ha for a 30 year rotation, an estimate can be made of the increased returns from different levels of pruning. There is a $4,000/ha difference between the greatest level of pruning and the control as can been seen in the extreme right column of Table 7.

Table 7. Percent yield by pruning level grade category for all logs. Price per thousand board feet of high grade lumber (C and D) is $1,050, low $330

CONCLUSIONS

The optimal target basal area for the production of sawtimber would be 22 m²/ha with an acceptable range of 16 to 28 m²/ha in order maximize volumes of merchantable lumber.

A mill study of 54 harvested stems conducted at age 40 showed pruning increased quality in the pruned portion of the stem (1 ½ log), but increased branch size above the pruning. Pruning didn’t detract from volume production. Pruning can increase value by up to $4,000 per hectare.


Acknowledgements

Our appreciation goes to Dr. David Patterson for his advice and personal communications, to Dr. Jamie Schuler for his suggestions and edits, and to Dr. James D. Burton who established this study so many decades ago.

Publications based on the reported plots

Bragg, D., M.G. Shelton, and B. Zeide. 2003. Impacts and management implications of ice storms on forests in the southern United States. Forest Ecology and Management 186: 99–123.
Leduc, D. J., and B. Zeide. 1987. The effect of thinning and pruning on the growth of planted loblolly pine stands. In the Fourth Biennial Southern Silvicultural Research Conference, p. 473-482. Compiled by D. R. Phillips. USDA For. Serv. Gen. Tech. Rep. SE-42. 598 p.
Leduc, D. J., and B. Zeide. 1987. Development of loblolly pine stands at various levels of density and pruning. Arkansas Farm Research 36(3):3.
Patterson, D.W. Personal communication. Several occasions September through October 2006.
Patterson, D.W. 2000. Does pruning the lower logs affect the lumber value of the upper logs? Presentation Abstracts. Forest Products Society 54 th Annual Meeting. South Lake Tahoe, Nevada.
Wiley, S., and B. Zeide. 1989. Thirty-year development of loblolly pine stands at various densities. In the Fifth Biennial Southern Silvicultural Research Conference, p. 199-204. Compiled by J. H. Miller. USDA For. Serv. Gen. Tech. Rep. SO-74. 618 p.
Wiley, S., and B. Zeide. 1990. Optimal density in loblolly pine: results of 30 years of growth. Arkansas Farm Research 39(1):5.
Wiley, S., and B. Zeide. 1991. Investigation of growth 14 years after glaze damage in a loblolly pine plantation. In the Sixth Biennial Southern Silvicultural Research Conference, volume 1, p. 272-281. Compiled and edited by S. S. Coleman and D. G. Neary. USDA For. Serv. Gen. Tech. Rep. SE-70. 504 p.
Wiley, S., and B. Zeide. 1992. Development of a loblolly pine plantation. Arkansas Agricultural Experiment Station Report Series 322. 44 p.
Zeide, B. 1999. Long-term observations: from trials and errors to process modeling. Pages 3-18 in C. Kleinn and M. Kohl, editors. Long Term Observations and Experiments in Forestry. Proceedings of IUFRO S4.11 International Symposium, CATIE, Costa Rica, February 23-26, 1999. 291 p.
Zeide, B., and D. Sharer. 2000. Good forestry at a glance: a guide for managing even-aged loblolly pine stands. Arkansas Agricultural Experiment Station, Arkansas Forest Resources Center Series 003. 19 p.
Zeide, B., and D. Sharer. 2001. Loblolly pine plantation management in a nutshell. 2001
Manual Forest Landowner 60(2): 10-15.
Zeide, B. 2002. Density and the growth of even-aged stands. Forest Science. 48:743-754.
Zeide, B. 2002. Sharing data. Forestry Chronicle 78(1): 152-153.
Zeide, B. and D. Sharer. 2002. Sustainable and profitable management of even-aged loblolly pine stands. Journal of Sustainable Forestry 14(1):93-106.
Zeide, B. 2004. Optimal stand density: a solution. Canadian Journal of Forest Research 34: 846–854.
Zeide, B. 2005. How to measure stand density. Tree - Structure and Function 19(1):1-14.
Zeide, B. 2006. Evolution of silvicultural thinning: from rejection to transcendence. Pages 322-327 in Connor, K.F., editor. Proceedings of the 13th Biennial Southern Silvicultural Research Conference. USDA Forest Service Gen.Tech.Rep. SRS-92. 640 p.