- Email: [email protected]
Maximum response of loblolly pine plantations to silvicultural management in the southern United States
Forest Ecology and Management 375 (2016) 105–111
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Maximum response of loblolly pine plantations to silvicultural management in the southern United States Dehai Zhao a,⇑, Michael Kane a, Robert Teskey a, Thomas R. Fox b, Timothy J. Albaugh b, H. Lee Allen c, Rafael Rubilar d a
Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA 30602, USA Department of Forest Resources and Environmental Conservation, Virginia Polytechnic and State University, Blacksburg, VA 24061, USA Department of Forestry and Environmental Resources, North Carolina State University, Raleigh, NC 27695, USA d Departamento de Silvicultura, Facultad de Ciencias Forestales, Universidad de Concepción, Concepción, Chile b c
a r t i c l e
i n f o
Article history: Received 25 February 2016 Received in revised form 17 May 2016 Accepted 22 May 2016
Keywords: Loblolly pine Maximum response Meta-analysis Plantation productivity Site quality Site-specific silviculture
a b s t r a c t Pine plantations in the southern US are among the most intensively managed forests in the world and their productivity has tripled over natural pine forests through application of intensive pine plantation establishment and management practices. As we are trying to increase carbon (C) sequestration through further enhancing pine plantation productivity by refinement of silvicultural regimes, whether a maximum productivity or the maximum potential C sequestration exists remains unclear. Our analysis of six long-term field trials indicated that a maximum productivity and a maximum response to silvicultural practices for loblolly pine (Pinus taeda L.) exist across the species geographic range in the southern US. The maximum response was inversely proportional to the base site quality, and silvicultural treatments never increased productivity above that maximum. Further analysis of loblolly pine culture and density studies demonstrated that the effects of planting density and cultural treatment intensity on biomass production strongly interacted with site quality in that lower quality sites responded more to silvicultural intensity than higher quality sites. The results highlight that we can optimize silvicultural prescriptions for specific sites by changing silvicultural intensity depending on the site quality. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction In the southern US, there are 87 million hectares of forestland including about 10 million hectares of naturally regenerated pine forests and 14 million hectares of pine plantations. These forests provide a wide range of environmental, social and economic values and services. They not only produce about 16% of global industrial wood, which is more than any other country (Prestemon and Abt, 2002), but also contain 36% of the sequestered forest carbon (C) in the conterminous US (Turner et al., 1995). Pine forests in the region annually sequester 76 Tg C, equivalent to 13% of regional greenhouse gas emissions (Johnsen et al., 2001). Over the past 50 years, the productivity of pine plantations has tripled (Fox et al., 2007b; Jokela et al., 2010) over natural pine forests due to silvicultural manipulations including site preparation, weeding, and fertilization, and the development of genetically improved seedling stock (Martin and Jokela, 2004; Zhao et al., 2009a,b, 2011). Pine plantations in the southern US have been among the most intensively ⇑ Corresponding author. E-mail address: [email protected] (D. Zhao). http://dx.doi.org/10.1016/j.foreco.2016.05.035 0378-1127/Ó 2016 Elsevier B.V. All rights reserved.
managed forests in the world. Now, researchers and forest landowners are trying to increase C sequestration by changing silvicultural practices to further enhance productivity of pine plantations. To determine the maximum potential carbon sequestration possible in these forests and to increase the productivity, profitability and sustainability of plantation silviculture requires answers to two fundamental questions: (a) How much more can forest productivity be increased by additional silvicultural treatments in the southern US? (b) Does the growth response to increased silvicultural intensity vary across the region? A substantial amount of research has focused on increasing productivity in loblolly pine plantations, and growth and yield responses to common silvicultural inputs have been documented (Borders and Bailey, 2001; Nilsson and Allen, 2003; Albaugh et al., 2004; Jokela et al., 2004; Fox et al., 2007a, 2007b; Zhao et al., 2008, 2009a, 2009b, 2011). However, the maximum achievable productivity has not been reported, with the exception of Farnum et al. (1983) who used a theoretical model to estimate a maximum potential productivity of loblolly pine on high quality sites in the Lower Coastal Plain of North Carolina to be 30 Mg ha1 y1.
106
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
The overall response to silvicultural treatments across sites has been commonly reported in literature. However, this approach disregards the response variation among different locations and soils. Site-specific response information can be very informative, but not often reported. Albaugh et al. (2015) observed that pine response to fertilization was influenced by soil drainage and texture, and only about 36% of sites had a significant treatment effect. In a long-term (26 year) study, Zhao et al. (2009b) found that complete competition control and repeated fertilization consistently and significantly increased slash pine productivity, with greater responses occurring on Spodosols than non-Spodosols. They also found that some site preparation treatments had significant effects only on non-Spodosols, while other treatments had no significant effect in plantations on either soil group. These findings suggested that there were complex responses of pines to silvicultural treatments, which appeared to be determined by resource limitations and sitespecific conditions. Various growth measures such as average height, average dbh, stand basal area, stand volume, and biomass have been used to describe the response to silvicultural treatments. However, different response patterns have been obtained, depending on which measure was used (Zhao et al., 2008, 2009b). Many important stand characteristics, such as volume and biomass, are related to stand age, stocking density and site quality. It is oftentimes hard to use volume or biomass to compare the results from different studies which have different ages and stocking. To avoid this problem, the change in site index can be used as the measure of response to management regimes. Site index is determined from the height of dominant and co-dominant trees in a stand at a base age, thus is little affected by stand density (Clutter et al., 1983). It is also determined for a fixed age, so that variable is also eliminated, leaving only site quality. Site index serves as an indicator of site quality and is a key driver in most growth and yield models. It is also a species-specific measure of actual or potential forest productivity (usually for even-aged stands). So a change in site index of a stand can indicate the change in productivity caused by different silvicultural practices. The objective of this study was to address two fundamental questions mentioned above, using a meta-analysis of six longterm field studies designed to obtain growth data for loblolly pine plantations managed at different levels of silvicultural intensity. Our hypotheses were that: (1) there is a maximum productivity of loblolly pine plantations in the southern US; (2) there is a maximum response to silvicultural treatments and the response is inversely related to site quality; and (3) there are significant interactions among silvicultural intensity, planting density and site quality. The first two hypotheses were tested by the change in site index using a linear quantile regression approach. The third one was tested by the change in biomass production using a mixedeffects modeling approach.
2. Materials and methods 2.1. Data The data came from six field studies that provided long-term growth data under carefully controlled conditions for loblolly pine plantations managed at varying levels of silvicultural intensity which are described below and summarized in Table 1. (1) The Plantation Management Research Cooperative (PMRC) Coastal Plain culture/density study (CPCD), which was established in 1995/1996 at 17 locations in the Lower Coastal Plain of Georgia (GA), Florida (FL) and South Carolina (SC), across five soil groups (Zhao et al., 2010). At each
location, there were six levels of planting densities (741, 1485, 2224, 2965, 3706 and 4448 trees ha1) and two levels of cultural treatment (intensive and operational). A splitplot design with one replication was used in which cultural intensities were randomly assigned to main plots and within a cultural intensity level the planting densities were randomly assigned subplots. In the CPCD study, the operational treatment consisted of bedding in the spring followed by a fall banded chemical site preparation. The intensive cultural treatment included bedding in the spring followed by a fall broadcast chemical site preparation, tip moth control through the first two growing seasons and repeated herbicide applications to achieve complete vegetation control throughout the rotation. At planting, 561 kg ha1 of 10-1010 fertilizer was applied on all plots. The operation treatment plots were fertilized with the equivalent of 224 kg N ha1 and 28 kg P ha1 before 8th and 12th growing seasons. The intensive cultural treatment plots also received 673 kg ha1 of 10-10-10 plus micronutrients and 131 kg ha1 of NH4NO3 in the spring of 3rd growing season, 131 kg ha1 NH4NO3 in the spring of 4th growing season, 336 kg ha1 NH4NO3 in the spring of 6th growing season, and 224 kg N ha1 and 28 kg P ha1 in the spring of 8th, 10th, 12th and 14th growing seasons. (2) The PMRC Upper Coastal Plain and Piedmont culture/density study (UPCD) which was established in 1997/1998 at 23 locations in the Upper Coastal Plain and Piedmont regions of GA, FL, SC, Alabama (AL) and Mississippi (MS), across seven broad soil classes (Zhao et al., 2010). Planting density treatments and experimental design were identical and the two cultural treatments somewhat different from that of the CPCD study. In the UPCD study, all tillage treatments included in site preparation were carried out on all treatment plots. Both the operational and intensive treatments included a broadcast chemical site preparation. The operational treatment included a first-year banded weed control. The intensive cultural treatment included additional herbicide application for complete competing vegetation control throughout the rotation. The same level of fertilization as in the CPCD study was applied in the operational and intensive treatments. (3) The PMRC Western Gulf culture/density study (WGCD) was established in 2001/2002/2003 at 18 locations in Arkansas (AR), Louisiana (LA), MS and Texas (TX), across four soil types. Each installation consisted of 10 plots, each plot representing a unique combination of five levels of planting density (494, 1112, 1730, 2347, 2963 trees ha1) and two levels of cultural intensity (operational vs intensive). The operational regime included soil group specific, mechanical site preparation, tip moth control during the first two growing seasons, and competition control and fertilization during the first growing season. The intensive regime included operational treatments plus additional competition control and fertilization treatments as described in Kane et al. (2015). (4) The Consortium for Accelerated Pine Production Studies (CAPPS) was initiated in 1987 with 6 installations in GA and monitored through age 25 years (Borders and Bailey, 2001). Each installation had six blocks in which four treatment plots were assigned one of four treatments: (1) herbicide (controlling competing woody and herbaceous vegetation using herbicides through the rotation); (2) fertilization (280 kg ha1 diammonium phosphate plus 112 kg ha1 potassium chloride applied in the spring followed by 56 kg ha1 diammonium phosphate applied in midsummer in the first and second year after planting, and
107
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111 Table 1 Information about the six long-term loblolly pine field studies. The abbreviations of these studies were identified in Section 2, and age refers to recent measurement age. Study
Experimental design
Treatments
Locations
Age
Reference
CPCD
Split-plot design
Zhao et al. (2010)
Split-plot design
16
Zhao et al. (2010)
WGCD
Split-plot design
12
Kane et al. (2015)
CAPPS
Randomized complete block design Randomized complete block design 2 2 factorial design
17 locations in GA, FL, SC 23 locations in GA, FL, SC, AL, MS 18 locations in AR, LA, MS, TX 6 locations in GA
18
UPCD
Combination of two levels of culture (Operational vs intensive) and six levels of planting density (741, 1483, 2224, 2965, 3706, 4448 trees ha1) Combination of two levels of culture (Operational vs intensive) and six levels of planting density (741, 1483, 2224, 2965, 3706, 4448 trees ha1) Combination of two levels of culture (Operational vs intensive) and five level of planting density (494, 1112, 1730, 2347, 2963 trees ha1) Four treatments (fertilization vegetation control), planted at 1600 trees ha1 Incomplete factorial of nutrient dose and frequency applied in young plantations (2–6 years old; 1216–2246 trees ha1) Four treatments (Irrigation fertilization) with density of 1176 trees ha1
25
22 trials in AL, FL, GA, LA, MS, NC, SC, VA, TX 1 location in NC
16
Borders and Bailey (2001) Albaugh et al. (2015)
29
Albaugh et al. (2004)
RG18 SETRES
168 kg ha1 ammonium nitrate applied in each spring since age three; (3) herbicide and fertilization (a combination of the herbicide treatment and fertilization treatment); and (4) control (no treatment following a double bedding site preparation). (5) The Forest Productivity Cooperative’s Regionwide 18 study (RG18) was a nutrient dose and application frequency study installed in juvenile (aged 2–6 years old) loblolly pine stands at 22 sites in AR, GA, FL, LA, MS, North Carolina (NC), SC, TX, and Virginia (VA), across six soil groups (Albaugh et al., 2015). Installation occurred from 1998 to 2003. At each site, eight or nine treatments were implemented where nitrogen was applied at different rates (0, 67, 134, 268 kg ha1) and frequencies (0, 1, 2, 4 and 6 years) in two or four replications. Phosphorus was added with nitrogen at amounts 0.1 times the nitrogen rate and other elements were added based on foliar nutrient analysis to insure that nutrient imbalances were not induced with treatment. (6) The Southeast Tree Research and Education Site (SETRES) was installed in NC in 1992 in an 8-year-old loblolly pine stand with a density of 1176 trees ha1 on infertile, somewhat excessively drained soils. The experimental design was a 2 2 factorial of two nutrition treatments (no fertilization or optimum fertilization based on foliar nutrient levels) and two irrigation treatments (with and without irrigation) in a randomized complete block, with four blocks (Albaugh et al., 2004). The planting stock used for these studies was genetically improved first or second generation open-pollinated families. Additional information about the experimental designs and
associated silvicultural treatments used in these studies can be found in the related references. These studies were established on major physiographic regions of the southern US across a variety of soil types (Fig. 1). Although these studies had different experimental designs with different silvicultural manipulations, each installation in all the studies had control plots (or operational treatment plots) and more intensive treatment plots. Pooling data from these studies provided a unique dataset for estimating growth responses to silvicultural treatments across the southern US. 2.2. Data calculation and statistical analysis The most recent measurement of non-thinned plots from the six studies was used to calculate the expressed site index with the function reported by Diéguez-Aranda et al. (2006). Site index (SI) was expressed in terms of average height of dominant and codominant trees at an index age of 25 years. We used the definition of dominant and codominant trees as the tallest 80% of trees in the stands. The base site index (SIbase ) was defined as the site index for operational plots planted at 1483 trees ha1 in the CPCD and UPCD studies, for operational plots planted at 1112 trees ha1 in the WGCD study, and for control plots in the CAPPS, RW18 and SETRES studies. The site index response was defined as the change in site index due to increased silvicultural intensity (DSI ¼ SI SIbase ), that is, the difference of the expressed site index of the more intensively managed plots minus the SIbase in the CPCD, UPCD and WGCD studies, or the difference of the expressed site index of plots with silvicultural treatment minus the SIbase within the same installation and block in the CAPPS, RG18 and SETRES studies.
Fig. 1. General locations of installations in the six designed field trials across the loblolly pine native distribution range in the southern US.
108
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
To investigate the effect of silvicultural activities on productivity, the scatter plot of the DSI and SIbase was first used to illustrate the relationship. Then, a linear quantile regression was used to ðsÞ
ðsÞ
quantify the relationship: DSI ¼ b0 þ b1 SIbase þ eðsÞ . The slope ðsÞ b1
for base site index describes the relationship between the change in site index and base site index for a given quantile s. Based on the estimated quantile regression, we could calculate the increase in site index due to silviculture on sites with different base site indices for selected quantiles. We defined the maximum response as the 0.99 conditional quantile for a given SIbase . A change in site index indicates a change in potential forest productivity from silvicultural management. To demonstrate the interaction of silvicultural treatment and site quality on forest productivity, we examined total aboveground biomass production in unthinned installations of the CPCD and UPCD studies. At age 15, there were a total of 21 installations that had not been thinned in these two studies. For each installation, the base site index was estimated from the operational plot planted at 1483 trees/ ha. Based on the base site index, 7 installations (7 installations 2 cultural intensities 6 planting densities = 84 plots) were grouped into a low-site-quality class (L: SI < 24.4 m), 6 installations (6 2 6 = 72 plots) were grouped into an intermediate-sitequality class (M: 24.4 6 SI < 27.4 m), and 8 installations (8 2 6 = 96 plots) were grouped into a high-site-quality class (H: SI P 27.4 m). Total aboveground biomass at age 15 was estimated with tree biomass equations developed by Zhao et al. (2015). Analysis of variance with a mixed-effects model for a split-split plot experimental design was used to examine effects of cultural intensity, planting density, site quality, and their interactions on stand-level aboveground biomass production. These main effects and interactions were treated as fixed-effects, while the installation effect and the interaction of installation and cultural intensity were treated as random-effects.
Fig. 2. A scatterplot of 850 observations of change in site index (DSI) and the base site index (SIbase ) with 0.99, 0.90, and 0.50 regression quantile estimates for the ^ ðs Þ þ b ^ðsÞ SIbase . The negative dependence of DSI on SIbase indicates a model DSI ¼ b 0 1 large potential site index response to silviculture on poor quality sites and a small response on high quality sites. The maximum response line: DSI ¼ 25:74 0:81SIbase is defined as the 0.99 conditional quantile given the base site index. The median (0.50) and 0.90 quantile responses are DSI ¼ 4:46 0:15SIbase and DSI ¼ 13:62 0:41SIbase , respectively.
3. Results There was a negative relationship between DSI and SIbase , indicating that as the base site index increased there was less response to growth-promoting treatments and that there was a maximum site index, around 32 m for loblolly pine in the southern US (Fig. 2). On the sites with that maximum site index, productivity did not increase with increased silvicultural inputs. Both DSI and its variation became smaller as the SIbase increased, which was described by the linear quantile regression. Negative values of ðsÞ
the slope b1 were significantly different from 0 across all quantiles (ss) (Fig. 3), indicating that the potential increase in site index due to silvicultural treatments was inversely proportional to the base site index. Interestingly, 26% of the 850 plots that received increased silvicultural inputs did not respond to treatment or even had a negativeDSI. These nonresponse or negative response plots comprised 7.5%, 13.0%, 27.4%, 18.0%, 36.0% and 16.7% of plots in the CPCD, UPCD, WGCD, CPPS, RG18, and SETRES, respectively. There was a clear maximum response to silviculture (Fig. 2), with the greatest maximum response on lower quality sites. The maximum response line is DSI ¼ 25:74 0:81SIbase , as defined by the 0.99 conditional quantile given the base site index. On sites with a site index of 30 m, for example, the maximum increase in site index due to silviculture is only about 1.4 m; on sites with a site index of 20 m, the maximum response is around 9.5 m (Table 2). The median (s ¼ 0:5) response, ranged from 1.56 to 0.11 at SIbase of 20 and 30 m, respectively (Table 2, Fig. 2). In general, the expressed maximum site index after the application of additional treatments was less than the maximum site index of 32 m, especially on low quality sites (Fig. 4).
ðs Þ
Fig. 3. Estimates of the slope (b1 ) in the quantile regression between the change in ðsÞ ðs Þ site index and base site index: DSI ¼ b0 þ b1 SIbase þ eðsÞ . The gray band shows the 95% confidence limits. The solid red line is the slope and dashed red lines connect endpoints of 95% confidence intervals estimated from least square regression for the model: DSI ¼ b0 þ b1 SIbase þ e.
Table 2 Mean increase in site index due to silviculture for base site indexes of 20, 25 and 30 m at age 25 years for selected quantiles (s). Base site index
s = 0.5
s = 0.75
s = 0.95
s = 0.99
20 25 30
1.56 0.83 0.11
3.14 1.87 0.60
6.45 4.11 1.77
9.52 5.46 1.41
The main effects of site quality, cultural intensity, and planting density, and the interaction of site quality and cultural intensity on total aboveground biomass of loblolly pine stands at age 15 were
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
109
higher stocking generally produced more aboveground biomass. However, the differences in total aboveground biomass among low, medium and high planting densities did not vary significantly among site quality classes. 4. Discussion
Fig. 4. The scatterplot of expressed site index for the plots with increased silviculture and the base site index for the corresponding control or operational plots. The red line (25:74 þ 0:19SIbase ) was drawn from the maximum response line (DSI ¼ 25:74 0:81SIbase ). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Table 3 The p-values of main fixed effects and their interactions on total aboveground biomass of loblolly pine stands at age 15. Source
p-value
Site quality class Cultural intensity Site quality and cultural intensity Planting density Site quality and planting density Cultural intensity and planting density
0.0101 <0.0001 0.0052 <0.0001 0.4751 0.1981
significant (Table 3). When the biomass growth response was compared among three levels of site quality, it was apparent that this was due to the substantial biomass response to silvicultural inputs found on low-quality sites (Fig. 5). On high-quality sites, the biomass response to silvicultural inputs was not significant. Higher stockings generally produced more aboveground biomass. Overall,
There were three key findings in this study: an apparent maximum expressed site index of loblolly pine of about 32 m at age 25; a decrease in the growth response (site index or biomass response) to silvicultural management as the base site index increased; and an apparent lack of a growth response, or a decrease in growth, after silvicultural treatments were applied to some plots across the range of base site indexes used in our analysis. The positive responses to silvicultural treatments observed on many sites in this study are consistent with results from past research on the positive impacts of competition control (Creighton et al., 1987; Miller et al., 2003) and fertilization (Fox et al., 2007a) on loblolly pine productivity on many sites. The apparent lack of a positive response in site index or decrease in expressed site index associated with additional silvicultural treatments on some sites may result from a number of causes. The variability in site quality between a base plot and its counterpart receiving the additional silvicultural treatment may affect the calculated response. If the plot receiving the additional silvicultural treatment has a somewhat higher or lower site quality than its counterpart plot not receiving treatment, the calculated response may be greater or less than the true response. Some of the negative responses in site index associated with additional treatment, especially those of 1 m site index or less, likely reflect site quality variability between plots. The lack of a marked positive response to additional cultural treatments on some sites may be because the treatment did not provide limiting resources (e.g. fertilization on a site where nutrients are not limiting productivity or vegetation control on a site with no or very low competition vegetation) or the trees are limited by some resources other than that provided by the treatment (e.g. fertilization on a site where soil moisture strongly limits growth). Albaugh et al. (2015) analyzed annual volume growth eight years after the RG18 study initiation and found about 64% of examined sites had no significant treatment effects. The cause of the apparent marked negative site index
Fig. 5. Aboveground biomass by cultural intensity, planting density, and site quality for loblolly pine plantations at age 15. Two levels of cultural intensity: operational (Op) and intensive (Int); Three levels of site quality: low-quality (L: SI < 24.4 m), intermediate-quality (M: 24.4 6 SI < 27.4 m); high-quality sites (H: SI P 27.4 m).
110
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
responses on some sites is unknown. Clearly we need to better understand why these apparent negative responses occurred. Possible causes include plot to plot site variability, pest activity, weather events, and treatment quality. Our first hypothesis that there is a maximum productivity of loblolly pine plantations in the southern US was supported by the results. The maximum SI (around 32 m at base age of 25 years) observed in this study appears to confirm the existence of a maximum potential productivity. Farnum et al. (1983) defined a pretty high maximum potential productivity (30 Mg ha1 yr1 biomass) by using a theoretical model based on light interception and observations from existing stands. Aboveground biomass at age 15 years in plots of the CPCD and UPCD studies ranged from 70 to 234 Mg ha1 and averaged 166 Mg ha1 (also see Fig. 5). Mean annual biomass increment ranged from 5 to 16 Mg ha1 yr1, depending on site quality, planting density, and cultural intensity. It is clear that even under more intensive management with genetically improved planting stock on high quality sites, loblolly pine achieved far less than that maximum potential productivity. In addition, our maximum level of productivity was achieved only on the high quality sites. Farnum et al. (1983) noted that the potential maximum productivity that they predicted was likely to be about 2.5 times higher than the actual measured maximum productivity in stands in the fields because of a variety of biotic and abiotic limitations to productivity. That is consistent with a maximum annual biomass increment of 15 Mg ha1 predicted using the 3-PG process model for a highly productive loblolly pine plantation located on a moist, weed-free and annually fertilized site in Waycross GA (Bryars et al., 2013). The parameterization of 3-PG benefited from much more extensive information about the physiology of loblolly pine than available to Farnum et al. (1983) but the two approaches converged on a similar maximum productivity for actual plantations in the field. The large difference between the actual maximum productivity of 15 Mg ha1 y1 (Bryars et al., 2013) or 16 Mg ha1 y1 (this study) and a theoretical maximum productivity of 30 Ma ha1 y1 suggests that there are a great many factors, including variability in weather conditions, effects of insects and diseases, competition and soil restriction that combine to limit the actual maximum loblolly pine productivity. In addition, how limiting resources change over time also needs to be explored (Farrior et al., 2013). Our second hypothesis, that there is a maximum response to increased silvicultural treatments which is inversely related to site quality, was supported. Especially with respect to sites of lower base SI, silvicultural treatments moved stands toward the maximum productivity but never increased the realized SI above that maximum (Fig. 4). It was shown that the overall response to silvicultural treatments diminished as base SI increased. In addition, when the base SI values were less than 20, the silvicultural treatments examined did not bring the expressed site index to come near the 99 percentile response line (Figs. 2 and 4). In fact, the potential for silvicultural improvement of yield, using currently available silvicultural options, appears limited on low base SI sites. This suggests that intensive silviculture on the poorest sites may be an inefficient use of resources. Our results also indicate that additional silvicultural inputs on high base quality sites has limited impact on expressed site index and therefore is an inefficient use of resources. An adjustment of silvicultural intensity based on potential maximum yield gain, with consideration of market factors, would optimize the use of chemicals, energy and time and still produce the improved growth rates. It may be possible to increase growth further at low and medium SI sites with new silvicultural regimes, more effective fertilizer amendments, competition control, further genetic improvement and different site preparation techniques. Those improvements may move sites closer to the maximum SI of 32 m, but they are unlikely to surpass it
based on currently available growth data. Whether future changes in the climate, or increases in atmospheric [CO2], can affect the maximum SI remains an unanswered question. The results also partially supported our third hypothesis that site quality significantly interacted with planting density and cultural intensity to affect biomass production, as the site quality interaction with cultural intensity was significant and the site quality interaction with planting density was not significant. The next step will be to identify which silvicultural practices were most effective on low, medium and high base SI sites. However, the development of site-specific silvicultural prescriptions was beyond the scope of this paper, but could have large potential benefits to loblolly pine plantation managers. We suggest that the analysis we have done using loblolly pine plantations has potential to improve the cost-effectiveness and efficiency of management practices used in other plantation tree species. 5. Conclusions There are important ramifications to the observed inverse relationship between the SIbase and the response to increased silvicultural inputs: (1) A maximum means that the productivity potential and thus carbon sequestration potential of loblolly pine plantations may be finite and can be defined; (2) site-specific silvicultural practices can be employed with both economic efficiencies and environmental benefits of reduced fertilizer use on high SI sites; (3) the cost of producing wood and fiber can be reduced by lowering the rotation age on high SI sites; and (4) genotypes can be developed to take advantage of high resource availability from more intensive management on low SI sites. Finally, investigating the biological reasons for a maximum limit to productivity will provide new insights into forest production ecology that will pertain to tree plantation species and perhaps forests in general. Acknowledgments This work was financially supported by USDA National Institute of Food and Agriculture (Agreement No. 2011-68002-30185). The authors appreciate the efforts of the Plantation Management Research Cooperative (PMRC) at the University of Georgia and the Forest Productivity Cooperative (FPC) at North Carolina State University and Virginia Polytechnic and State University in establishing, maintaining, and monitoring the field studies. References Albaugh, T.J., Allen, H.L., Dougherty, P.M., Johnsen, K.H., 2004. Long term growth responses of loblolly pine to optimal nutrient and water resource availability. For. Ecol. Manage. 192, 3–19. Albaugh, T.J., Fox, T.R., Allen, H.L., Rubilar, R.A., 2015. Juvenile southern pine response to fertilization is influenced by soil drainage and texture. Forests 6, 2799–2819. Borders, B.E., Bailey, R.L., 2001. Loblolly pine – pushing the limits of growth. South. J. Appl. For. 25, 69–73. Bryars, C., Maier, C., Zhao, D., Kane, M., Borders, B., Will, R., Teskey, R., 2013. Fixed physiological parameters in the 3-PG model produced accurate estimates of loblolly pine growth on sites in different geographic regions. For. Ecol. Manage. 289, 501–514. Clutter, J.L., Fortson, J.C., Pienaar, L.V., Brister, G.H., Bailey, R.L., 1983. Timber Management: A Quantitative Approach. John Wiley & Sons, New York. Creighton, J.L., Zutter, B.R., Glover, G.R., Gjerstad, D.H., 1987. Planted pine growth and survival response to herbaceous vegetation control, treatment duration, and herbicide application technique. South. J. Appl. For. 11 (4), 223–228. Diéguez-Aranda, U., Burkhart, H.E., Amateis, R.L., 2006. Dynamic site model for loblolly pine (Pinus taeda L.) plantations in the United States. For. Sci. 52, 262– 272. Farnum, P., Timmis, R., Kulp, J.L., 1983. Biotechnology of forest Yield. Science 219, 694–702. Farrior, C.E., Tilman, D., Dybzinski, R., Reich, P.B., Levin, S.A., Pacala, S.W., 2013. Resource limitation in a competitive context determines complex plant responses to experimental resource additions. Ecology 94 (11), 2505–2517.
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111 Fox, T.R., Allen, H.L., Albaugh, T.J., Rubilar, R., Carlson, C.A., 2007a. Tree nutrition and forest fertilization of the pine plantations in the southern United States. South. J. Appl. For. 31, 5–11. Fox, T.R., Jokela, E.J., Allen, H.L., 2007b. The development of pine plantation silviculture in the southern United States. J. For. 105, 337–347. Johnsen, K.H., Wear, D., Oren, R., Teskey, R.O., Sanchez, F., Will, R., Butnor, J., Markewitz, D., Richter, D., Rials, T., Allen, H.L., Seiler, J., Ellsworth, D., Maier, C., Katul, G., Dougherty, P.M., 2001. Meeting global policy commitments: Carbon sequestration and southern pine forests. J. For. 99 (4), 14–21. Jokela, E.J., Dougherty, P.M., Martin, T.A., 2004. Production dynamics of intensively managed loblolly pine stands in the southern United States: a synthesis of seven long-term experiments. For. Ecol. Manage. 192 (1), 117–130. Jokela, E.J., Martin, T.A., Vogel, J.G., 2010. Twenty-five years of intensive forest management with southern pines: important lessons learned. J. Forest. 108 (7), 338–347. Kane, M., Wang, M., Rheney, J., Zhao, D., 2015. Western Gulf Culture Density Study: Results Through Age 12 PMRC Technical Report 2015-1. Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA, p. 55. Martin, T.A., Jokela, E.J., 2004. Stand development and production dynamics of loblolly pine under a range of cultural treatments in north-central Florida USA. For. Ecol. Manage. 192, 39–58. Miller, J.H., Zutter, B.R., Zedaker, S.M., Edwards, M.B., Newbold, R.A., 2003. Growth and yield relative to competition for loblolly pine plantations to midrotation – a southeastern United States regional study. South. J. Appl. For. 27 (4), 237–252. Nilsson, U., Allen, H.L., 2003. Short- and long-term effects of site preparation, fertilization and vegetation control on growth and stand development of planted loblolly pine. For. Ecol. Manage. 175, 367–377.
111
Prestemon, J.P., Abt, R.C., 2002. Timber products supply and demand. In: Wear, D.N., Greis, J.G. (Eds.), The Southern Forest Resources Assessment. USDA Forest Service Southern Research Station, Asheville, NC, pp. 299–325. Turner, D.P., Koerper, G.J., Harmon, M.E., Lee, J.J., 1995. A carbon budget for forests of the conterminous United States. Ecol. Appl. 5, 421–436. Zhao, D., Kane, M., Borders, B.E., Harrison, M., 2008. Pine growth response to different site-preparation methods with and without post-plant herbaceous weed control on North Florida’s Lower Coastal Plain. For. Ecol. Manage. 255, 2512–2523. Zhao, D., Kane, M., Borders, B.E., Harrison, M., Rheney, J.W., 2009a. Site preparation and competing vegetation control affect loblolly pine long-term productivity in the southern Piedmont/Upper Coastal Plain of the United States. Ann. For. Sci. 66 (705). Zhao, D., Kane, M., Borders, B.E., Harrison, M., 2009b. Long-term effects of site preparation treatments, complete competition control, and repeated fertilization on growth of slash pine plantations in the flatwoods of the southeastern United States. For. Sci. 55, 403–410. Zhao, D., Kane, M., Borders, B.E., 2010. Development and applications of the relative spacing model for loblolly pine plantations. For. Ecol. Manage. 259, 1922–1929. Zhao, D., Kane, M., Borders, B.E., 2011. Growth responses to planting density and management intensity in loblolly pine plantations in the southeastern USA Lower Coastal Plain. Ann. For. Sci. 68, 625–635. Zhao, D., Kane, M., Markewitz, D., Teskey, R., Clutter, M., 2015. Additive tree biomass equations for midrotation loblolly pine plantations. For. Sci. 61 (4), 613–623.
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Maximum response of loblolly pine plantations to silvicultural management in the southern United States Dehai Zhao a,⇑, Michael Kane a, Robert Teskey a, Thomas R. Fox b, Timothy J. Albaugh b, H. Lee Allen c, Rafael Rubilar d a
Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA 30602, USA Department of Forest Resources and Environmental Conservation, Virginia Polytechnic and State University, Blacksburg, VA 24061, USA Department of Forestry and Environmental Resources, North Carolina State University, Raleigh, NC 27695, USA d Departamento de Silvicultura, Facultad de Ciencias Forestales, Universidad de Concepción, Concepción, Chile b c
a r t i c l e
i n f o
Article history: Received 25 February 2016 Received in revised form 17 May 2016 Accepted 22 May 2016
Keywords: Loblolly pine Maximum response Meta-analysis Plantation productivity Site quality Site-specific silviculture
a b s t r a c t Pine plantations in the southern US are among the most intensively managed forests in the world and their productivity has tripled over natural pine forests through application of intensive pine plantation establishment and management practices. As we are trying to increase carbon (C) sequestration through further enhancing pine plantation productivity by refinement of silvicultural regimes, whether a maximum productivity or the maximum potential C sequestration exists remains unclear. Our analysis of six long-term field trials indicated that a maximum productivity and a maximum response to silvicultural practices for loblolly pine (Pinus taeda L.) exist across the species geographic range in the southern US. The maximum response was inversely proportional to the base site quality, and silvicultural treatments never increased productivity above that maximum. Further analysis of loblolly pine culture and density studies demonstrated that the effects of planting density and cultural treatment intensity on biomass production strongly interacted with site quality in that lower quality sites responded more to silvicultural intensity than higher quality sites. The results highlight that we can optimize silvicultural prescriptions for specific sites by changing silvicultural intensity depending on the site quality. Ó 2016 Elsevier B.V. All rights reserved.
1. Introduction In the southern US, there are 87 million hectares of forestland including about 10 million hectares of naturally regenerated pine forests and 14 million hectares of pine plantations. These forests provide a wide range of environmental, social and economic values and services. They not only produce about 16% of global industrial wood, which is more than any other country (Prestemon and Abt, 2002), but also contain 36% of the sequestered forest carbon (C) in the conterminous US (Turner et al., 1995). Pine forests in the region annually sequester 76 Tg C, equivalent to 13% of regional greenhouse gas emissions (Johnsen et al., 2001). Over the past 50 years, the productivity of pine plantations has tripled (Fox et al., 2007b; Jokela et al., 2010) over natural pine forests due to silvicultural manipulations including site preparation, weeding, and fertilization, and the development of genetically improved seedling stock (Martin and Jokela, 2004; Zhao et al., 2009a,b, 2011). Pine plantations in the southern US have been among the most intensively ⇑ Corresponding author. E-mail address: [email protected] (D. Zhao). http://dx.doi.org/10.1016/j.foreco.2016.05.035 0378-1127/Ó 2016 Elsevier B.V. All rights reserved.
managed forests in the world. Now, researchers and forest landowners are trying to increase C sequestration by changing silvicultural practices to further enhance productivity of pine plantations. To determine the maximum potential carbon sequestration possible in these forests and to increase the productivity, profitability and sustainability of plantation silviculture requires answers to two fundamental questions: (a) How much more can forest productivity be increased by additional silvicultural treatments in the southern US? (b) Does the growth response to increased silvicultural intensity vary across the region? A substantial amount of research has focused on increasing productivity in loblolly pine plantations, and growth and yield responses to common silvicultural inputs have been documented (Borders and Bailey, 2001; Nilsson and Allen, 2003; Albaugh et al., 2004; Jokela et al., 2004; Fox et al., 2007a, 2007b; Zhao et al., 2008, 2009a, 2009b, 2011). However, the maximum achievable productivity has not been reported, with the exception of Farnum et al. (1983) who used a theoretical model to estimate a maximum potential productivity of loblolly pine on high quality sites in the Lower Coastal Plain of North Carolina to be 30 Mg ha1 y1.
106
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
The overall response to silvicultural treatments across sites has been commonly reported in literature. However, this approach disregards the response variation among different locations and soils. Site-specific response information can be very informative, but not often reported. Albaugh et al. (2015) observed that pine response to fertilization was influenced by soil drainage and texture, and only about 36% of sites had a significant treatment effect. In a long-term (26 year) study, Zhao et al. (2009b) found that complete competition control and repeated fertilization consistently and significantly increased slash pine productivity, with greater responses occurring on Spodosols than non-Spodosols. They also found that some site preparation treatments had significant effects only on non-Spodosols, while other treatments had no significant effect in plantations on either soil group. These findings suggested that there were complex responses of pines to silvicultural treatments, which appeared to be determined by resource limitations and sitespecific conditions. Various growth measures such as average height, average dbh, stand basal area, stand volume, and biomass have been used to describe the response to silvicultural treatments. However, different response patterns have been obtained, depending on which measure was used (Zhao et al., 2008, 2009b). Many important stand characteristics, such as volume and biomass, are related to stand age, stocking density and site quality. It is oftentimes hard to use volume or biomass to compare the results from different studies which have different ages and stocking. To avoid this problem, the change in site index can be used as the measure of response to management regimes. Site index is determined from the height of dominant and co-dominant trees in a stand at a base age, thus is little affected by stand density (Clutter et al., 1983). It is also determined for a fixed age, so that variable is also eliminated, leaving only site quality. Site index serves as an indicator of site quality and is a key driver in most growth and yield models. It is also a species-specific measure of actual or potential forest productivity (usually for even-aged stands). So a change in site index of a stand can indicate the change in productivity caused by different silvicultural practices. The objective of this study was to address two fundamental questions mentioned above, using a meta-analysis of six longterm field studies designed to obtain growth data for loblolly pine plantations managed at different levels of silvicultural intensity. Our hypotheses were that: (1) there is a maximum productivity of loblolly pine plantations in the southern US; (2) there is a maximum response to silvicultural treatments and the response is inversely related to site quality; and (3) there are significant interactions among silvicultural intensity, planting density and site quality. The first two hypotheses were tested by the change in site index using a linear quantile regression approach. The third one was tested by the change in biomass production using a mixedeffects modeling approach.
2. Materials and methods 2.1. Data The data came from six field studies that provided long-term growth data under carefully controlled conditions for loblolly pine plantations managed at varying levels of silvicultural intensity which are described below and summarized in Table 1. (1) The Plantation Management Research Cooperative (PMRC) Coastal Plain culture/density study (CPCD), which was established in 1995/1996 at 17 locations in the Lower Coastal Plain of Georgia (GA), Florida (FL) and South Carolina (SC), across five soil groups (Zhao et al., 2010). At each
location, there were six levels of planting densities (741, 1485, 2224, 2965, 3706 and 4448 trees ha1) and two levels of cultural treatment (intensive and operational). A splitplot design with one replication was used in which cultural intensities were randomly assigned to main plots and within a cultural intensity level the planting densities were randomly assigned subplots. In the CPCD study, the operational treatment consisted of bedding in the spring followed by a fall banded chemical site preparation. The intensive cultural treatment included bedding in the spring followed by a fall broadcast chemical site preparation, tip moth control through the first two growing seasons and repeated herbicide applications to achieve complete vegetation control throughout the rotation. At planting, 561 kg ha1 of 10-1010 fertilizer was applied on all plots. The operation treatment plots were fertilized with the equivalent of 224 kg N ha1 and 28 kg P ha1 before 8th and 12th growing seasons. The intensive cultural treatment plots also received 673 kg ha1 of 10-10-10 plus micronutrients and 131 kg ha1 of NH4NO3 in the spring of 3rd growing season, 131 kg ha1 NH4NO3 in the spring of 4th growing season, 336 kg ha1 NH4NO3 in the spring of 6th growing season, and 224 kg N ha1 and 28 kg P ha1 in the spring of 8th, 10th, 12th and 14th growing seasons. (2) The PMRC Upper Coastal Plain and Piedmont culture/density study (UPCD) which was established in 1997/1998 at 23 locations in the Upper Coastal Plain and Piedmont regions of GA, FL, SC, Alabama (AL) and Mississippi (MS), across seven broad soil classes (Zhao et al., 2010). Planting density treatments and experimental design were identical and the two cultural treatments somewhat different from that of the CPCD study. In the UPCD study, all tillage treatments included in site preparation were carried out on all treatment plots. Both the operational and intensive treatments included a broadcast chemical site preparation. The operational treatment included a first-year banded weed control. The intensive cultural treatment included additional herbicide application for complete competing vegetation control throughout the rotation. The same level of fertilization as in the CPCD study was applied in the operational and intensive treatments. (3) The PMRC Western Gulf culture/density study (WGCD) was established in 2001/2002/2003 at 18 locations in Arkansas (AR), Louisiana (LA), MS and Texas (TX), across four soil types. Each installation consisted of 10 plots, each plot representing a unique combination of five levels of planting density (494, 1112, 1730, 2347, 2963 trees ha1) and two levels of cultural intensity (operational vs intensive). The operational regime included soil group specific, mechanical site preparation, tip moth control during the first two growing seasons, and competition control and fertilization during the first growing season. The intensive regime included operational treatments plus additional competition control and fertilization treatments as described in Kane et al. (2015). (4) The Consortium for Accelerated Pine Production Studies (CAPPS) was initiated in 1987 with 6 installations in GA and monitored through age 25 years (Borders and Bailey, 2001). Each installation had six blocks in which four treatment plots were assigned one of four treatments: (1) herbicide (controlling competing woody and herbaceous vegetation using herbicides through the rotation); (2) fertilization (280 kg ha1 diammonium phosphate plus 112 kg ha1 potassium chloride applied in the spring followed by 56 kg ha1 diammonium phosphate applied in midsummer in the first and second year after planting, and
107
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111 Table 1 Information about the six long-term loblolly pine field studies. The abbreviations of these studies were identified in Section 2, and age refers to recent measurement age. Study
Experimental design
Treatments
Locations
Age
Reference
CPCD
Split-plot design
Zhao et al. (2010)
Split-plot design
16
Zhao et al. (2010)
WGCD
Split-plot design
12
Kane et al. (2015)
CAPPS
Randomized complete block design Randomized complete block design 2 2 factorial design
17 locations in GA, FL, SC 23 locations in GA, FL, SC, AL, MS 18 locations in AR, LA, MS, TX 6 locations in GA
18
UPCD
Combination of two levels of culture (Operational vs intensive) and six levels of planting density (741, 1483, 2224, 2965, 3706, 4448 trees ha1) Combination of two levels of culture (Operational vs intensive) and six levels of planting density (741, 1483, 2224, 2965, 3706, 4448 trees ha1) Combination of two levels of culture (Operational vs intensive) and five level of planting density (494, 1112, 1730, 2347, 2963 trees ha1) Four treatments (fertilization vegetation control), planted at 1600 trees ha1 Incomplete factorial of nutrient dose and frequency applied in young plantations (2–6 years old; 1216–2246 trees ha1) Four treatments (Irrigation fertilization) with density of 1176 trees ha1
25
22 trials in AL, FL, GA, LA, MS, NC, SC, VA, TX 1 location in NC
16
Borders and Bailey (2001) Albaugh et al. (2015)
29
Albaugh et al. (2004)
RG18 SETRES
168 kg ha1 ammonium nitrate applied in each spring since age three; (3) herbicide and fertilization (a combination of the herbicide treatment and fertilization treatment); and (4) control (no treatment following a double bedding site preparation). (5) The Forest Productivity Cooperative’s Regionwide 18 study (RG18) was a nutrient dose and application frequency study installed in juvenile (aged 2–6 years old) loblolly pine stands at 22 sites in AR, GA, FL, LA, MS, North Carolina (NC), SC, TX, and Virginia (VA), across six soil groups (Albaugh et al., 2015). Installation occurred from 1998 to 2003. At each site, eight or nine treatments were implemented where nitrogen was applied at different rates (0, 67, 134, 268 kg ha1) and frequencies (0, 1, 2, 4 and 6 years) in two or four replications. Phosphorus was added with nitrogen at amounts 0.1 times the nitrogen rate and other elements were added based on foliar nutrient analysis to insure that nutrient imbalances were not induced with treatment. (6) The Southeast Tree Research and Education Site (SETRES) was installed in NC in 1992 in an 8-year-old loblolly pine stand with a density of 1176 trees ha1 on infertile, somewhat excessively drained soils. The experimental design was a 2 2 factorial of two nutrition treatments (no fertilization or optimum fertilization based on foliar nutrient levels) and two irrigation treatments (with and without irrigation) in a randomized complete block, with four blocks (Albaugh et al., 2004). The planting stock used for these studies was genetically improved first or second generation open-pollinated families. Additional information about the experimental designs and
associated silvicultural treatments used in these studies can be found in the related references. These studies were established on major physiographic regions of the southern US across a variety of soil types (Fig. 1). Although these studies had different experimental designs with different silvicultural manipulations, each installation in all the studies had control plots (or operational treatment plots) and more intensive treatment plots. Pooling data from these studies provided a unique dataset for estimating growth responses to silvicultural treatments across the southern US. 2.2. Data calculation and statistical analysis The most recent measurement of non-thinned plots from the six studies was used to calculate the expressed site index with the function reported by Diéguez-Aranda et al. (2006). Site index (SI) was expressed in terms of average height of dominant and codominant trees at an index age of 25 years. We used the definition of dominant and codominant trees as the tallest 80% of trees in the stands. The base site index (SIbase ) was defined as the site index for operational plots planted at 1483 trees ha1 in the CPCD and UPCD studies, for operational plots planted at 1112 trees ha1 in the WGCD study, and for control plots in the CAPPS, RW18 and SETRES studies. The site index response was defined as the change in site index due to increased silvicultural intensity (DSI ¼ SI SIbase ), that is, the difference of the expressed site index of the more intensively managed plots minus the SIbase in the CPCD, UPCD and WGCD studies, or the difference of the expressed site index of plots with silvicultural treatment minus the SIbase within the same installation and block in the CAPPS, RG18 and SETRES studies.
Fig. 1. General locations of installations in the six designed field trials across the loblolly pine native distribution range in the southern US.
108
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
To investigate the effect of silvicultural activities on productivity, the scatter plot of the DSI and SIbase was first used to illustrate the relationship. Then, a linear quantile regression was used to ðsÞ
ðsÞ
quantify the relationship: DSI ¼ b0 þ b1 SIbase þ eðsÞ . The slope ðsÞ b1
for base site index describes the relationship between the change in site index and base site index for a given quantile s. Based on the estimated quantile regression, we could calculate the increase in site index due to silviculture on sites with different base site indices for selected quantiles. We defined the maximum response as the 0.99 conditional quantile for a given SIbase . A change in site index indicates a change in potential forest productivity from silvicultural management. To demonstrate the interaction of silvicultural treatment and site quality on forest productivity, we examined total aboveground biomass production in unthinned installations of the CPCD and UPCD studies. At age 15, there were a total of 21 installations that had not been thinned in these two studies. For each installation, the base site index was estimated from the operational plot planted at 1483 trees/ ha. Based on the base site index, 7 installations (7 installations 2 cultural intensities 6 planting densities = 84 plots) were grouped into a low-site-quality class (L: SI < 24.4 m), 6 installations (6 2 6 = 72 plots) were grouped into an intermediate-sitequality class (M: 24.4 6 SI < 27.4 m), and 8 installations (8 2 6 = 96 plots) were grouped into a high-site-quality class (H: SI P 27.4 m). Total aboveground biomass at age 15 was estimated with tree biomass equations developed by Zhao et al. (2015). Analysis of variance with a mixed-effects model for a split-split plot experimental design was used to examine effects of cultural intensity, planting density, site quality, and their interactions on stand-level aboveground biomass production. These main effects and interactions were treated as fixed-effects, while the installation effect and the interaction of installation and cultural intensity were treated as random-effects.
Fig. 2. A scatterplot of 850 observations of change in site index (DSI) and the base site index (SIbase ) with 0.99, 0.90, and 0.50 regression quantile estimates for the ^ ðs Þ þ b ^ðsÞ SIbase . The negative dependence of DSI on SIbase indicates a model DSI ¼ b 0 1 large potential site index response to silviculture on poor quality sites and a small response on high quality sites. The maximum response line: DSI ¼ 25:74 0:81SIbase is defined as the 0.99 conditional quantile given the base site index. The median (0.50) and 0.90 quantile responses are DSI ¼ 4:46 0:15SIbase and DSI ¼ 13:62 0:41SIbase , respectively.
3. Results There was a negative relationship between DSI and SIbase , indicating that as the base site index increased there was less response to growth-promoting treatments and that there was a maximum site index, around 32 m for loblolly pine in the southern US (Fig. 2). On the sites with that maximum site index, productivity did not increase with increased silvicultural inputs. Both DSI and its variation became smaller as the SIbase increased, which was described by the linear quantile regression. Negative values of ðsÞ
the slope b1 were significantly different from 0 across all quantiles (ss) (Fig. 3), indicating that the potential increase in site index due to silvicultural treatments was inversely proportional to the base site index. Interestingly, 26% of the 850 plots that received increased silvicultural inputs did not respond to treatment or even had a negativeDSI. These nonresponse or negative response plots comprised 7.5%, 13.0%, 27.4%, 18.0%, 36.0% and 16.7% of plots in the CPCD, UPCD, WGCD, CPPS, RG18, and SETRES, respectively. There was a clear maximum response to silviculture (Fig. 2), with the greatest maximum response on lower quality sites. The maximum response line is DSI ¼ 25:74 0:81SIbase , as defined by the 0.99 conditional quantile given the base site index. On sites with a site index of 30 m, for example, the maximum increase in site index due to silviculture is only about 1.4 m; on sites with a site index of 20 m, the maximum response is around 9.5 m (Table 2). The median (s ¼ 0:5) response, ranged from 1.56 to 0.11 at SIbase of 20 and 30 m, respectively (Table 2, Fig. 2). In general, the expressed maximum site index after the application of additional treatments was less than the maximum site index of 32 m, especially on low quality sites (Fig. 4).
ðs Þ
Fig. 3. Estimates of the slope (b1 ) in the quantile regression between the change in ðsÞ ðs Þ site index and base site index: DSI ¼ b0 þ b1 SIbase þ eðsÞ . The gray band shows the 95% confidence limits. The solid red line is the slope and dashed red lines connect endpoints of 95% confidence intervals estimated from least square regression for the model: DSI ¼ b0 þ b1 SIbase þ e.
Table 2 Mean increase in site index due to silviculture for base site indexes of 20, 25 and 30 m at age 25 years for selected quantiles (s). Base site index
s = 0.5
s = 0.75
s = 0.95
s = 0.99
20 25 30
1.56 0.83 0.11
3.14 1.87 0.60
6.45 4.11 1.77
9.52 5.46 1.41
The main effects of site quality, cultural intensity, and planting density, and the interaction of site quality and cultural intensity on total aboveground biomass of loblolly pine stands at age 15 were
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
109
higher stocking generally produced more aboveground biomass. However, the differences in total aboveground biomass among low, medium and high planting densities did not vary significantly among site quality classes. 4. Discussion
Fig. 4. The scatterplot of expressed site index for the plots with increased silviculture and the base site index for the corresponding control or operational plots. The red line (25:74 þ 0:19SIbase ) was drawn from the maximum response line (DSI ¼ 25:74 0:81SIbase ). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Table 3 The p-values of main fixed effects and their interactions on total aboveground biomass of loblolly pine stands at age 15. Source
p-value
Site quality class Cultural intensity Site quality and cultural intensity Planting density Site quality and planting density Cultural intensity and planting density
0.0101 <0.0001 0.0052 <0.0001 0.4751 0.1981
significant (Table 3). When the biomass growth response was compared among three levels of site quality, it was apparent that this was due to the substantial biomass response to silvicultural inputs found on low-quality sites (Fig. 5). On high-quality sites, the biomass response to silvicultural inputs was not significant. Higher stockings generally produced more aboveground biomass. Overall,
There were three key findings in this study: an apparent maximum expressed site index of loblolly pine of about 32 m at age 25; a decrease in the growth response (site index or biomass response) to silvicultural management as the base site index increased; and an apparent lack of a growth response, or a decrease in growth, after silvicultural treatments were applied to some plots across the range of base site indexes used in our analysis. The positive responses to silvicultural treatments observed on many sites in this study are consistent with results from past research on the positive impacts of competition control (Creighton et al., 1987; Miller et al., 2003) and fertilization (Fox et al., 2007a) on loblolly pine productivity on many sites. The apparent lack of a positive response in site index or decrease in expressed site index associated with additional silvicultural treatments on some sites may result from a number of causes. The variability in site quality between a base plot and its counterpart receiving the additional silvicultural treatment may affect the calculated response. If the plot receiving the additional silvicultural treatment has a somewhat higher or lower site quality than its counterpart plot not receiving treatment, the calculated response may be greater or less than the true response. Some of the negative responses in site index associated with additional treatment, especially those of 1 m site index or less, likely reflect site quality variability between plots. The lack of a marked positive response to additional cultural treatments on some sites may be because the treatment did not provide limiting resources (e.g. fertilization on a site where nutrients are not limiting productivity or vegetation control on a site with no or very low competition vegetation) or the trees are limited by some resources other than that provided by the treatment (e.g. fertilization on a site where soil moisture strongly limits growth). Albaugh et al. (2015) analyzed annual volume growth eight years after the RG18 study initiation and found about 64% of examined sites had no significant treatment effects. The cause of the apparent marked negative site index
Fig. 5. Aboveground biomass by cultural intensity, planting density, and site quality for loblolly pine plantations at age 15. Two levels of cultural intensity: operational (Op) and intensive (Int); Three levels of site quality: low-quality (L: SI < 24.4 m), intermediate-quality (M: 24.4 6 SI < 27.4 m); high-quality sites (H: SI P 27.4 m).
110
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111
responses on some sites is unknown. Clearly we need to better understand why these apparent negative responses occurred. Possible causes include plot to plot site variability, pest activity, weather events, and treatment quality. Our first hypothesis that there is a maximum productivity of loblolly pine plantations in the southern US was supported by the results. The maximum SI (around 32 m at base age of 25 years) observed in this study appears to confirm the existence of a maximum potential productivity. Farnum et al. (1983) defined a pretty high maximum potential productivity (30 Mg ha1 yr1 biomass) by using a theoretical model based on light interception and observations from existing stands. Aboveground biomass at age 15 years in plots of the CPCD and UPCD studies ranged from 70 to 234 Mg ha1 and averaged 166 Mg ha1 (also see Fig. 5). Mean annual biomass increment ranged from 5 to 16 Mg ha1 yr1, depending on site quality, planting density, and cultural intensity. It is clear that even under more intensive management with genetically improved planting stock on high quality sites, loblolly pine achieved far less than that maximum potential productivity. In addition, our maximum level of productivity was achieved only on the high quality sites. Farnum et al. (1983) noted that the potential maximum productivity that they predicted was likely to be about 2.5 times higher than the actual measured maximum productivity in stands in the fields because of a variety of biotic and abiotic limitations to productivity. That is consistent with a maximum annual biomass increment of 15 Mg ha1 predicted using the 3-PG process model for a highly productive loblolly pine plantation located on a moist, weed-free and annually fertilized site in Waycross GA (Bryars et al., 2013). The parameterization of 3-PG benefited from much more extensive information about the physiology of loblolly pine than available to Farnum et al. (1983) but the two approaches converged on a similar maximum productivity for actual plantations in the field. The large difference between the actual maximum productivity of 15 Mg ha1 y1 (Bryars et al., 2013) or 16 Mg ha1 y1 (this study) and a theoretical maximum productivity of 30 Ma ha1 y1 suggests that there are a great many factors, including variability in weather conditions, effects of insects and diseases, competition and soil restriction that combine to limit the actual maximum loblolly pine productivity. In addition, how limiting resources change over time also needs to be explored (Farrior et al., 2013). Our second hypothesis, that there is a maximum response to increased silvicultural treatments which is inversely related to site quality, was supported. Especially with respect to sites of lower base SI, silvicultural treatments moved stands toward the maximum productivity but never increased the realized SI above that maximum (Fig. 4). It was shown that the overall response to silvicultural treatments diminished as base SI increased. In addition, when the base SI values were less than 20, the silvicultural treatments examined did not bring the expressed site index to come near the 99 percentile response line (Figs. 2 and 4). In fact, the potential for silvicultural improvement of yield, using currently available silvicultural options, appears limited on low base SI sites. This suggests that intensive silviculture on the poorest sites may be an inefficient use of resources. Our results also indicate that additional silvicultural inputs on high base quality sites has limited impact on expressed site index and therefore is an inefficient use of resources. An adjustment of silvicultural intensity based on potential maximum yield gain, with consideration of market factors, would optimize the use of chemicals, energy and time and still produce the improved growth rates. It may be possible to increase growth further at low and medium SI sites with new silvicultural regimes, more effective fertilizer amendments, competition control, further genetic improvement and different site preparation techniques. Those improvements may move sites closer to the maximum SI of 32 m, but they are unlikely to surpass it
based on currently available growth data. Whether future changes in the climate, or increases in atmospheric [CO2], can affect the maximum SI remains an unanswered question. The results also partially supported our third hypothesis that site quality significantly interacted with planting density and cultural intensity to affect biomass production, as the site quality interaction with cultural intensity was significant and the site quality interaction with planting density was not significant. The next step will be to identify which silvicultural practices were most effective on low, medium and high base SI sites. However, the development of site-specific silvicultural prescriptions was beyond the scope of this paper, but could have large potential benefits to loblolly pine plantation managers. We suggest that the analysis we have done using loblolly pine plantations has potential to improve the cost-effectiveness and efficiency of management practices used in other plantation tree species. 5. Conclusions There are important ramifications to the observed inverse relationship between the SIbase and the response to increased silvicultural inputs: (1) A maximum means that the productivity potential and thus carbon sequestration potential of loblolly pine plantations may be finite and can be defined; (2) site-specific silvicultural practices can be employed with both economic efficiencies and environmental benefits of reduced fertilizer use on high SI sites; (3) the cost of producing wood and fiber can be reduced by lowering the rotation age on high SI sites; and (4) genotypes can be developed to take advantage of high resource availability from more intensive management on low SI sites. Finally, investigating the biological reasons for a maximum limit to productivity will provide new insights into forest production ecology that will pertain to tree plantation species and perhaps forests in general. Acknowledgments This work was financially supported by USDA National Institute of Food and Agriculture (Agreement No. 2011-68002-30185). The authors appreciate the efforts of the Plantation Management Research Cooperative (PMRC) at the University of Georgia and the Forest Productivity Cooperative (FPC) at North Carolina State University and Virginia Polytechnic and State University in establishing, maintaining, and monitoring the field studies. References Albaugh, T.J., Allen, H.L., Dougherty, P.M., Johnsen, K.H., 2004. Long term growth responses of loblolly pine to optimal nutrient and water resource availability. For. Ecol. Manage. 192, 3–19. Albaugh, T.J., Fox, T.R., Allen, H.L., Rubilar, R.A., 2015. Juvenile southern pine response to fertilization is influenced by soil drainage and texture. Forests 6, 2799–2819. Borders, B.E., Bailey, R.L., 2001. Loblolly pine – pushing the limits of growth. South. J. Appl. For. 25, 69–73. Bryars, C., Maier, C., Zhao, D., Kane, M., Borders, B., Will, R., Teskey, R., 2013. Fixed physiological parameters in the 3-PG model produced accurate estimates of loblolly pine growth on sites in different geographic regions. For. Ecol. Manage. 289, 501–514. Clutter, J.L., Fortson, J.C., Pienaar, L.V., Brister, G.H., Bailey, R.L., 1983. Timber Management: A Quantitative Approach. John Wiley & Sons, New York. Creighton, J.L., Zutter, B.R., Glover, G.R., Gjerstad, D.H., 1987. Planted pine growth and survival response to herbaceous vegetation control, treatment duration, and herbicide application technique. South. J. Appl. For. 11 (4), 223–228. Diéguez-Aranda, U., Burkhart, H.E., Amateis, R.L., 2006. Dynamic site model for loblolly pine (Pinus taeda L.) plantations in the United States. For. Sci. 52, 262– 272. Farnum, P., Timmis, R., Kulp, J.L., 1983. Biotechnology of forest Yield. Science 219, 694–702. Farrior, C.E., Tilman, D., Dybzinski, R., Reich, P.B., Levin, S.A., Pacala, S.W., 2013. Resource limitation in a competitive context determines complex plant responses to experimental resource additions. Ecology 94 (11), 2505–2517.
D. Zhao et al. / Forest Ecology and Management 375 (2016) 105–111 Fox, T.R., Allen, H.L., Albaugh, T.J., Rubilar, R., Carlson, C.A., 2007a. Tree nutrition and forest fertilization of the pine plantations in the southern United States. South. J. Appl. For. 31, 5–11. Fox, T.R., Jokela, E.J., Allen, H.L., 2007b. The development of pine plantation silviculture in the southern United States. J. For. 105, 337–347. Johnsen, K.H., Wear, D., Oren, R., Teskey, R.O., Sanchez, F., Will, R., Butnor, J., Markewitz, D., Richter, D., Rials, T., Allen, H.L., Seiler, J., Ellsworth, D., Maier, C., Katul, G., Dougherty, P.M., 2001. Meeting global policy commitments: Carbon sequestration and southern pine forests. J. For. 99 (4), 14–21. Jokela, E.J., Dougherty, P.M., Martin, T.A., 2004. Production dynamics of intensively managed loblolly pine stands in the southern United States: a synthesis of seven long-term experiments. For. Ecol. Manage. 192 (1), 117–130. Jokela, E.J., Martin, T.A., Vogel, J.G., 2010. Twenty-five years of intensive forest management with southern pines: important lessons learned. J. Forest. 108 (7), 338–347. Kane, M., Wang, M., Rheney, J., Zhao, D., 2015. Western Gulf Culture Density Study: Results Through Age 12 PMRC Technical Report 2015-1. Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA, p. 55. Martin, T.A., Jokela, E.J., 2004. Stand development and production dynamics of loblolly pine under a range of cultural treatments in north-central Florida USA. For. Ecol. Manage. 192, 39–58. Miller, J.H., Zutter, B.R., Zedaker, S.M., Edwards, M.B., Newbold, R.A., 2003. Growth and yield relative to competition for loblolly pine plantations to midrotation – a southeastern United States regional study. South. J. Appl. For. 27 (4), 237–252. Nilsson, U., Allen, H.L., 2003. Short- and long-term effects of site preparation, fertilization and vegetation control on growth and stand development of planted loblolly pine. For. Ecol. Manage. 175, 367–377.
111
Prestemon, J.P., Abt, R.C., 2002. Timber products supply and demand. In: Wear, D.N., Greis, J.G. (Eds.), The Southern Forest Resources Assessment. USDA Forest Service Southern Research Station, Asheville, NC, pp. 299–325. Turner, D.P., Koerper, G.J., Harmon, M.E., Lee, J.J., 1995. A carbon budget for forests of the conterminous United States. Ecol. Appl. 5, 421–436. Zhao, D., Kane, M., Borders, B.E., Harrison, M., 2008. Pine growth response to different site-preparation methods with and without post-plant herbaceous weed control on North Florida’s Lower Coastal Plain. For. Ecol. Manage. 255, 2512–2523. Zhao, D., Kane, M., Borders, B.E., Harrison, M., Rheney, J.W., 2009a. Site preparation and competing vegetation control affect loblolly pine long-term productivity in the southern Piedmont/Upper Coastal Plain of the United States. Ann. For. Sci. 66 (705). Zhao, D., Kane, M., Borders, B.E., Harrison, M., 2009b. Long-term effects of site preparation treatments, complete competition control, and repeated fertilization on growth of slash pine plantations in the flatwoods of the southeastern United States. For. Sci. 55, 403–410. Zhao, D., Kane, M., Borders, B.E., 2010. Development and applications of the relative spacing model for loblolly pine plantations. For. Ecol. Manage. 259, 1922–1929. Zhao, D., Kane, M., Borders, B.E., 2011. Growth responses to planting density and management intensity in loblolly pine plantations in the southeastern USA Lower Coastal Plain. Ann. For. Sci. 68, 625–635. Zhao, D., Kane, M., Markewitz, D., Teskey, R., Clutter, M., 2015. Additive tree biomass equations for midrotation loblolly pine plantations. For. Sci. 61 (4), 613–623.