Citation
Effects of Silvicultural Management Intensity and Genetics on Soil Respiration and Belowground Carbon Allocation in Loblolly Pine Plantations

Material Information

Title:
Effects of Silvicultural Management Intensity and Genetics on Soil Respiration and Belowground Carbon Allocation in Loblolly Pine Plantations
Creator:
Drum, Chelsea Gill
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (157 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Forest Resources and Conservation
Committee Chair:
JOKELA,ERIC J
Committee Co-Chair:
MARTIN,TIMOTHY A
Committee Members:
GEZAN,SALVADOR
SCHUUR,EDWARD A
VOGEL,JASON
Graduation Date:
8/9/2014

Subjects

Subjects / Keywords:
Carbon ( jstor )
Fertilization ( jstor )
Forest litter ( jstor )
Forest soils ( jstor )
Forests ( jstor )
Nutrients ( jstor )
Plantations ( jstor )
Soil respiration ( jstor )
Soil science ( jstor )
Understory ( jstor )
Forest Resources and Conservation -- Dissertations, Academic -- UF
belowground -- genetics -- loblolly -- pine -- respiration -- silviculture -- soil
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Forest Resources and Conservation thesis, M.S.

Notes

Abstract:
In the southeastern U.S., fertilization and weed control treatments, along with the deployment of genetically improved planting stock, are routinely used to increase aboveground productivity. This project examined the effects of intensive management and genetic selection of loblolly pine (Pinus taeda L.) on soil respiration (SR) and belowground carbon (C) allocation. In Florida, two field installations of two families of loblolly pine, fast grower and slow grower, were studied in a replicated, family block design with two levels of nitrogen and phosphorus fertilization (high and operational intensity) on Spodosols. Measurements of root biomass and repeated measurements of forest growth, SR, and litterfall were made over multiple years. Soil respiration and litterfall measurements were used to estimate Total Belowground Carbon Flux (TBCF), which provided an estimate of C allocation to roots. Soil respiration varied temporally, with significant effects (p<0.05) of family x time, intensity x time, and location x time for both the bed and inter-bed and root exclusion positions. Heterotrophic respiration did not vary between management intensities and comprised approximately 80% of total SR. Modelled respiration rates resulted in an intensity x family interaction for SR and TBCF, indicating that TBCF did not differ for either family under intensive management, but they did differ under the operational management regime. These findings, therefore, lend support that genotype x environment interactions can occur not only aboveground but also belowground. The results also argue that genotype x environment interactions should be considered in the southeastern United States when estimating forest C budgets. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2014.
Local:
Adviser: JOKELA,ERIC J.
Local:
Co-adviser: MARTIN,TIMOTHY A.
Statement of Responsibility:
by Chelsea Gill Drum.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Resource Identifier:
968131544 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

Soil Science Society of America Journal þt This work was presented at the 12th North American Forest Soils Conference, Whitesh, MT, 16–20 June 2013, in the Production Systems for Biomass and Bioenergy session. Soil Sci. Soc. Am. J. doi:10.2136/sssaj2013.08.0345nafsc *Corresponding author (ejokela@u.edu). Received 13 Aug. 2013. © Soil Science Society of America, 5585 Guilford Rd., Madison WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.Inter-rotational Effects of Fertilization and Weed Control on Juvenile Loblolly Pine Productivity and Nutrient Dynamics North American Forest Soils Conference ProceedingsD uring the last six decades, plantation silviculture in the southern United States has gone through a series of changes that has enhanced the pro ductive capacity of southern pine stands (Jokela et al., 2004; Fox et al., 2007a). Coastal Plain pine forests in the southern United States are among the most intensively managed in the world. Fertilization and understory competition control are the common silvicultural practices used, and when combined with suitable site preparation techniques and the deployment of genetically improved planting stock, improvements in yield can range from twoto fourfold compared with extensively managed loblolly pine stands (Allen et al., 1990; Colbert et al., 1990; Neary et al., 1990a; Fox et al., 2007a). As the global population continues to increase, intensive forest management systems will undoubtedly play a major role in the U.S. Southeast and other parts Praveen Subedi Eric J. Jokela*School of Forest Resources and Conservation Univ. of Florida Gainesville, FL 32611 Jason G. VogelDep. of Ecosystem Science and Management Texas A&M Univ. College Station, TX 77843Timothy A. MartinSchool of Forest Resources and Conservation Univ. of Florida Gainesville, FL 32611 The inter-rotational effects of fertilization and weed control treatments on the productivity and soil nutrient availability of loblolly pine (Pinus taeda L.) stands growing on a North Florida Spodosol were investigated using two replicated, randomized complete block design experiments. The rst rotation treatments were: control (C), fertilizer only (F), weed control only (W), and fertilizer + weed control (FW). One experiment was actively retreated as in the previous rotation, while the second was left untreated (C C , C F , C W , and C FW ). A common full-sib loblolly pine family was planted in both experiments. After three growing seasons, the second-rotation pine growth consistently outperformed the rst rotation. The actively retreated FW treatment had fourfold higher aboveground pine biomass than the C treatment (7.7 Mg ha1 ); the untreated CF (17.9 Mg ha1 ) treatment had 1.5-fold higher pine biomass than the C FW treatment. While a suite of improved cultural practices (e.g., advanced genetics and site preparation) and environmental factors may be responsible for higher growth responses in the second rotation than the rst in both experiments, lower growth response in the CFW treatment compared with the CF treatment was associated with lower soil P availability (r = 0.8, p < 0.01) and historical P movement from the E to Bh and Bt horizons. These results suggest that the understory vegetation and forest oor from the rst rotation served as an important nutrient sink, especially for P, which then subsequently became a nutrient source (through mineralization) in the second rotation. Historical P fertilization on atwoods Spodosols may thus alleviate the need for P fertilization during stand establishment when an intact understory was present in the previous stand.Abbreviations: HSD, honestly signicant difference. Published June 30, 2014

PAGE 2

þt Soil Science Society of America Journalof the world for meeting the ever burgeoning demand for forest products and other related ecosystem services (Sedjo and Botkin, 1997). However, considering the relatively short history and use of intensive forest management systems (since the 1970s), questions arise regarding their long-term sustainability. ese con cerns generally center on long-term site productivity (Fox, 2000) and maintenance of biodiversity (Jeries et al., 2010). Understory plant communities in managed forest stands are oen thought of as competitors for aboveand belowground site resources such as water, nutrients, light, and growing space (Morris et al., 1993; Collet et al., 1996; Zutter et al., 1999; Zhang et al., 2013). From the perspective of ecosystem diver sity, however, these plant communities are important elements. ey may help support long-term site sustainability by increasing the species richness and functional diversity of the forest ecosystem and also aid in ecosystem adaptation to abiotic and biotic stresses and disturbances through, for example, nutrient cycling processes (Tilman et al., 2001; Zak et al., 2003; Folke et al., 2004; Nilsson and Wardle, 2005). Although understory competition for available soil nutrients represents one of the primary causes for lower growth rates in young southern pine stands (Neary et al., 1990b), a study conducted by Smethurst and Nambiar (1995) on the second rotation crop of Pinus radiata D. Don showed that allowing competing vegetation in the understory reduced N leaching and increased N mineralization following their senescence. On sandy soils, where soil organic C serves an important function in aiding moisture and nutrient retention, understory vegetation may also play a role in both maintaining C and serving as a nutrient sink for elements such as N and P, especially during the early stages of stand development (Boring et al., 1981; Gholz and Fisher, 1982; Gholz et al., 1985; Smethurst and Nambiar, 1995; Blazier et al., 2005; Rifai et al., 2010). Understory sup pression may reduce the potential for C and nutrient sinks in soils, as some studies have observed reductions in soil C and N pools with repeated herbicide applications (Laiho et al., 2003; Echeverría et al., 2004; Sarkhot et al., 2007; Sartori et al., 2007; Rifai et al., 2010; Vogel et al., 2011). erefore, it is important to understand whether the silvicultural treatments used to support short-rotation, intensively managed pine plantations aect un derstory reinitiation, soil C, and nutrient levels. It follows that as soil organic matter decomposes, nutrients like N and P are min eralized and made available for pine uptake, leaching, and other biogeochemical processes. Silvicultural treatments that could potentially increase or decrease soil organic matter could have the potential to inuence the long-term nutrient supply to the site (Jurgensen et al., 1997). Previous research conducted on sandy Spodosols in the U.S. Lower Coastal Plain have documented that N and P are both growth-limiting nutrient elements and that fertilization is a cost-eective treatment that forest managers can use to enhance growth and nancial returns (Pritchett and Llewellyn, 1966; Bengtson, 1979; Pritchett and Comerford, 1982; Allen, 1987; Jokela and Stearns-Smith, 1993; Fox et al., 2007b; Albaugh et al., 2009). Because higher nutrient demands associated with these intensively managed stands have the potential to induce micro nutrient deciencies (Stone, 1990; Jokela et al., 1991b; Allen et al., 2005) and limit growth (Vogel and Jokela, 2011), fertiliza tion on some sites now includes B, Cu, and Mn, in addition to N, P, and K (Jokela et al., 1991a; Albaugh et al., 2007; Vogel and Jokela, 2011). In that context, it is important to understand the role of residual nutrients from past fertilization activities, if any, to support the growth of newly planted stands. For exam ple, P has been shown to readily recycle in fertilized pine stands (Polglase et al., 1992a), and the residual P from past fertilization has the potential to meet the nutrient demands and early growth requirements of newly planted stands (Ballard, 1978; Comerford et al., 2002; Everett and Palm-Leis, 2009). Comparative studies over multiple rotations will enable direct assessment of manage ment practices on long-term site productivity. Direct assessment of long-term site productivity can be an arduous task, and the examples that exist in the literature have rarely attempted to combine both fertilization and weed control assessments (Keeves, 1966; Nambiar, 1996). Moreover, dier ences in site management and the planting stock genetics used between rotations can confound the results (Nambiar, 1996). e study described here used a replicated eld design to answer two questions: 1. In a nutrient-stressed environment, are inter-rotational growth responses of juvenile loblolly pine aected by the previous silvicultural treatment history (fertilization and weed control)? 2. Does the historical treatment of the understory vegetation community aect its competitive role in the second rotation relative to soil nutrient availability and the growth of loblolly pine? ese questions were addressed by examining and comparing the aboveground biomass accumulation, distribution, and nutrient content of the overstory and understory species across a range of silvicultural treatment histories that had varying levels of soil nutrient availability. Eorts were made to duplicate as closely as possible the rst-rotation silvicultural treatments and included using a common genetic source of loblolly pine.MATERIALS AND METHODSStudy AreaTo evaluate the factors that limit the biological growth po tential of southern pines, the Intensive Management Practices Assessment Center (IMPAC) at the University of Florida estab lished an experimental study site in 1983 (Swindel et al., 1988). e IMPAC experimental site is located approximately 10 km north of Gainesville, FL (29n 30 a N, 82n 20 a W longitude; mean elevation of 45 m). e long-term mean annual temperature (1984–2012) of the study site is 20.6n C, and it receives an annual rainfall of about 1178 mm (NOAA, 2012). e climate is warm and humid. Poorly drained Pomona ne sands (sandy, siliceous, hyperthermic Ultic Alaquods) are the predominant soils at the study site.

PAGE 3

www.soils.org/publications/sssajþt Study Designe original IMPAC experiment was a 2 q 2 q 2 factorial, with treatments of species (loblolly and slash pine [Pinus elliottii var. elliottii Engelm.]), complete and sustained weed control, and annual fertilization arranged in a randomized split-plot (species as whole plots) design with three replications. is resulted in four treatments within each species: control (C), weed control only (W), fertilizer only (F), and both fertilizer and weed control (FW). e entire experimental area was prepared using a singlepass bedding treatment. Genetically improved (rst generation, open pollinated) 1–0 bareroot stock of both loblolly and slash pine were hand planted in January 1983 (Swindel et al., 1988; Colbert et al., 1990; Martin and Jokela, 2004). A fertilizer re gime with balanced levels of macroand micronutrients was ap plied for the rst 10 yr to the F and FW treatments, aer which it was stopped in May 1993 and then resumed during the 16th to 18th growing seasons (1998–2000; Jokela and Martin, 2000). Fertilizers were applied in narrow bands (30-cm semicircle) around the base of each tree or planting location. Total nutrient additions during the life of the original study for the F and FW treatments for both species were (kg ha1 ): 1088 N, 230 P, 430 K, 108 Ca, 72 Mg, 72 S, 4.1 Mn, 5.4 Fe, 0.9 Cu, 4 Zn, and 0.9 B. Competing understory vegetation was controlled annually in the W and FW treatments for the rst 10 yr (1983–1993) using a combination of chemical and mechanical methods (Colbert et al., 1990; Neary et al., 1990b; Dalla-Tea and Jokela, 1994). e weed control treatment was stopped aer canopy closure because the growth of competing understory vegetation was suppressed in the W and FW plots. Earlyand mid-rotation growth dynam ics for the original study were reported by Colbert et al. (1990), Jokela and Martin (2000), and Martin and Jokela (2004). Jokela et al. (2010) summarized the growth dynamics for this original experiment during the 25-yr study period. Likewise, Vogel et al. (2011) documented the total C and N pools at the end of the rotation for the original experiment. e original IMPAC study was whole-tree harvested in May 2009, with the intent of overlaying a second experiment using the same treatment plots. Harvested trees were processed o the treatment plots to ensure no inputs of nutrients into the soil via harvest residues. Considering the need for long-term site monitoring to understand the eects of the past management history, the IMPAC II study site was initiated in June 2009. Original plots in the rst rotation were reestablished and used to examine both the untreated carryover and actively managed retreatment eects on the growth of the second-rotation experiment. e IMPAC II experiment now consists of two randomized complete block designs (three replications each), having four treatments (C, F, FW, and W) for the actively managed retreatment design and four treatments for the untreated carryover design (Cc, CF , C FW , C W ; subscripts indicate rst-rotation treatments, C before the subscripts indicates untreated carryover) (Fig. 1; Table 1). e carryover experiment was established on the previous slash pine plots, and the actively managed retreatment experiment was established on the previous loblolly pine plots. Before harvesting, all treatment plot corners were physically monumented and the understory vegetation in the C and F plots was mulched in place (April 2009) to retain this nutrient pool within the plot boundaries. Mulching was not necessary for the W and FW plots because of the sustained weed control treatment history from the previous rotation. Following harvest, the entire study area was later bedded in June, with a second bedding pass conducted in August of the same year. Similar to the last rotation, trees were planted in each plot at a 1.8by 3.0-m spacing, with measurement plots (0.02 ha) con sisting of 40 trees per plot (eight trees each in ve beds). Each of the measurement plots was provided with a treated buer of three trees and two beds, resulting in a 0.08-ha treatment plot. An untreated buer of six tree spaces was provided between two adjacent treatment plots. Across the treatment plots, an untreat ed buer of four beds was maintained (Fig. 1). A single, full-sib, and elite performing loblolly family was used to regenerate the entire study (i.e., actively managed and untreated experiments) in December 2009 using containerized seedlings. Before planting, only the active retreatment plots that had received chemical site preparation and weed control in the rst rotation (W and FW) were treated using a broadcast application of 0.84 kg a.e. ha1 imazapyr (2-[4,5-dihydro-4-methyl-4-(1methylethyl)-5-oxo-1H -imidazol-2-yl]-3-pyridinecarboxylic acid), 1.12 kg a.e. ha1 triclopyr (2-[(3,5,6-trichloro-2-pyridinyl) oxy]acetic acid), and 0.14 kg ha1 of metsulfuron-methyl (methyl 2-[[[[(4-methoxy-6-methyl-1,3,5-triazin-2-yl)amino]carbonyl] amino]sulfonyl]benzoate) in October 2009. In October 2010, these same plots received a directed spray application of triclopyr (3%) and imazapyr (1%) to control Ilex glabra (L.) A. Gray and Fig. 1. Layout of the actively managed retreatment experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and the untreated carryover experiment (CC , C F , C W , and C FW , respectively) on Spodosols near Gainesville, FL.

PAGE 4

þt Soil Science Society of America Journalother understory competitors. Also, in September 2011 the actively managed W and FW plots received another directed spray of 3% glyphosate [N -(phosphonomethyl)glycine] to maintain a weed-free environment. All treatments (actively managed re treated and untreated carryover) received a single application of 9.1% pronil (5-amino-1-[2,6-dichloro-4-(triuoromethyl) phenyl]-4-[(triuoromethyl)sulnyl]-1 H -pyrazole-3-carboni trile) in the form of PTM (BASF Corp.) in March 2010 to con trol Nantucket pine tip moth (Rhyacionia ustrana Comstock). e untreated carryover plots did not receive any additional chemical treatments (herbicide or fertilizer), with the exception of a banded 0.2 kg a.e. ha1 imazapyr application in May 2010 to control Dicanthelium sp. and to aid seedling survival in all treat ment plots. is same banded herbicide treatment was also ap plied to the actively managed C and F plots. e actively managed retreated (F and FW) experiment re ceived fertilizer at the end of July 2011 and beginning of September 2012. Consistent with the last rotation treatments, the total nutrient additions during the rst three growing seasons for the F and FW treatments were (kg ha1 ): 120 N, 53 P, 99 K, 40 Ca, 19 Mg, 56 S, 1.3 Mn, 0.5 Fe, 0.2 Cu, 0.5 Zn, and 0.2 B. As done in the rst rotation experiment, the fertilizer was applied in narrow bands (30-cm semicircle) around the base of each tree or planting location.Data Preparation and AnalysisEstimation of the total aboveground biomass for ages 1 to 3 yr was made using existing allometric equations previously developed for loblolly pine for the same family, ages, and soil type (inventory data and Adegbidi et al., 2002). e following equation was used to estimate the total aboveground biomass of loblolly pine using annual inventory data collected in 2011, 2012, and 2013: t n t n 01 ln ln YX CfC [1] where Y is the biomass component (kg dry wt.), X is the tree height (m) for ages 1 and 2 yr and diameter at breast height (cm) for age 3 yr, and C 0 and C 1 are the coecients of regression. For estimation of component biomass, highly signicant allometric equations of the same form as Eq. [1] were developed using the destructive harvest data for 1-, 2and 3-yr-old loblolly pines collected by Adegbidi et al. (2002) (Table 2). eir destructive harvest data were generated from the same genetic family growing on similar soils as the current study. Corrections for logarithmic bias were made on all estimates of biomass accumulation (Baskerville, 1972; Sprugel, 1983). Nutrient analyses of the pine foliage at age 2 yr were conducted by collecting fully elongated needles (approximately 25 fascicles) that were sampled from ve random trees in each measure ment plot in December 2011. ese foliar tissues were then analyzed for macroand micronutrients at the Micro-Macro International Laboratory in Athens, GA. About 0.5 g of ground tissue was rst dry-ashed in a mue furnace, and then the samples were brought up to volume with aqua regia (3:1 HNO3 /HCl). e extracts were analyzed using inductively coupled plasma atomic emission spectroscopy (ICP–AES). Total N was analyzed in a CNS analyzer (Leco Corp.) using the Dumas method (Campbell, 1992). Stem wood with bark and branch tissues was collected from four randomly selected pine trees in each measurement plot in December 2011. ese samples were oven dried at 65n C to a constant weight and then ground in a Wiley mill to pass through a 1-mm sieve. ese tissues were then analyzed for macroand micronutrients at the Micro-Macro International Laboratory using the same methods described above. Aboveground nutrient content (kg ha1 ) in loblolly pine was estimated separately for each treatment using component-part biomass estimates and av erage nutrient concentration data (Subedi, 2013). Estimates of aboveground understory biomass were made using a clip plot survey conducted in August and September 2011. Within each plot, six quadrats (1 m2 ) were randomly established in each measurement plot and stratied equally between the bed and interbed positions. All standing vegetation that fell within the quadrats was clipped at ground level and sorted separately by spe cies. For overhanging vegetation, only the portion that fell within the quadrat was clipped. All samples were oven dried at 65n C to a constant weight. e dried leaves, twigs, or branches of individual understory species in a plot were ground in a Wiley mill to pass through a 1-mm sieve. In addition, six culms (with roots) of Andropogon sp. were manually excavated from both bed and in terbed positions of the CFW and C W treatments of the untreated carryover experiment, where Andropogon sp. was predominant, to estimate nutrient accumulation in roots. Andropogon sp. samples were then separated into shoots and roots. Roots were washed with distilled water to rinse o soil particles. Both shoots and rinsed roots were oven dried at 65n C to a constant weight before subsam pling. e subsamples of shoot and root tissues were also ground in a Wiley mill to pass through a 1-mm sieve. All tissues were analyzed for macroand micronutrients at the Micro-Macro International Laboratory. Nutrient content in the understory was determined separately by species for each treatment using average nutrient con centrations and biomass estimates for each plot. Soil nutrient availability was assessed in the actively man aged and untreated carryover experiments using ion-exchange Table 1. Treatments applied to juvenile loblolly pine plantations growing on Spodosols in North Florida at the Intensive Management Practices Assessment Center (IMPAC) II study. Experiment First-rotation treatment Second-rotation treatment Treatment abbreviation Actively managed retreated control control C fertilizer only fertilizer only F fertilizer + weed control fertilizer + weed control FW weed control only weed control only W Untreated carryover control untreated C C fertilizer only untreated C F fertilizer + weed control untreated C FW weed control only untreated C W

PAGE 5

www.soils.org/publications/sssajþt membranes (PRS probes, Western Ag Innovations). Although the use of ion-exchange membranes in forest ecosystems is not as common as in agricultural ecosystems, they have been widely used to determine nutrient supply rates in the soil (Hart and Firestone, 1989; Ziadi et al., 1999; Qian and Schoenau, 2002). Four cation PRS probes capable of adsorbing all nutrient cations and four anion PRS probes capable of adsorbing all nutrient anions were buried randomly in the upper 15 cm of the beds of all mea surement plots in August 2011. Aer 8 wk of burial, these probes were removed from the soil and rinsed free of ad hering soil particles with deionized and distilled water. All probes were eluted using a 0.5 mol L1 HCl solution for 1 h. e eluate was then analyzed colorimetrically using an automated ow injection analysis system for NO3 –N and NH4 + –N to obtain the total N supply. For P, K, Ca, Mg, S, B, Cu, Mn, Zn, and Fe, ICP–AES was used. All analyses were conducted by Western Ag Innovations. Soil nutrient concentrations in the untreated carryover ex periment were estimated by collecting soil samples from all plots at depth intervals of 0 to 10, 10 to 20, 20 to 50, and 50 to 100 cm in November 2012 using a 7.6-cm-diameter auger. Eight samples were collected from each treatment plot: four from the bed and four from the interbed positions. e samples from the same depth intervals were thoroughly mixed and weighed. Approximately 15% of the mixed sample was then subsampled. Roots were re moved from the subsamples. Almost 100 g of subsample was then air dried and ground in a mortar and pestle to pass thorough a 2-mm sieve. Soil macroand micronutrients were extracted using the Mehlich III procedure (Mehlich, 1984). All samples were analyzed at the Micro-Macro International Laboratory. Analysis of variance (ANOVA) for a randomized complete block design was used to test the eects of fertilizer and weed control on the aboveground biomass, nutrient accumulation, and soil nutrient supply rates for both the actively managed retreated and the untreated carryover experiments. To ensure that the data met assumptions of normality and homoscedasticity, Kolmogorov–Smirnov and equal variance tests, respectively, were used (Massey, 1951). For data not meeting the assumptions of normality and homoscedasticity, appropriate transformations were made before conducting ANOVA in SAS (SAS Institute, 2007). Tukey’s Studentized range (honestly signicant dier ence [HSD]) test was used to separate dierences among treat ment means at an B level of 0.05 unless noted otherwise.RESULTSInter-rotational Comparison of Height Response of Three-Year-Old Loblolly PinesComparison of loblolly pine tree heights between rotations indicated that at age 3 yr, the second rotation stands performed better than the rst (Fig. 2). All actively managed retreated plots in the second rotation had signicantly greater tree heights than those in the rst rotation. For example, with the FW treatment, the average tree height in the second rotation averaged 5.7 m compared with 3.9 m during the rst rotation (Colbert et al., 1990). Likewise for the F, W, and C treatments, average heights in the second vs. the rst rotation averaged 5.0 vs. 3.1, 4.5 vs. 3.0, and 3.6 vs. 1.3 m, respectively. In the untreated carryover experiment, all treatments in the second rotation also had greater average tree heights than the rst rotation. However, the trend was dierent than that observed for the actively managed retreated experiment; the CF treatment had signicantly greater average tree height (5.2 m) than the CFW (4.3 m), CC (3.9 m), and CW (3.8 m) treatments (Fig. 2).Total Aboveground Biomass Accumulation— Loblolly Pinee eects of fertilization and weed control on the total aboveground biomass accumulation of the second-rotation lob lolly pine stands were quantied annually for the rst 3 yr for both the actively managed retreated and untreated carryover experiments (Fig. 3). In general, continuation of the fertilization Table 2. Parameters for an allometric equation for estimating foliage, stem wood with bark, branches, and aboveground tree biomass in young loblolly pine stands growing on Spodosols of the southeastern United States. Biomass Age ln( y ) = C o + C 1 ln( x )† R 2 SE n C 0 C 1 Value P Value P yr Foliage 1 and 2 1.56 <0.01 1.94 <0.01 0.84 0.343 48 Stem wood with bark 1 and 2 2.47 <0.01 2.27 <0.01 0.95 0.203 48 Branch 1 and 2 2.73 <0.01 2.54 <0.01 0.86 0.421 48 Aboveground‡ 1 and 2 1.03 <0.01 2.18 <0.01 0.90 0.302 48 Aboveground‡ 3 1.2 <0.01 1.81 <0.01 0.91 0.153 56 † þ y is the biomass component (kg dry wt.) and x is the tree height (m) for age 1 and 2 yr or diameter at breast height (cm) for age 3 yr. ‡ From Adegbidi et al. (2002). Fig. 2. Heights of the rst-rotation 3-yr-old loblolly pines at the original Intensive Management Practices Assessment Center study sites compared with those in the second-rotation actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and untreated carryover experiment (CC , C F , C W , and C FW , respectively) on Spodosols in North Florida. Error bars represent standard errors of the mean.

PAGE 6

þt Soil Science Society of America Journaland weed control treatments in the second rotation’s actively man aged experiment signicantly increased the total aboveground biomass accumulation. e FW treatment accumulated almost 28 Mg ha1 compared with 8 Mg ha1 for the control (C; 3.5fold response). Dierent patterns were observed for the secondrotation untreated carryover experiment; growth responses for the CF treatment (18 Mg ha1 ) were almost 1.8-fold greater than the untreated control (10 Mg ha1 in CC ). Interestingly, despite the previous history of repeated fertilizer and weed control applications, the carryover CFW treatment (12 Mg ha1 ) did not dier signicantly in aboveground loblolly pine biomass accumulation from the CC and C W treatments from ages 1 to 3 yr (Fig. 3B).Total Aboveground Biomass Accumulation— Understory VegetationAs expected, fertilizer additions in the second rotation sig nicantly increased the amount of understory biomass accumula tion compared with the weed control treatments (Fig. 4A). At age 2 yr, the F treatment accumulated almost 7.7 Mg ha1 of aboveþ­ ground understory biomass compared with 1.6 and 1.1 Mg ha1 for the W and FW treatments, respectively. However, understory vegetation redevelop ment was evident in the second-rotation untreated carryover experiment, especially in the CFW treat ment (Fig. 4B). For example, understory biomass accumulation in the CFW treatment (3.4 Mg ha1 ) did not dier signicantly from either the CC (4.8 Mg ha1 ) or CF (4.8 Mg ha1 ) treatments. On the contrary, the CW treatment (1.7 Mg ha1 ) had the lowest amount of understory biomass accumula tion compared with the CC treatment. Understory vegetation composition was in uenced by the silvicultural treatment histories. In general, grass-like species dominated the understory in the weed control treatments, and shrubby spe cies dominated the understory in the absence of the weed control treatments (Table 3). Andropogon sp. (0.7 Mg ha1 ) and Dicanthelium sp. (0.3 Mg ha1 ) together contributed about 96% of the total understory aboveg round biomass in the FW treatment. Ilex glabra (3.8 Mg ha1 ) and Serenoa repens (W. Bartram) Small (0.91 Mg ha1 ), in con trast, contributed almost 61% of the total understory aboveg round biomass in the F treatment. Similar responses in under story composition were observed in the untreated carryover experiment; Andropogon sp. (3 Mg ha1 ) alone contributed to almost 89% of the total understory aboveground biomass in the C FW treatment, and Ilex glabra (2.0 Mg ha1 ) and Serenoa repens (1.1 Mg ha1 ) together accounted for nearly 65% of the total un derstory aboveground biomass accumulation in the CF treatment.Total Aboveground Biomass Accumulatione total aboveground biomass (pine + understory) accu mulation in the actively managed experiment followed the trend F > FW > C > W (Fig. 4A). Understory vegetation alone accounted for almost 62, 53, 21, and 10% of the total aboveground biomass in the C, F, W, and FW treatments, respectively. Pine bio mass accounted for about 90% of the total aboveground biomass in the FW and 47% in the F treatment. Similarly, in the untreated carryover experiment, the total aboveground biomass accumulation followed the general trend CF > C C > C FW > C W (Fig. 4B). Pine biomass contributed to almost 64, 49, 58, and 69% of the total aboveground biomass in the CF , C C , C FW , and C W treatments, re spectively. e understory vegetation accounted for al most 51% of the total aboveground biomass in the CC treatment and 31% in the CW treatment.Nutrient Accumulation Nutrient Accumulation in Loblolly PineEstimates of aboveground nutrient accumula tions in 2-yr-old loblolly pine are shown in Fig. 5 and 6 for both macroand micronutrients. In general, nu trient accumulation followed the trends reported for Fig. 3. Total aboveground loblolly pine biomass accumulation for second-rotation stands growing in (A) the actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and (B) the untreated carryover experiment (CC , C F , C W , and C FW , respectively) on Spodosols in North Florida. Error bars represent standard deviations. Within a given stand age, treatments followed by same letter are not signicantly different (Tukey’s honestly signicant difference) at B = 0.05. Fig. 4. Total aboveground biomass accumulation among treatments in a second-rotation 2-yr-old loblolly pine stand growing in the (A) actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and (B) untreated carryover experiment (CC , C F , C W , and C FW , respectively) on Spodosols in North Florida. Among treatments, components followed by the same letter were not signicantly different at B = 0.05. Error bars represent standard deviations.

PAGE 7

www.soils.org/publications/sssajþt aboveground biomass accumulation. Nutrient accumulation increased with increasing intensity of silvicultural treatments for the actively managed retreated experiment. Nitrogen, P, and K accumulations in the FW (N, 124.1 kg ha1 ; P, 8.5 kg ha 1 ; K, 38.6 kg ha1 ) treatment were almost 3.4-, 3.9and 3.4-fold higher, respectively, than the C treatment (N, 37.0 kg ha 1 ; P, 2.2 kg ha1 ; K, 11.2 kg ha1 ). Similarly, B, Mn, and Zn accumulations in the actively managed FW treatment plots were almost 2.8-, 4.1-, and 3.2-fold higher than the C treat ment (B, 62 vs. 22 g ha1 ; Mn, 750 vs. 183 g ha1 ; Zn, 322 vs. 100 g ha1 ). Nutrient accumulation in loblolly pine also followed the same trend as biomass accumulation in the untreated carryover experiment (Fig. 5 and 6). The CF treatment had significantly higher nutrient accumulation than the CC treat ment. For instance, almost 1.9-, 1.8-, and 2.1-fold higher N, P, and K contents, respectively, were associated with the CF treatment when compared with the CC treatment (N, 92 vs. 47.6 kg ha1 ; P, 5.4 vs. 3.0 kg ha1 ; K, 34.2 vs. 16.5 kg ha1 ). Similarly, B and Zn accumulations were also higher in the C F than the CC treatment (B, 51 vs. 29 g ha1 ; Zn, 231 vs. 117 g ha1 ). Table 3. Percentage contribution by species to the total understory aboveground biomass in the actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and untreated carryover experiment (C C , C F , C W , and C FW , respectively) in North Florida. (Treatments may not sum to 100 because of rounding error.) Species Actively managed retreated Untreated carryover C F FW W C C C F C FW C W Andropogon sp. 3.1 17.2 65.4 47.5 4.8 15.7 89.4 68.4 Carex spp. 1.0 <0.1 1.8 1.9 1.2 0.1 0.1 3.4 Chrysopsisgraminifolia(Michx.) Elliott 0.3 Cyperus spp. 1.4 2.6 2.6 0.1 0.1 Dicanthelium sp. 1.3 0.6 30.6 34.9 2.3 0.1 1.5 18.2 Eleocharis baldwinii (Torr.) Chapm. <0.1 <0.1 0.4 <0.1 0.2 Erechtites hieraciifolius (L.) Raf. ex DC. <0.1 0.1 Eupatorium capillifolium (Lam.) Small 0.5 Eupatorium spp. 0.2 0.2 2.2 1.8 0.4 <0.1 Gamochaeta purpurea (L.) Cabrera 0.7 Gaylussacia dumosa (Andrews) Torr. & A. Gray 0.2 0.7 0.1 Gaylussacia frondosa (L.) Torr. & A. Gray ex Torr. <0.1 Hypericum spp. 0.4 0.3 0.1 4.0 0.7 1.8 <0.1 Ilex glabra (L.) A. Gray 44.06 49.2 49.2 42.1 Juncus spp. <0.1 0.2 Lachnanthes caroliniana (Lam.) Dandy <0.11 <0.1 3.9 0.5 0.1 Ludwigia decurrens Walter <0.1 Lyonia ferruginea (Walter) Nutt. <0.1 <0.1 0.2 <0.1 Lyonia lucida (Lam.) K. Koch 15.82 5.91 5.3 Panicum hemitomon Schult. 1.4 Myrica cerifera (L.) Small 6.8 Paspalum sp. <0.1 <0.1 0.2 Persea borbonia (L.) Spreng. <0.1 <0.1 0.5 Persea palustris (Raf.) Sarg. 0.5 Pteridium aquilinum (L.) Kuhn <0.1 0.1 <0.1 0.2 Quercus nigra L. 0.3 0.6 1.1 0.2 2.2 0.1 Rhexia alifanus Walter 0.1 Rhexia sp. 0.1 Rhexia virginica L. <0.1 Rhus copallinum L. 8.7 6.7 0.6 2.3 Rubus sp. 0.2 0.1 2.4 Scleria spp. 1.7 <0.1 6.8 0.5 0.1 <0.1 2.4 Serenoa repens (W. Bartram) Small 19.3 11.9 24.1 23.1 1.6 Sorghastrum secundum (Elliott) Nash <0.1 <0.1 Smilax rotundifolia L. 2.0 1.7 1.6 0.1 1.1 0.6 0.7 0.4 Solidago stulosa Mill. 0.1 Sporobolus curtissii (Vasey ex Beal) Small ex Scribn. <0.1 0.8 Vaccinium myrsinites Lam. 0.3 1.1 0.3 3.2 Vitis rotundifolia Michx. 1.9 <0.1 8.4 <0.1 Woodwardia virginica (L.) Sm. 0.6

PAGE 8

þt Soil Science Society of America Journalþ Nutrient Accumulation in the Aboveground Understory VegetationNutrient accumulation in the understory vegetation of the actively managed experiment was aected by the intensive silvicultural treatments like fertilization and weed control (Fig. 5 and 6). As expected, sustained control of competing understory vege tation reduced understory nutrient accumulation in the FW (N, 9.7 kg ha1 ; P, 1.4 kg ha1 ; B, 5 g ha1 ) and W (N, 15.6 kg ha1 ; P, 1.9 kg ha1 ; B, 9 g ha1 ) treatments compared with the F treat ment (N, 68 kg ha1 ; P, 7.4 kg ha1 ; B, 121 g ha1 ). No signicant dierences in understory nutrient accumulation were found between the F and C treatments (Fig. 5 and 6). In the untreated carryover experiment, the CF and C W treatments did not signicantly inuence the N and P accu mulation in the second-rotation understory vegetation (Fig. 5). However, the CFW treatment had signicantly lower Ca, Cu, and Mn accumulation in the understory compared with the CF treatment (Ca, 5.8 vs. 26.3 kg ha1 ; Cu, 6 vs. 21 g ha1 ; Mn, 370 vs. 1103 g ha1 ). Moreover, when compared with the CC treatment, the CW treatment had signicantly lower K, Ca, Cu, and Zn accumulation in the understory, diering by almost 71% (22.3 kg ha1 in CC vs. 6.4 kg ha1 in CW ), 90% (19.9 kg ha1 in C C vs. 2.1 kg ha1 in CW ), 76% (21 g ha1 in CC vs. 5 g ha1 in C W ), and 73% (161 g ha1 in CC vs. 44 g ha1 in CW ), respectively. is dierence was primarily due to lower understory bio mass in the CW treatment (? 63% less) than the CC treatment. Woody understory species like Ilex glabra and Serenoa re pens were the major accumulators of nutrients like N, P, B, Cu, Mn, and Zn in the C and F treatments of the actively managed experiment and the CC and C F treatments of the untreated carryover experiment (Subedi, 2013; Fig. 7). For example, Ilex glabra alone accumulated almost 23 and 20 kg ha1 of N and 3.3 and 1.7 kg ha1 of P in the F and C treatments, respectively. Similarly, it accumulated almost 43% (? 18.2 kg ha1 ) and 32% ( ? 12.7 kg ha1 ) of the total understory N in the CF and C C treatments, respectively. Boron, Mn, and Zn accumu lation in Ilex glabra accounted for almost 64% (? 77 g ha1 ), 82% (? 959 g ha1 ), and 52% (? 884 g ha1 ) , respectively, of the total understory micronutrient pools in the F treatment. Almost 17% (? 225 g ha1 ) of the total understory Mn pool in the C treatment was contributed by Serenoa repens. Ilex glabra (717 g ha1 ) and Vitis rotundifolia Michx. (147 g ha1 ) accu mulated nearly 65 and 13%, respectively, of the total understo ry Mn pool in the CF treatment. Herbaceous species like Andropogon sp. and Dicanthelium sp. were the major nutrient accumulators in the FW, W, CFW , Fig. 5. Macronutrient accumulation in the aboveground biomass of 2-yr-old loblolly pine and understory vegetation for the actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and untreated carryover experiment (C C , C F , C W , and C FW , respectively) on Spodosols in North Florida. Among treatments, components followed by the same letter were not signicantly different at B = 0.05. Error bars represent standard deviations.

PAGE 9

www.soils.org/publications/sssajþt and C W treatments (Subedi, 2013; Fig. 7). Andropogon sp. accu mulated approximately 93% (8.6 kg ha1 ) and 80% (13 kg ha1 ) of the total understory N pool and 93% (1.3 kg ha1 ) and 89% (2.6 kg ha1 ) of the total understory P pool in the FW and CFW treatments, respectively. Manganese and Zn accumulation in Andropogon sp. was almost 262 g ha1 ( ? 71% of the total un derstory Mn pool) and 48 g ha1 ( ? 60% of the total understory Zn pool) in the CFW treatment. Andropogon sp. was a major N accumulator in the W treated plots (9.9 kg ha1 , 63% of the total understory N pool).Soil Nutrient SupplySoil nutrient supply rates in the second-rotation loblolly pine stands, measured using ion exchange membranes, were sig nicantly higher for nutrients like N, Ca, Mn, Cu, Zn, and S in the actively managed FW treatment than the C treatment (B = 0.1; Table 4) at age 2 yr. Tukey’s HSD (B = 0.1) revealed that the P supply rate was signicantly higher in the F than the W treatment (36.2 vs. 4.1 mg m2 during an 8-wk period) (Fig. 8A). Strong correlations between aboveground pine biomass and soil nutrient supply rates during the growing season were observed for several nutrients (Table 5). For instance, correla tions between aboveground pine biomass and soil supply rates for N and Zn were 0.67 and 0.73, respectively, for the actively managed experiment. In the untreated carryover experiment, the CF treatment had a signicantly higher supply of nutrients like P, Mn, and Zn in the upper soil surface when compared with the CC treatment ( B = 0.1; Fig. 8B). e P supply rate in the CF treatment was almost 2.3-fold higher than the CC treatment (CF , 20.8 mg m2 ; C C , 8.9 mg m2 during an 8-wk period). Interestingly, the CFW treatment had no signicant inuence on the surface supply of soil nutrients when compared with the CC treatment, except for Zn. In the untreated carryover experiment, strong correlations were observed between the aboveground loblolly pine biomass and soil P (r = 0.83), Mn (r = 0.79), and Cu (r = 0.73) supply rates (Table 5). Of these nutrients, only the soil P supply could signicantly predict the total aboveground biomass accumula tion of loblolly pine (forward stepwise regression: R 2 = 0.68) (Fig. 9). Analysis of variance of the Mehlich III extractable soil P concentration in the untreated carryover plots showed sig nicant treatment dierences. With the exception of the CW treatment, both the CFW and C F treatments had signicantly higher mean soil P concentrations than the CC treatment. e least square mean soil P concentration was highest for the CFW treatment (19.5 mg kg1 ) followed by the CF (12.6 mg kg1 ), C W (8.5 mg kg1 ), and C C (6.1 mg kg1 ) treatments for the 0to 100-cm depth (Tukey’s HSD at B = 0.1). In addition, soil P concentrations increased with soil depth (Fig. 10). For instance, the mean soil P concentration in the 50to 100-cm soil depth (19.7 mg kg1 ) was almost 3.1and 2.2-fold higher than the 0to 20-cm (6.3 mg kg1 ) and 20to 50-cm (9.01 mg kg1 ) soil depths, respectively.DISCUSSION Understanding the long-term impacts of intensive forest management on the productivity of planted pines over successive rotations will inform discussions of forest sustainability (Powers, 1999; Fox, 2000; Wear and Greis, 2002). Productivity and growth of pines are aected by multiple factors such as soil nutrient availability, genetics, and competition (Allen et al., 1990; Neary et al., 1990b; Li et al., 1999; Jokela et al., 2010). e long-term replicated experiments in this study incorporated Fig. 6. Micronutrient accumulation in the aboveground biomass of 2-yr-old loblolly pine and understory vegetation for the actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and untreated carryover experiment (C C , C F , C W , and C FW , respectively) on Spodosols in North Florida. Among treatments, components followed by the same letter were not signicantly different at B = 0.05. Error bars represent standard deviations.

PAGE 10

þt Soil Science Society of America Journalrotation-long applications of fertilizer and sustained understory competition control. is experimental framework provided a unique opportunity to investigate inter-rotational silvicultural impacts on growth dynamics and the competitive environment in loblolly pine stands. e evolution of forest management practices and the avail ability of improved genetic stock have increased the productivity of southern pine plantations during the past six decades (Fox et al., 2007a). In this study, inter-rotational comparisons of loblolly pine growth demonstrated that height and aboveground biomass accumulation were consistently greater in the second rotation than the rst. is response, in part, was probably due to the de ployment of genetically superior pine seedlings (a single full-sib pine family in the second rotation vs. multiple rst-generation open-pollinated pine families in the rst rotation) (Li et al., 1999; Jansson and Li, 2004), improved site preparation (Miller et al., 1991; Borders and Bailey, 2001; Jones et al., 2009) and bedding techniques (double bedding vs. single bedding) (Lauer and Zutter, 2001), control of Panicum spp. (Morris et al., 1993) and tip moth (Williston and Barras, 1977; Cade and Hedden, Fig. 7. Aboveground biomass and nutrient accumulation in (A) herbaceous plants and grasses and (B) shrubs and vines of the actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and (C) herbaceous plants and grasses and (D) shrubs and vines of the untreated carryover experiments (CC , C F , C W , and C FW , respectively) in North Florida. Means within biomass or nutrient treatments followed by the same letter were not signicantly different at B = 0.1. Error bars represent standard errors of the mean.Table 4. Soil nutrient supply rates for 8 wk beginning in August 2011 (0–15 cm) for the actively managed retreatment experiment in North Florida. The experiment received control (C), fertilizer only (F), fertilizer + weed control (FW), and weed control only (W) treatments in the current and previous rotations. Treatment Soil nutrient supply rate N P K Ca Mg S B Cu Mn Zn Fe ——————————————————————————— mg m 2 ——————————————————————————— C 8.4 b (10.7)† 10.2 ab (6.2) 65.3 a (22.6) 269.2 b (126.8) 162.1 a (81.9) 26.7 b (18.9) 0.2 a (<0.1) 0.1 b (<0.1) 3.1 b (2.4) 1.0 b (0.4) 2.3 a (0.8) F 62.1 ab (41.6) 36.2 a (3.3) 103.9 a (56.0) 531.4 ab (102.2) 264.9 a (28.7) 31.1 ab (13.9) 0.2 a (0.1) 0.9 a (0.3) 9.1 ab (3.9) 4.4 a (1.3) 5.3 a (3.6) FW 222.2 a (167.6) 29.0 a (21.9) 78.3 a (22.8) 740.1 a (286.1) 299.8 a (121.6) 88.3 a (53.7) 0.2 a (0.1) 0.7 a (0.3) 9.4 a (3.0) 5.4 a (1.7) 5.8 a (2.4) W 9.6 b (3.6) 4.1 b (1.4) 74.3 a (38.6) 270.4 b (84.1) 173.9 a (38.3) 25.0 b (8.5) 0.1 a (<0.1) 0.1 b (<0.1) 3.2 b (1.7) 1.1 b (0.5) 3.2 a (1.1) † Standard deviations in parentheses. Within each nutrient, treatments followed by the same letter are not signicantly different at B = 0.1.

PAGE 11

www.soils.org/publications/sssajþt 1987), and an elevation in atmospheric CO2 concentrations ( ? 344 N L L 1 in 1983 vs. 393 N L L 1 in 2011) (Nemani et al., 2003; Moore et al., 2006) in the second rotation. Fertilizer additions and understory competition control have been shown to benet the growth of loblolly pine on nu trient-stressed sites by reducing nutrient deciencies (Swindel et al., 1988; Colbert et al., 1990; Borders et al., 2004; Martin and Jokela, 2004; Roth et al., 2007; Jokela et al., 2010). Results from this study clearly demonstrated that these benets were ex tended into the second rotation. ird-year estimates of aboveg round biomass accumulation (FW treatment, 28 Mg ha1 ), for example, were similar to those reported by Colbert et al. (1990; FW treatment, 32 Mg ha1 ) at age 4 yr for this same site in the previous rotation. Although this exemplies the yield for the highest levels of silvicultural inputs (FW) for the actively managed experiment, results were similar for the untreated car ryover experiment. e CF treatment outperformed all other treatments in the untreated carryover experiment. ird-year estimates of aboveground biomass accumulation in the CF treat ment (17.8 Mg ha1 ) were similar to the value of 19.7 Mg ha1 reported by Adegbidi et al. (2005) for intensively managed 3-yrold loblolly pine stands growing on similar soils. During the early stages of canopy development, demand for soil nutrients is critical to the development of photosynthetic tissues, and both the fertilization and weed control treatments can contribute to increased soil nutrient supply (Miller, 1981; Neary et al., 1990a). Higher N and P contents in the aboveground bio mass of pines in the FW treatment were indicative of the general nutrient uptake rates of these rapidly growing stands. Total N (124.1 kg ha1 ) and P (8.5 kg ha1 ) pools in the aboveground pine biomass at age 2 yr in the FW treatment of our study closely matched those reported by Adegbidi et al. (2005) for this same family at age 3 yr growing on similar sites in the Lower Coastal Plain of Georgia (117 kg ha1 N and 8.3 kg ha1 P, respectively). Despite greater nutrient immobilization in the aboveground pine biomass for the F and FW treatments, the high foliar N concentration (22.6 g.kg1 ) observed in the FW treatment (Subedi, 2013) and higher soil nutrient supply rates observed in the upper soil surface (15cm) of the F and FW treatments (Fig. 8) highlight the benets that long-term nutrient additions and competing vegetation control have on soil N and P availability in sandy Spodosols. e potential of fertilizer amendments, especially P, to persist and enhance long-term soil nutrient availability and aect inter-rotational levels of productivity has been previously documented for pine plantations across dierent regions (Comerford et al., 2002; Crous et al., 2007; Everett and PalmLeis, 2009; Kiser and Fox, 2012). e notable dierences in residual soil nutrient availability and second-rotation growth rates found between the CF and C FW treatments was unex pected, given that comparable levels of fertilizer additions were Fig. 8. Phosphorus, Mn, and Zn supply in the (A) actively managed retreated experiment (control [C], fertilization [F], weed control [W], and fertilization + weed control [FW]) and (B) untreated carryover experiment (CC , C F , C W , and C FW , respectively) on Spodosols in North Florida. Within a nutrient, treatments with the same letter indicate no signicant differences at B = 0.1. Error bars represent standard deviations.Table 5. Correlation coefcients (r ) between soil nutrient supply rates during the growing season and aboveground biomass accumulation in loblolly pine stands for the actively managed retreated and untreated carryover experiments on Spodosols in North Florida. Within an experiment, data were combined across treatments before analysis (n = 12). Nutrient Actively managed retreated Untreated carryover r P r P N 0.67 0.017 0.30 0.344 P 0.42 0.170 0.83 0.001 K 0.07 0.827 0.37 0.241 Ca 0.73 0.007 0.40 0.198 Mg 0.61 0.037 0.31 0.335 S 0.61 0.034 0.15 0.633 B 0.07 0.840 0.13 0.686 Cu 0.51 0.094 0.73 0.007 Mn 0.62 0.031 0.79 0.002 Zn 0.73 0.007 0.55 0.066 Fe 0.57 0.051 0.10 0.748

PAGE 12

þt Soil Science Society of America Journalmade for these treatments in the previous rotation. A number of factors related to dierences in soil nutrient pools, nutrient mineralization rates, movement and availability of soil nutrients in the upper solum, and immobilization of soil nutrients in the understory vegetation may have been responsible. e higher levels of biomass accumulation associated with the CF treatment compared with the CFW and C C treatments was probably due to higher background soil nutrient pools as a result of past fertilizer additions, forest oor incorporation, and understory mulching before stand establishment (Fig. 3 and 8). Based on the N pools measured at the end of the previous rota tion, almost 1.4 Mg ha1 of N present in the forest oor and the understory vegetation was estimated to have been incorporated in the CF treatment compared with 0.95 Mg N ha1 in the CFW and 0.9 Mg N ha1 in the CC treatment (Vogel et al., 2011). Similar levels of soil nutrient enrichment probably occurred for P. Using forest oor mass, understory biomass, and P data for this same site (Vogel et al., 2011; Neary et al., 1990b; Polglase et al., 1992b), estimated P pools in the forest oor and understory vegetation were also higher in the CF treatment (32.5 kg ha1 ) than the CFW (26.3 kg ha1 ) and C C (8.1 kg ha1 ) treatments. e nutrient pools in the forest oor and understory from the rst rotation probably served as a nutrient source on decom position and subsequent mineralization (Tisdale, 2008; Maier et al., 2012) and thereby supported the greater second-rotation lob lolly pine growth in the CF treatment. With forest oor incorpo ration during site preparation, Maier et al. (2012) observed 18% higher stand volume compared with the control in a 6-yr-old loblolly pine plantation in South Carolina. Lower soil P (? 61% lower) and Mn (? 62% lower) supply rates for the CFW treat ment in our study supports the nutrient sink–source relationship between successive rotations, especially in highlighting the role that the understory and forest oor play in this process. Despite higher soil N pools at the end of the rst rotation, nonsignicant treatment dierences in resin-available soil N supply 2 yr aer forest oor incorporation and understory mulching suggests a depletion of soil N via leaching or immobilization on these sandy soils (Miller, 1981; Kissel et al., 2009). Nevertheless, the strong correlation between pine growth and soil P supply rates in the upper solum suggests that higher background soil nutrient pools and increased P availability probably improved the growth of loblolly pines in the CF treatment compared with the CFW treatment at this site (Allen et al., 1990; Albaugh et al., 2007; Fox et al., 2007b). e quality of the substrate, along with temperature and soil moisture, inuences the decomposition of the forest oor and understory mulch and subsequent nutrient mineralization rates (Polglase et al., 1992a, 1992b; Gonçalves and Carlyle, 1994; Scott and Binkley, 1997; Grierson et al., 1999; Piatek and Allen, 1999). Because P in understory vegetation is presumably less re calcitrant than that in the litterfall from the pines (Polglase et al., 1992a), the higher soil P supply observed in the CF treatment was probably due to higher P mineralization rates compared with the CFW and C W treatments. Polglase et al. (1992a) and Grierson et al. (1999) observed higher P mineralization rates in fertilized plots that contained understory vegetation for this same site. In addition, Polglase et al. (1992b) reported higher phenolic concentrations in the pine litter associated with the weed control treatment. Because higher phenolic concentrations in litter could have the potential to hinder litter decomposition, either by forming a decomposition-resistant complex with proteins (Hagerman et al., 1998) or non-proteins (Benoit and Starkey, 1968) or inhibiting microbial activity (Harrison, 1971; Schimel et al., 1996), the potential of P mineralization from the rst-rotation forest oor in the CFW and C W plots was presum ably lower. On nutrient-poor sandy soils, nutrient loss either through leaching or through volatilization (N) in the absence of under story vegetation is of concern in young pine stands (Outcalt and White, 1981; Smethurst and Nambiar, 1995; Piatek and Allen, 2001; Tessier and Raynal, 2003; Kissel et al., 2009; Zerpa and Fox, 2011). Phosphorus leaching from the E to the Al-dominant Bh and Bt horizons was observed in the untreated carryover ex Fig. 9. Relationship between soil P supply and the aboveground pine biomass in the untreated carryover experiment on Spodosols in North Florida. Plots were treated in the previous rotation as control (CC ), fertilization (C F ), weed control (CW ), and fertilizer + weed control (C FW ). Error bars represent standard deviations. Fig. 10. Mehlich III extractable P concentrations in the soils of the untreated carryover experiment in North Florida. Plots were treated in the previous rotation as control (CC ), fertilization (C F ), weed control (CW ), and fertilizer + weed control (CFW ). Error bars represent standard errors of the mean.

PAGE 13

www.soils.org/publications/sssajþt periment (Fig. 10). In the CFW treatment, Mehlich III extract able P concentrations were almost twofold higher in the 20to 50-cm (19.7 mg kg1 ) and 50to 100-cm (27.5 mg kg1 ) depths compared with the 0to 20-cm depth (11.4 mg kg1 ). Because root development of 2-yr-old loblolly pines are restricted main ly to the upper 25 cm (Adegbidi et al., 2004), P present in the 20to 50and 50to 100-cm depths may have been inaccessible to the pines at this stage of stand development. As a result, aboveground pine growth at age 2 yr was not dierent between the CFW and C C treatments, despite higher average Mehlich III extractable P concentrations (0–100 cm) in the CFW treatment. By the end of the third growing season, however, the growth re sponse of pines in the CFW treatment, although not signicant, was beginning to separate from the CC treatment (Fig. 3). is response would probably become more pronounced as stand development proceeds and nutrients present in the deeper soil horizons become accessible to the pines on further root develop ment (Adegbidi et al., 2004). Competition for soil nutrients from understory vegeta tion is a major growth limitation in southern pine plantations (Neary et al., 1990a, 1990b; Colbert et al., 1990). With nutrient additions, understory vegetation growth can be dramatically in creased (Turner and Long, 1975; Persson, 1981; VanderSchaaf et al., 2002). us, without fertilization nutrient immobilization in the understory vegetation places large demands on soil nutrient availability for pines in aggrading stands. For example, there was almost 2.8and 3.8-fold higher N and P accumulation, respectively, in the aboveground understory vegetation in the F plots than that reported by Gholz et al. (1985) for unfertilized slash pine plantations on similar soils (N, 24.7 kg ha1 ; P, 1.9 kg ha1 ). In addition to changes in understory biomass, silvicultural treatments mediated shis in the understory community composition (Neary et al., 1990a; Miller et al., 1991). For this same site, Neary et al. (1990a) reported in the rst rotation a shi to a shrub-dominated understory community following fertilizer additions 6 yr aer site preparation and planting; the shrub biomass increased by 137% and herbaceous biomass decreased by 76% relative to untreated controls. e herbaceous community dominance documented in the weed control treatments (both in the actively managed [FW and W] and carryover experiments [C FW and C W ]) was also consistent with the ndings of Jones et al. (2009), who reported a dominance of early seral species on sites established with chemical site preparation. Along with changes in community composition, dieren tial nutrient uptake rates of these understory communities inu enced nutrient availability. e shrub community, dominated by Ilex glabra and Serenoa repens, had higher nutrient uptake poten tial as demonstrated by higher aboveground N and P content for the F treatment than the C treatment, which had similar amounts of understory biomass accumulation (Fig. 7B). Nitrogen and P concentrations (content/biomass ratio) in the shrubs of the F treatment (N, 9.3 mg kg1 ; P, 0.9 mg kg1 ) of the actively man aged experiment and the CF treatment (N, 9.2 mg kg1 ; P, 0.7 mg kg1 ) of the untreated carryover experiment were almost twice the values reported by Neary et al. (1990a) for fertilized plots (N, 4.8 mg kg1 ; P, 0.3 mg kg1 ) at this site. In that context, elimination of shrubby competition from the FW treatment in creased the pine biomass by fourfold (p = 0.06) compared with twofold increase in the F treatment (p = 0.07). Dominance of the herbaceous understory species like Andropogon in the CFW plots also represented a signicant source of competition for loblolly pine growth. Although the root/shoot ratio of N and P accumulation in Andropogon sp. amounted to 1:3 and 1:10, respectively, in the CFW treatment (Table 6), lower C costs associated with the dense, brous root system of Andropogon sp. (Eissenstat, 1997) facilitated an uptake and immobilization of almost 13.3 kg ha1 of N and 2.6 kg ha1 of P in the aboveground component. Strong competitive inu ences of Andropogon sp. on loblolly pine growth have been pre viously documented (Mitchell et al., 1999; Zutter et al., 1999). For example, Zutter et al. (1999) reported a sixfold decline in the root length density of loblolly pine with as few as four Andropogon sp. plants m2 of soil surface. In addition, allelo pathic eects of Andropogon sp. have been shown to inhibit the nodulation of N2 –xing plants and thereby reduce soil N avail ability (Rice, 1972). us, strong competition from Andropogon sp. for the available soil nutrients in the upper soil surface prob ably inuenced the reduced growth observed for loblolly pine in the CFW treatment.SUMMARY AND CONCLUSIONSis study was designed to examine the eects of previous forest management activities on inter-rotational productivity in young loblolly pine stands. Continued application of fertilizer and sustained elimination of competing vegetation favored early growth of pines in the second rotation. Growth, as measured using aboveground biomass and total average height, was greater in the second rotation for both the actively managed retreat ment and untreated carryover experiments. Notably, the carry over CF treatment was signicantly more productive than the Table 6. Macroand micronutrient accumulation in the belowground biomass of Andropogon sp. for the treatments that received weed control in the rst rotation (weed control only [CW ] and fertilizer + weed control [CFW ]) in the untreated carryover experiment in North Florida. Treatment Root biomass b/a ratio† Macronutrients Micronutrients N P K Ca Mg S B Cu Mn Zn kg ha 1 ————— kg ha 1 ————— ——— g ha 1 ——— C FW 864 0.28 (0.09) 4.4 0.3 1.5 0.3 0.2 0.4 2.9 1.8 8.7 14.4 C W 285 0.24 (0.03) 1.7 0.1 0.4 0.1 0.1 0.1 0.8 0.4 2.1 4.8 † Belowground/aboveground biomass ratio with standard deviation in parentheses.

PAGE 14

þt Soil Science Society of America JournalC FW , C W , and C C treatments. Long-term fertilization and weed control treatments also contributed to shis in the understory community composition to more competitive shrubs (F and CF treatments) or Andropogon sp. dominated communities (CFW and C W treatments) that aected nutrient immobilization and loblolly pine growth. While greater growth responses in the second rotation in all treatments might be due to advanced genet ics, improved site preparation techniques, Panicum spp. control, tip moth control, and elevated atmospheric CO2 concentrations, these results also suggest that the forest oor and understory veg etation from the previous rotation was an important nutrient sink, especially for P, in the untreated carryover plots. Following harvest and regeneration, this nutrient pool presumably became a nutrient source that helped meet the nutritional demands of the second-rotation stand. Historical P movement from the E to the Bh and Bt horizons, in the absence of understory vegetation, especially for the CFW treatment, could create an early limitation to growth on P-limited sites. Strong correlation between pine growth and resin-available P supply (r = 0.83) in the surface soils of the untreated carryover experiment, for instance, highlights the important role of P supply in early pine growth on Spodosols. erefore, from a forest management perspective on atwoods sites that were previously fertilized with P, the newly regener ated stands could benet from nutrient management practices like understory mulching and forest oor incorporation, which could reduce the need for P fertilization at establishment when an intact understory was present in the previous stand.A þcknowledgments CKNOWLEDGMENTSis research was funded by the member companies of the Forest Biology Research Cooperative (FBRC) at the University of Florida. Rayonier–Forest Resources is gratefully acknowledged for providing access to and maintenance of the study site.REFERENCES Adegbidi, H.G., N.B. Comerford, E.J. Jokela, and N.F. Barros. 2004. Root development of young loblolly pine in Spodosols in southeast Georgia. Soil Sci. Soc. Am. J. 68:596–604. doi:10.2136/sssaj2004.5960 Adegbidi, H.G., E.J. Jokela , and N.B. Comerford . 2005 . Factors inuencing production eciency of intensively managed loblolly pine plantations in a 1to 4-year-old chronosequence. For. Ecol. Manage. 218 : 245 – 258 . doi:10.1016/j.foreco.2005.08.016 Adegbidi, H.G., E.J. Jokela , N.B. Comerford , and N.F. Barros . 2002 . Biomass development for intensively managed loblolly pine plantations growing on Spodosols in the southeastern USA . For. Ecol. Manage. 167 : 91 – 102 . doi:10.1016/S0378-1127(01)00691-0 Albaugh , T.J. , H.L. Allen , and T.R. Fox . 2007 . Historical patterns of forest fertilization in the southeastern United States from 1969 to 2004. South. J. Appl. For. 31 : 129 – 137 . Albaugh , T. , H.L. Allen , T.R. Fox , C.A. Carlson, and R.A. Rubilar . 2009 . Opportunities for fertilization of loblolly pine in the sandhills of the southeastern United States . South. J. Appl. For. 33 : 129 – 136 . Allen , H.L. 1987 . Forest fertilizers: Nutrient amendment, stand productivity, and environmental impact. J. For. 85 : 37 – 46 . Allen , H.L., P.M. Dougherty , and R.G. Campbell . 1990 . Manipulation of water and nutrients: Practice and opportunity in southern U.S. pine forests . For. Ecol. Manage. 30 : 437 – 453 . doi:10.1016/0378-1127(90)90153-3 Allen , H.L., T.R. Fox , and R.G. Campbell . 2005 . What is ahead for intensive pine plantation silviculture in the South? South. J. Appl. For. 29 : 62 – 69 . Ballard , R. 1978 . Eect of rst rotation phosphorus applications on fertilizer requirements of second rotation radiata pine. N.Z. J. For. Sci. 8 : 135 – 145 . Baskerville, G.L. 1972 . Use of logarithmic regression in the estimation of plant biomass . Can. J. For. Res. 2 : 49 – 53 . doi:10.1139/x72-009 Bengtson , G.W. 1979 . Forest fertilization in the United States: Progress and outlook . J. For. 77 : 222 – 229 . Benoit, R. , and R. Starkey . 1968 . Inhibition of decomposition of cellulose and some other carbohydrates by tannin . Soil Sci. 105 : 291 – 296 . doi:10.1097/00010694-196805000-00002 Blazier , M.A., T.C. Hennessey , and S. Deng . 2005 . Eects of fertilization and vegetation control on microbial biomass carbon and dehydrogenase activity in a juvenile loblolly pine plantation . For. Sci. 51 : 449 – 459 . Borders, B.E., and R.L. Bailey . 2001 . Loblolly pine: Pushing the limits of growth . South. J. Appl. For. 25 : 69 – 74 . Borders, B.E., R.E. Will , D. Markewitz , A. Clark , R. Hendrick , R.O. Teskey , and Y. Zhang . 2004 . Eect of complete competition control and annual fertilization on stem growth and canopy relations for a chronosequence of loblolly pine plantations in the Lower Coastal Plain of Georgia . For. Ecol. Manage. 192 : 21 – 37 . doi:10.1016/j.foreco.2004.01.003 Boring , L.R., C.D. Monk , and W.T. Swank . 1981 . Early regeneration of a clear-cut southern Appalachian forest. Ecology 62 : 1244 – 1253 . doi:10.2307/1937289 Cade, S.C., and R.L. Hedden . 1987 . Growth impact of pine tip moth on loblolly pine plantations in the Ouachita Mountains of Arkansas . South. J. Appl. For. 11 : 128 – 133 . Campbell , C. 1992 . Determination of total nitrogen in plant tissue by combustion. In: C.O. Plank, editor, Plant analysis reference procedures for the southern US. South. Coop. Ser. Bull. 368. Univ. of Georgia Agric. Exp. Stn., Athens. p. 20 – 22 . Colbert , S.R., E.J. Jokela , and D.G. Neary . 1990 . Eects of annual fertilization and sustained weed control on dry matter partitioning, leaf area, and growth eciency of juvenile loblolly and slash pine. For. Sci. 36 : 995 – 1014 . Collet, C. , H. Frochot, J.-M. Guehl , and A. Ferhi . 1996 . Eect of two forest grasses diering in their growth dynamics on the water relations and the growth of Quercus petraea seedlings . Can. J. Bot. 74 : 1562 – 1571 . doi:10.1139/b96-189 Comerford , N.B. , M. McLeod , and M. Skinner . 2002 . Phosphorus form and bioavailability in the pine rotation following fertilization: P fertilization inuences P form and potential bioavailability to pine in the subsequent rotation . For. Ecol. Manage. 169 : 203 – 211 . doi:10.1016/S0378-1127(01)00680-6 Crous , J. , A. Morris , and M. Scholes . 2007 . e signicance of residual phosphorus and potassium fertiliser in countering yield decline in a fourth rotation of Pinus patula in Swaziland . South. Hemisphere For. J. 69 : 1 – 8 . doi:10.2989/SHFJ.2007.69.1.1.163 Dalla-Tea , F. , and E.J. Jokela . 1994 . Needlefall returns and resorption rates of nutrients in young intensively managed slash and loblolly pine stands . For. Sci. 40 : 650 – 662 . Echeverría , M.E., D. Markewitz , L.A. Morris , and R.L. Hendrick . 2004 . Soil organic matter fractions under managed pine plantations of the southeastern USA . Soil Sci. Soc. Am. J. 68 : 950 – 958 . doi:10.2136/sssaj2004.9500 Eissenstat , D.M. 1997 . Trade-os in root form and function. In: L.E. Jackson, editor, Ecology in agriculture. Academic Press, San Diego. p. 173 – 199 . Everett , C.J., and H. Palm-Leis . 2009 . Availability of residual phosphorus fertilizer for loblolly pine. For. Ecol. Manage. 258 : 2207 – 2213 . doi:10.1016/j.foreco.2008.11.029 Folke , C. , S. Carpenter , B. Walker , M. Scheer , T. Elmqvist, L. Gunderson, and C.S. Holling . 2004 . Regime shis, resilience, and biodiversity in ecosystem management. Annu. Rev. Ecol. Evol. Syst. 35 : 557 – 581 . doi:10.1146/annurev.ecolsys.35.021103.105711 Fox , T.R. 2000 . Sustained productivity in intensively managed forest plantations . For. Ecol. Manage. 138 : 187 – 202 . doi:10.1016/S0378-1127(00)00396-0 Fox , T.R. , H.L. Allen , T.J. Albaugh , R. Rubilar , and C.A. Carlson . 2007b . Tree nutrition and forest fertilization of pine plantations in the southern United States . South. J. Appl. For. 31 : 5 – 11 . Fox , T.R. , E.J. Jokela , and H.L. Allen . 2007a . e development of pine plantation silviculture in the southern United States . J. For. 105 : 337 – 347 . Gholz , H.L., and R.F. Fisher . 1982 . Organic matter production and distribution in slash pine (Pinus elliottii) plantations . Ecology 63 : 1827 – 1839 . doi:10.2307/1940124 Gholz , H.L., R.F. Fisher , and W.L. Pritchett. 1985 . Nutrient dynamics in slash pine plantation ecosystems . Ecology 66 : 647 – 659 . doi:10.2307/1940526

PAGE 15

www.soils.org/publications/sssajþt Gonçalves , J.L.M., and J.C. Carlyle. 1994 . Modelling the inuence of moisture and temperature on net nitrogen mineralization in a forested sandy soil . Soil Biol. Biochem. 26 : 1557 – 1564 . doi:10.1016/0038-0717(94)90098-1 Grierson, P.F. , N.B. Comerford , and E.J. Jokela . 1999 . Phosphorus mineralization and microbial biomass in a Florida Spodosol: Eects of water potential, temperature and fertilizer application . Biol. Fertil. Soils 28 : 244 – 252 . doi:10.1007/s003740050489 Hagerman , A.E., K.M. Riedl , G.A. Jones, K.N. Sovik , N.T. Ritchard , P.W. Hartzfeld , and T.L. Riechel . 1998 . High molecular weight plant polyphenolics (tannins) as biological antioxidants . J. Agric. Food Chem. 46 : 1887 – 1892 . doi:10.1021/jf970975b Harrison, A.F. 1971 . e inhibitory eect of oak leaf litter tannins on the growth of fungi, in relation to litter decomposition . Soil Biol. Biochem. 3 : 167 – 172 . doi:10.1016/0038-0717(71)90011-3 Hart , S.C., and M.K. Firestone . 1989 . Evaluation of three in situ soil nitrogen availability assays . Can. J. For. Res. 19 : 185 – 191 . doi:10.1139/x89-026 Jansson, G. , and B. Li . 2004 . Genetic gains of full-sib families from disconnected diallels in loblolly pine. Silvae Genet. 53 : 60 – 64 . Jeries, S.B. , T.R. Wentworth , and H.L. Allen . 2010 . Long-term eects of establishment practices on plant communities across successive rotations in a loblolly pine (Pinus taeda) plantation . For. Ecol. Manage. 260 : 1548 – 1556 . doi:10.1016/j.foreco.2010.08.003 Jokela , E.J. , H.L. Allen , and W.W. McFee . 1991a . Fertilization of southern pines at establishment. In: M.L. Duryea and P.M. Dougherty, editors, Forest regeneration manual. Springer-Verlag, Dordrecht, the Netherlands. p. 263 – 277 . Jokela , E.J. , P.M. Dougherty , and T.A. Martin . 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 : 117 – 130 . doi:10.1016/j.foreco.2004.01.007 Jokela , E.J. , and T.A. Martin . 2000 . Eects of ontogeny and soil nutrient supply on production, allocation, and leaf area eciency in loblolly and slash pine stands . Can. J. For. Res. 30 : 1511 – 1524 . doi:10.1139/x00-082 Jokela , E.J. , T.A. Martin , and J.G. Vogel . 2010 . Twenty-ve years of intensive forest management with southern pines: Important lessons learned . J. For. 108 : 338 – 347 . Jokela , E.J. , and S.C. Stearns-Smith . 1993 . Fertilization of established southern pine stands: Eects of single and split nitrogen treatments . South. J. Appl. For. 17 : 135 – 138 . Jokela , E.J. , E.L. Stone , and W.W. McFee . 1991b . Micronutrient deciency in slash pine: Response and persistence of added manganese. Soil Sci. Soc. Am. J. 55 : 492 – 496 . doi:10.2136/sssaj1991.03615995005500020033x Jones, P.D. , S.L. Edwards, S. Demarais, and A.W. Ezell . 2009 . Vegetation community responses to dierent establishment regimes in loblolly pine ( Pinus taeda) plantations in southern MS, USA . For. Ecol. Manage. 257 : 553 – 560 . doi:10.1016/j.foreco.2008.09.033 Jurgensen , M.F., A.E. Harvey , R.T. Graham, D.S. Page-Dumroese, Jr ., M.J. Tonn , M.J. Larsen, and T.B. Jain . 1997 . Impacts of timber harvesting on soil organic matter, nitrogen, productivity, and health of Inland Northwest forests . For. Sci. 43 : 234 – 251 . Keeves , A. 1966 . Some evidence of loss of productivity with successive rotations of Pinus radiata in the south-east of South Australia . Aust. For. 30 : 51 – 63 . doi:10.1080/00049158.1966.10675398 Kiser , L.C. , and T.R. Fox . 2012 . Soil accumulation of nitrogen and phosphorus following annual fertilization of loblolly pine and sweetgum on sandy sites . Soil Sci. Soc. Am. J. 76 : 2278 – 2288 . doi:10.2136/sssaj2012.0118 Kissel , D.E. , M.L. Cabrera , N. Vaio , J.R. Craig , J.A. Rema , and L.A. Morris . 2009 . Forest oor composition and ammonia loss from urea in a loblolly pine plantation . Soil Sci. Soc. Am. J. 73 : 630 – 637 . doi:10.2136/sssaj2007.0377 Laiho, R. , F. Sanchez , A. Tiarks , P.M. Dougherty , and C.C. Trettin . 2003 . Impacts of intensive forestry on early rotation trends in site carbon pools in the southeastern US. For. Ecol. Manage. 174 : 177 – 189 . doi:10.1016/S0378-1127(02)00020-8 Lauer , D.K. , and B.R. Zutter . 2001 . Vegetation cover response and second-year loblolly and slash pine response following bedding and preand post-plant herbicide applications in Florida . South. J. Appl. For. 25 : 75 – 83 . Li , B. , S.E. McKeand , and R.J. Weir . 1999 . Tree improvement and sustainable forestry: Impact of two cycles of loblolly pine breeding in the U.S.A . For. Genet. 6 : 229 – 234 . Maier , C.A., K.H. Johnsen , P. Dougherty , D. McInnis , P. Anderson, and S. Patterson . 2012 . Eect of harvest residue management on tree productivity and carbon pools during early stand development in a loblolly pine plantation . For. Sci. 58 : 430 – 445 . Martin , T.A. , and E.J. Jokela . 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 . doi:10.1016/j.foreco.2004.01.004 Massey , F.J. 1951 . e Kolmogorov–Smirnov test for goodness of t. J. Am. Stat. Assoc. 46 : 68 – 78 . doi:10.1080/01621459.1951.10500769 Mehlich, A. 1984 . Mehlich 3 soil test extractant: A modication of Mehlich 2 extractant. Commun. Soil Sci. Plant Anal. 15 : 1409 – 1416 . doi:10.1080/00103628409367568 Miller , H.G. 1981 . Forest fertilization: Some guiding concepts . Forestry 54 : 157 – 167 . doi:10.1093/forestry/54.2.157 Miller , J.H., B.R. Zutter , S.M. Zedaker , M.B. Edwards, J.D. Haywood, and R.A. Newbold . 1991 . A regional study on the inuence of woody and herbaceous competition on early loblolly pine growth . South. J. Appl. For. 15 : 169 – 179 . Mitchell , R.J. , B.R. Zutter , D.H. Gjerstad , G.R. Glover , and C.W. Wood . 1999 . Competition among secondary-successional pine communities: A eld study of eects and responses . Ecology 80 : 857 – 872 . doi:10.1890/0012-9658(1999)080[0857:CASSPC]2.0.CO;2 Moore , D.J.P., S. Aref , R.M. Ho , J.S. Pippen , J.G. Hamilton , and E.H. De Lucia . 2006 . Annual basal area increment and growth duration of Pinus taeda in response to eight years of free-air carbon dioxide enrichment. Global Change Biol. 12 : 1367 – 1377 . doi:10.1111/j.1365-2486.2006.01189.x Morris , L.A. , S.A. Moss , and W.S. Garbett. 1993 . Competitive interference between selected herbaceous and woody plants and Pinus taeda L. during two growing seasons following planting . For. Sci. 39 : 166 – 187 . Nambiar , E.K.S. 1996 . Sustained productivity of forests is a continuing challenge to soil science. Soil Sci. Soc. Am. J. 60 : 1629 – 1642 . doi:10.2136/sssaj1996.03615995006000060006x Neary , D.G. , E.J. Jokela , N.B. Comerford , S.R. Colbert , and T.E. Cooksey . 1990a . Understanding competition for soil nutrients: e key to site productivity on southeastern Coastal Plain Spodosols. In: S.P. Gessel et al., editors, Sustained productivity of forest soils. Proceedings of the 7th North American Forest Soils Conference, Vancouver. July 24–28 1988. For. Publ., Faculty of For., Univ. of British Columbia, Vancouver, BC, Canada. p. 432–450. Neary , D.G. , D.L. Rockwood, N.B. Comerford , B.F. Swindel , and T.E. Cooksey . 1990b . Importance of weed control, fertilization, irrigation, and genetics in slash and loblolly pine early growth on poorly drained Spodosols . For. Ecol. Manage. 30 : 271 – 281 . doi:10.1016/0378-1127(90)90142-X Nemani , R.R. , C.D. Keeling , H. Hashimoto, W.M. Jolly , S.C. Piper , C.J. Tucker , et al . 2003 . Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science 300 : 1560 – 1563 . doi:10.1126/science.1082750 Nilsson, M.C., and D.A. Wardle. 2005 . Understory vegetation as a forest ecosystem driver: Evidence from the northern Swedish boreal forest. Front. Ecol. Environ 3 : 421 – 428 . doi:10.1890/1540-9295(2005)003[0421:UVAAFE]2.0.CO;2 NOAA . 2012 . Annual climatological summary, Florida. Natl. Clim. Data Ctr., Natl. Environ. Satellite, Data, and Inf. Serv., Asheville, NC. Outcalt , K.W. , and E.H. White. 1981 . Understory biomass and nutrients 2 year aer timber harvest in northern Minnesota . Can. J. For. Res. 11 : 305 – 308 . Persson, H. 1981 . e eect of fertilization and irrigation on the vegetation dynamics of a pine–heath ecosystem . Plant Ecol. 46–47 : 181 – 192 . doi:10.1007/BF00118396 Piatek , K.B., and H.L. Allen . 1999 . Nitrogen mineralization in a pine plantation een years aer harvesting and site preparation . Soil Sci. Soc. Am. J. 63 : 990 – 998 . doi:10.2136/sssaj1999.634990x Piatek , K.B., and H.L. Allen . 2001 . Are forest oors in mid-rotation stands of loblolly pine (Pinus taeda) a sink for nitrogen and phosphorus? Can. J. For. Res. 31 : 1164 – 1174 . doi:10.1139/x01-049 Polglase , P.J. , E.J. Jokela , and N.B. Comerford . 1992a . Nitrogen and phosphorus release from decomposing needles of southern pine plantations . Soil Sci. Soc. Am. J. 56 : 914 – 920 . doi:10.2136/sssaj1992.03615995005600030039x Polglase , P.J. , E.J. Jokela , and N.B. Comerford . 1992b . Phosphorus, nitrogen, and carbon fractions in litter and soil of southern pine plantations . Soil Sci. Soc. Am. J. 56 : 566 – 573 . doi:10.2136/sssaj1992.03615995005600020036x Powers , R.F. 1999 . On the sustainable productivity of planted forests . New For.

PAGE 16

þt Soil Science Society of America Journal 17 : 263 – 306 . doi:10.1023/A:1006555219130 Pritchett, W.L. , and N.B. Comerford . 1982 . Long-term response to phosphorus fertilization on selected southeastern Coastal Plain soils . Soil Sci. Soc. Am. J. 46 : 640 – 644 . doi:10.2136/sssaj1982.03615995004600030038x Pritchett, W.L. , and W.R. Llewellyn . 1966 . Response of slash pine (Pinus elliottii Engelm. var. elliottii) to phosphorus in sandy soils . Soil Sci. Soc. Am. J. 30 : 509 – 512 . doi:10.2136/sssaj1966.03615995003000040031x Qian , P. , and J.J. Schoenau . 2002 . Practical applications of ion exchange resins in agricultural and environmental soil research . Can. J. Soil Sci. 82 : 9 – 21 . doi:10.4141/S00-091 Rice, E.L. 1972 . Allelopathic eects of Andropogon virginicus and its persistence in old elds . Am. J. Bot. 59 : 752 – 755 . doi:10.2307/2441147 Rifai, S.W. , D. Markewitz , and B. Borders . 2010 . Twenty years of intensive fertilization and competing vegetation suppression in loblolly pine plantations: Impacts on soil C, N, and microbial biomass . Soil Biol. Biochem. 42 : 713 – 723 . doi:10.1016/j.soilbio.2010.01.004 Roth , B.E., E.J. Jokela , T.A. Martin , D.A. Huber , and T.L. White. 2007 . Genotype q environment interactions in selected loblolly and slash pine plantations in the southeastern United States . For. Ecol. Manage. 238 : 175 – 188 . doi:10.1016/j.foreco.2006.10.010 Sarkhot, D.V. , N.B. Comerford , E.J. Jokela , and J.B. Reeves . 2007 . Eects of forest management intensity on carbon and nitrogen content in dierent soil size fractions of a North Florida Spodosol . Plant Soil 294 : 291 – 303 . doi:10.1007/s11104-007-9255-z Sartori , F. , D. Markewitz , and B.E. Borders . 2007 . Soil carbon storage and nitrogen and phosphorous availability in loblolly pine plantations over 4 to 16 years of herbicide and fertilizer treatments . Biogeochemistry 84 : 13 – 30 . doi:10.1007/s10533-007-9072-8 SAS Institute . 2007 . SAS OnlineDoc 9.2. SAS Inst., Cary, NC. Schimel , J.P. , K.V. Cleve , R.G. Cates , T.P. Clausen , and P.B. Reichardt. 1996 . Eects of balsam poplar (Populus balsamifera) tannins and low molecular weight phenolics on microbial activity in taiga oodplain soil: Implications for changes in N cycling during succession . Can. J. Bot. 74 : 84 – 90 . doi:10.1139/b96-012 Scott , N.A., and D. Binkley . 1997 . Foliage litter quality and annual net N mineralization: Comparison across North American forest sites . Oecologia 111 : 151 – 159 . doi:10.1007/s004420050219 Sedjo, R.A. , and D. Botkin . 1997 . Using forest plantations to spare natural forests . Environment 39 : 14 – 30 . doi:10.1080/00139159709604776 Smethurst, P.J. , and E.K.S. Nambiar . 1995 . Changes in soil carbon and nitrogen during the establishment of a second crop of Pinus radiata. For. Ecol. Manage. 73 : 145 – 155 . doi:10.1016/0378-1127(94)03491-E Sprugel , D.G. 1983 . Correcting for bias in log-transformed allometric equations . Ecology 64 : 209 – 210 . doi:10.2307/1937343 Stone , E.L. 1990 . Boron deciency and excess in forest trees: A review . For. Ecol. Manage. 37 : 49 – 75 . doi:10.1016/0378-1127(90)90046-E Subedi , P. 2013 . Inter-rotational eects of fertilization and weed control treatments on the productivity and soil nutrient availability in juvenile loblolly pine plantations. M.S. thesis. Univ, of Florida, Gainesville. Swindel , B.F., D.G. Neary , N.B. Comerford , D.L. Rockwood, and G.M. Blakeslee. 1988 . Fertilization and competition control accelerate early southern pine growth on atwoods . South. J. Appl. For. 12 : 116 – 121 . Tessier , J.T. , and D.J. Raynal . 2003 . Vernal nitrogen and phosphorus retention by forest understory vegetation and soil microbes . Plant Soil 256 : 443 – 453 . doi:10.1023/A:1026163313038 Tilman, D. , P.B. Reich, J. Knops, D. Wedin , T. Mielke, and C. Lehman . 2001 . Diversity and productivity in a long-term grassland experiment. Science 294 : 843 – 845 . doi:10.1126/science.1060391 Tisdale, J.L. 2008 . uantifying the eects of organic residues on soil nitrogen and phosphorus availability. M.S. thesis. Dep. of For. and Environ. Resour., North Carolina State Univ., Raleigh. Turner , J. , and J.N. Long . 1975 . Accumulation of organic matter in a series of Douglas-r stands . Can. J. For. Res. 5 : 681 – 690 . doi:10.1139/x75-094 VanderSchaaf , C.L., J.A. Moore , and J.L. Kingery . 2002 . e eect of multinutrient fertilization on understory vegetation annual production . West. J. Appl. For. 17 : 147 – 153 . Vogel , J.G. , and E.J. Jokela . 2011 . Micronutrient limitations in two managed southern pine stands planted on Florida Spodosols . Soil Sci. Soc. Am. J. 75 : 1117 – 1124 . doi:10.2136/sssaj2010.0312 Vogel , J.G. , L.J. Suau , T.A. Martin , and E.J. Jokela . 2011 . Long-term eects of weed control and fertilization on the carbon and nitrogen pools of a slash and loblolly pine forest in north-central Florida . Can. J. For. Res. 41 : 552 – 567 . doi:10.1139/X10-234 Wear , D.N., and J.G. Greis . 2002 . Southern forest resource assessment: Summary of ndings . J. For. 100 : 6 – 14 . Williston, H.L., and S.J. Barras . 1977 . Impact of tip moth injury on growth and yield of 16-year-old loblolly and shortleaf pine. Res. Note SO-221. U.S. For. Serv., South. For. Exp. Stn., Asheville, NC. Zak , D.R. , W.E. Holmes, D.C. White, A.D. Peacock , and D. Tilman . 2003 . Plant diversity, soil microbial communities, and ecosystem function: Are there any links? Ecology 84 : 2042 – 2050 . doi:10.1890/02-0433 Zerpa , J.L. , and T.R. Fox . 2011 . Controls of volatile ammonia losses from loblolly pine plantations fertilized with urea in the southeast USA . Soil Sci. Soc. Am. J. 75 : 257 – 266 . doi:10.2136/sssaj2010.0101 Zhang , J. , R.F. Powers , W.W. Oliver , and D.H. Young . 2013 . Response of ponderosa pine plantations to competing vegetation control in northern California, USA: A meta-analysis . Forestry 86 : 3 – 11 . doi:10.1093/forestry/cps054 Ziadi, N. , R.R. Simard , G. Allard , and J. Lafond . 1999 . Field evaluation of anion exchange membranes as a N soil testing method for grasslands . Can. J. Soil Sci. 79 : 281 – 294 . doi:10.4141/S98-062 Zutter , B.R., R.J. Mitchell , G.R. Glover , and D.H. Gjerstad . 1999 . Root length and biomass in mixtures of broomsedge with loblolly pine or sweetgum . Can. J. For. Res. 29 : 926 – 933 . doi:10.1139/x99-059



PAGE 1

EFFECTS OF SILVICULTURAL MANAGEMENT INTENSITY AND GENETIC S ON SOIL RESPIRATION AND BELOWGROUND CARBON ALLOCATION IN LOBLOLLY PINE PLANTATIONS By CHELSEA GILL DRUM A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TH E DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014

PAGE 2

© 2014 Chelsea Gill Drum

PAGE 3

To my family for the inspiration and encouragement that has guided me through all of my adventures . To my loving husband for his support throughout my studies .

PAGE 4

4 ACKNOWLEDGMENTS I would like to share my gratitude with my major advisor, Dr. Eric J. Jokela, for his support, guidance, and time. Without his continued support, I would not have achieved the level of expertise that is required to succeed not only in graduate school, but as a the sometimes elusive nature of belowground results. I am also thankful for all of his efforts throughout my graduate study and for his patience, help, and develo pment of my understanding of belowground processes. Many thanks go to Dr. Salvador Gezan for his constant efforts to ensure my statistical analyses meet the high standards required for publication. I would like to thank Dr. Timothy A. Martin for his suppor t during my studies , helpful advice , and time . I am grateful to Dr. Edward A. G. Schuur for his assistance with the development of this study and the invaluable use of his lab and time. I am very grateful for the funding provided by the Forest Biology Research Cooperative Program, both of whom made my graduate studies possible. Much of the field dat a could not have been collected without the tireless efforts technicians and members. I would like to thank the following people for their hard work and collaboration throughout this study: G. Lokuta, A. Milligan, and B. Gottloeb (collecting inven tory data, soil samples, litterfall samples, forest floor samples, and measuring soil CO 2 efflux); J. McCafferty, B. Caudill, and B. Ruffi n (collecting and processing litterfall) . I would especially like to thank Dr. Rosvel G. Bracho Garillo for teaching m e soil CO 2 efflux sampling techniques , for his countless hours in the field,

PAGE 5

5 and for his guidance. I would also like to thank A. Garcia, M. Wightman, and P. Subedi for their moral support and comradeship . Many thanks go to my friends, especially John, Kim , Bailey, and Zak , for their camaraderie throughout my graduate studies. I would like to recognize my parents, who from my earliest have encouraged my love of science and the outdoors. I would like to thank my brother for his continual confidence in my abi lities. Finally, I thank my husband for his patience and dedication towards the conclusion of this study.

PAGE 6

6 TABLE OF CONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ................................ . 4 LIST OF TABLES ................................ ................................ ................................ ........... 8 LIST OF FIGURE S ................................ ................................ ................................ ...... 11 LIST OF ABBREVIATIONS ................................ ................................ .......................... 14 ABSTRACT ................................ ................................ ................................ .................. 15 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ... 17 Backgr ound ................................ ................................ ................................ ........... 17 Problem ................................ ................................ ................................ ................. 22 Objectives ................................ ................................ ................................ .............. 22 2 EFFECTS OF FERTILIZATION AND WEED CONTROL TREATMENTS AND GENETICS ON SOIL CO 2 EFFLUX, ROOT BIOMASS, AND ABOVEGROUND BIOMASS IN PINUS TAEDA (L.) PLANTATIONS ................................ ................. 24 Methods ................................ ................................ ................................ ................ 27 Study Areas ................................ ................................ ................................ .... 27 Study Designs ................................ ................................ ................................ . 29 Soil Respiration ................................ ................................ ............................... 32 Carbon Compo nents for Loblolly Pine ................................ ............................. 32 Root Measurement ................................ ................................ .......................... 33 Data Analyses ................................ ................................ ................................ . 33 Results ................................ ................................ ................................ .................. 34 Aboveground Loblolly Pine Carbon Accumulation ................................ ........... 34 Mean Soil Respiration: Bed and Inter bed Comparison ................................ ... 35 Mean Soil Respiration: Bed and Root Exclusion Comparison ......................... 37 Fertilization of Bed and Inter bed at ACF ................................ ........................ 39 Root Carbon ................................ ................................ ................................ .... 39 Discussion ................................ ................................ ................................ ............. 40 Summary and Conclusions ................................ ................................ .................... 48 3 EFFECTS OF FERTILIZATION AND WEED CONTROL TREATMENTS AND GENETICS ON ANNUAL SOIL RESPIRATION, LITTERFALL, AND TOTAL BELOWGROUND CARBON FLUX IN PINUS TAEDA (L.) PLANTATIONS ........... 85 Methods ................................ ................................ ................................ ................ 89

PAGE 7

7 Study Areas ................................ ................................ ................................ .... 89 Study Designs ................................ ................................ ................................ . 90 Soil Respiration ................................ ................................ ............................... 93 Litterfall ................................ ................................ ................................ ........... 93 Aboveground C Pool s ................................ ................................ ..................... 93 Net primary production ................................ ................................ .............. 94 Forest Floor ................................ ................................ ................................ ..... 94 Soil Respiration Models ................................ ................................ ................... 94 Results ................................ ................................ ................................ .................. 97 Environmental Variables and Soil CO 2 Efflux ................................ .................. 98 Annual Soil Respiration ................................ ................................ ................... 99 Annual Litterfall ................................ ................................ ............................. 100 Total Belowground Carbon Flux ................................ ................................ .... 101 Bed and Root Exclusion Soil Respiration Comparison ................................ .. 102 Forest Floor ................................ ................................ ................................ ... 103 Carbon Turnover Time ................................ ................................ .................. 104 Discussion ................................ ................................ ................................ ........... 105 Summary and Conc lusions ................................ ................................ .................. 113 4 CONCLUSIONS ................................ ................................ ................................ .. 141 LIST OF REFERENCES ................................ ................................ ............................ 147 BIOGRAPHICAL SKETCH ................................ ................................ ......................... 157

PAGE 8

8 LIST OF TABLES Table page 2 1 Cumulative fertilization rates for the loblolly pine experimental sites, Austin Cary Forest (ACF) and Sanderson in north Florida. ................................ ........... 50 2 2 Analysis of variance for aboveground carbon mass by tree component for the Austin Cary Forest in north Florida. . ................................ ................................ . 50 2 3 Means of aboveground carbon mass by tree component for the Austin Cary Forest in north Florida. ................................ ................................ ...................... 51 2 4 Analysis of variance for aboveground carbon mass by tree component for the Sanderson site in north Florida.. ................................ ................................ ........ 52 2 5 Means of aboveground carbon mass by tree component for the Sanderson site in north Florida. ................................ ................................ ........................... 53 2 6 Analysis of variance for soil respiration rates for bed and inter bed positions at the Austin Cary Forest in north Florida. ................................ ......................... 53 2 7 Analysis of variance for soil respiration rates for bed and inter bed positions at the Sanderson site in north Florida. ................................ ............................... 54 2 8 Analysis of variance for soil respiration rates for bed and root exclusion positions at the Austin Cary Forest in north Florida. ................................ .......... 55 2 9 Analysis of variance for soil respiration rates for bed and root exclusion positions at the Sanderson site in north Florida. ................................ ................ 56 2 10 Analysis of variance for soil respiration rates for bed and inter bed positions during pre fertilization and post fertilization periods at the Austin C ary Forest in north Florida. ................................ ................................ ................................ . 57 2 11 Means of soil respiration rates for bed and inter bed positions during pre and post fertilization periods at the Austin Cary Forest in north Florida. ................... 58 2 12 Analysis of variance for root carbon mass by root d iameter class for the Austin Cary Forest in north Florida.. ................................ ................................ .. 59 2 13 Means of root carbon mass by diameter class for the Austin Ca ry Forest in north Florida.. ................................ ................................ ................................ .... 59 2 14 Analysis of variance for root carbon mass by root diameter class for the Sanderson site in no rth Florida. ................................ ................................ ......... 61

PAGE 9

9 2 15 Means of root carbon mass by diameter class for the Sanderson site in north Florida.. ................................ ................................ ................................ ............. 61 2 16 Arrhenius temperature response parameters and Q10 values for the Austin Cary Forest in north Florida. ................................ ................................ .............. 62 2 17 Arrhenius temperature response parameters and Q10 values for the Sanderson site in north Florida. . . ................................ ................................ ....... 62 3 1 Analysis of variance for net primary productivity for the Austin Cary Forest in north Florida. ................................ ................................ ................................ ... 115 3 2 Means for net primary productivity for the Austin Cary Forest in north Florida. 115 3 3 Analysis of variance for net primary productivity for the Sanderson site in north Florida. ................................ ................................ ................................ ... 115 3 4 Means for net primary productivity for the Sanderson site in north Florida. ...... 116 3 5 Coefficient estimates and R 2 emp of soil re spiration models for the Austin Cary Forest in north Florida. ................................ ................................ .................... 117 3 6 Coefficient estimates and R 2 emp of soil respiration models for the Sanderson site in north Florida. ................................ ................................ ......................... 118 3 7 Soil respiration regression coefficients for the Aus tin Cary Forest and Sanderson site in north Florida. ................................ ................................ ....... 119 3 8 Repeated measures analysis of variance for annual soil respiration f or bed and inter bed positions at the Austin Cary Forest in north Florida. ................... 120 3 9 Mean annual soil respiration for the bed and inter bed positions at the Austin Cary Forest in north Florida. ................................ ................................ ............ 120 3 10 Repeated measures analysis of variance for annual soil re spiration for bed and inter bed positions at the Sanderson site in north Florida. ........................ 121 3 11 Mean annual soil respiration for the be d and inter bed positions at the Sanderson site in north Florida. ................................ ................................ ....... 121 3 12 Repeated measures analysis of variance for annual litte rfall for bed and inter bed positions at the Austin Cary Forest in north Florida. ................................ . 122 3 13 Mean annual litterfall for the bed and inter bed positions at the Austin Cary Forest in north Florida. ................................ ................................ .................... 122 3 14 Repeated measures analysis of variance for annual litterf all for bed and inter bed positions at the Sanderson site in north Florida. ................................ ....... 123

PAGE 10

10 3 15 Mean annual litterfall for the bed and inter bed positions at the Sanderson site in north Florida.. ................................ ................................ ........................ 124 3 16 Repeated measures analysis of variance for total belowground carbon flux for bed and inter bed positions at the Austin Cary Forest in north Florida. ....... 125 3 17 Mean total belowground carbon flu x for the bed and inter bed positions at the Austin Cary Forest in north Florida. ................................ ................................ . 125 3 18 Repeated measures analysis of variance for total belowground carbon flux for bed and inter bed positions at the Sanderson site in north Florida. ............ 126 3 19 Mean total belo wground carbon flux for the bed and inter bed positions at the Sanderson site in north Florida. ................................ ................................ ....... 126 3 20 Repeated measures analysis of variance for annual soil respiration for bed and root exclusion positions at the Austin Cary Forest in north Florida. ........... 127 3 21 Repeated measures analysis of variance for annual soil respiration for bed and root exclusion positions at the Sanderson site in north Florida. ................ 127 3 22 Analysis of variance for forest floor carbon for the Austin Cary Forest in north Florida.. ................................ ................................ ................................ ........... 128 3 23 Mean forest floor carbon for the Austin Cary Forest in north Florida. ............... 128 3 24 Analysis of var iance for forest floor carbon for the Sanderson site in north Florida.. ................................ ................................ ................................ ........... 129 3 25 Mean forest floor carbon for the Sanderson site i n north Florida. .................... 130

PAGE 11

11 LIST OF FIGURES Figure page 2 1 Mean for total aboveground loblolly pine carbon accumulation at the Sanderson site in north Florida across two years of measurement for slow and fast growers. ................................ ................................ ............................... 63 2 2 Mean of total aboveground loblolly pine carbon accumulation at the Sanderson site in north Florida across two years of measurement for high and operational intensity management. ................................ ............................. 64 2 3 Means for soil respiration rates for bed and inter bed positions and families at the Austin Cary Forest and the Sanderson site in north Florida. ........................ 65 2 4 Means for soil respiration rates by intensity at the Austin Cary Forest and the Sanderson site in north Florida. ................................ ................................ ......... 66 2 5 Means for soil respiration rates for bed and inter bed positions at the Austin Cary Forest and the Sanderson site in north Florida. ................................ ......... 67 2 6 Means for soil respiration rates by family at the Austin Cary Forest and the Sanderson site in north Florida. ................................ ................................ ......... 68 2 7 Means for soil respiration rates for bed and inter bed positions and intensities at the Sanderson site in north Florida. ................................ ............................... 69 2 8 Means for soil respiration rates for the interaction of management intensity and position at the Sanderson site in north Flo rida. ................................ ........... 70 2 9 Means for soil respiration rates for the interaction of management intensity, family, and position at the Sanderson site in north Florida. ................................ 71 2 10 Means for soil respiration rates for bed and root exclusion positions at the Austin Cary Forest and the Sanderson site in north Florida. .............................. 72 2 11 Means for soil respiration rates for bed and root exclusion positions and intensity at the Austin Cary Forest and the Sanderson site in north Florida. ...... 73 2 12 Means for soil respiration rates for bed and root exc lusion positions and family at the Sanderson site in north Florida.. ................................ .................... 74 2 13 Means for soil respiration rates for the interaction of bed and root exclusion positions, family, and measurement date at the Sanderson site in north Florida. ................................ ................................ ................................ .............. 75 2 14 Means for soil respiration rates for the interaction of bed and root exclusion positions and intensity at the Sanderson site in north Florida. ........................... 76

PAGE 12

12 2 15 Means for soil respiration rates for bed and root exclusion positions with an intensity x family interaction at the Austin Cary Forest in north Florida.. ............ 77 2 16 Means for soil respiration rates for pre and post fertilization periods and intensity at the Austin Cary Forest in north Florida. ................................ ........... 78 2 17 Temperature response curves for soil respiration at the Austin Cary Forest in north Florida. ................................ ................................ ................................ ..... 7 9 2 18 Temperature response curves for soil respiration at the Sanderson site in north Florida. ................................ ................................ ................................ ..... 80 2 19 Means for soil moisture (%) through time at the Sanderson site for the bed and root exclusion positions of measurement. ................................ ................... 81 2 20 high and operational, at the Austin Cary Forest for three measurement positions ................................ ................................ ................................ ............ 82 2 21 Florisa for three measurement positions ................................ ............................ 83 2 22 Palmer drought severity index over the study period for north central Florida. ... 84 3 1 The interaction of soil respiration, soil temperature, and relative humidity at the Austin Cary Forest in north Florida over three years of measurement. ...... 131 3 2 The response of soil respiration to soil temperature at the Sanderson site in north Florida over two years of measurement. ................................ ................. 132 3 3 Effects of management treatments and positions on the mean fluxes of soil respiration, litterfall, and total belowground carbon flux at the Austin Cary Forest and Sanderson site. ................................ ................................ ............. 133 3 4 Effects of management treatments and families on the mean fluxes of soil respiration, litterfa ll, and total belowground carbon flux at the Austin Cary Forest and Sanderson site. ................................ ................................ ............. 134 3 5 Means for annual soil CO 2 efflux for the interaction of management intensity x family x measurement position at the Austin Cary Forest in north Florida.. ... 135 3 6 Means for annual soil CO 2 efflux for the interaction of measurement position and family at the Sanderson site in north Florida. ................................ ............ 136 3 7 Means for annual soil CO 2 efflux for the interaction of measurement position and management intensity at the Sanderson site in north Florida.. .................. 137

PAGE 13

13 3 8 Means for forest floor C for the interaction of sampling position and management intensity at the Austin Cary Forest in north Florida.. ................... 138 3 9 Means for forest floor C for the interaction of sampling position and horizon at the Austin Cary Forest in north Florida.. ................................ .......................... 139 3 10 Means for forest floor C for the interaction of horizon and family at the Sanderson site in north Florida. ................................ ................................ ....... 140

PAGE 14

14 LIST OF ABBREVIATIONS F Family H Humidity I Intensity Inter Inter bed collar position M Measurement Oper Operational intensity management P Collar position R Radiation RA Autotrophic soil respiration RE Root exclusion RH Heterotrophic soil respiration SR Soil respiration TA Air temperature TA10 Air temperature at 10 meters TS Soil temperature VPD Vapor pressure deficit

PAGE 15

15 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for th e Degree of Master of Science EFFECTS OF SILVICULTURAL MANAGEMENT INTENSITY AND GENETIC S ON SOIL RESPIRATION AND BELOWGROUND CARBON ALLOCATION IN LOBLOLLY PINE PLANTATIONS By Chelsea Gill Drum August 2014 Chair: Eric J. Jokela Major: Forest Resources and Conservation I n the southeastern U.S. , fertilization and weed control treatments, along with the deployment of genetically improved plantin g stock, are routinely used to increase aboveground productivity. This project examined the effects of intensive management and genetic selection of loblolly pine ( Pinus taeda L.) on soil respiration (SR) and below g round carbon (C) allocation . In Florida, two field installations of two families of loblolly pine, design with two levels of nitrogen and phos phorus fertilization (high and operational intensity ) on Spodosols . Measurements of root biomass and repeated measurements of forest growth, SR , and litterfall were made over multiple years. Soil respiration an d litterfall measurements were used to estimate Total Belowground Carbon Flux (TBCF), which provided an estimat e of C allocation to roots. Soil respiration varied temporally, with significant effects (p<0.05) of family x time, intensity x time, and location x time for both the bed and inter bed and root exclusion positions . Heterotrophic respiration did not vary b etween management intensities and comprised approximately 80% of total SR. Modelled respiration rates

PAGE 16

16 resulted in an intensity x family interaction for SR and TBCF, indicating that TBCF d id not differ for either family under intensive management, but they did differ under the operational management regime . These findings, therefore, lend support that genotype x environment interactions can occur not only aboveground but also belowground. The results also argue that genotype x environment interactions shoul d be considered in the southeastern United States when estimating forest C budgets .

PAGE 17

17 CHAPTER 1 INTRODUCTION Background Forest management intensity is increasing globally, but most especially in the southeastern United States where loblolly pine ( Pinus taeda L.) forests are intensively managed for wood production (Jokela et al., 2004). Managed loblolly pine forests are among the most productive forests in the world and s ignificant gains in the productivity of pine plantations have resulted from fer tilization, weed control, and the selection of elite genotypes for disease resistance and faster growth (Fox et al., 2007) . Intensive management is expected to become more common in southern pine silviculture as forested land area decreases from competing uses and forest product demand s increase (Mickler et al., 2004). Currently, global climate change and carbon (C) sequestration are at the forefront of scientific and political discourse. C arbon cycles between the atmosphere and the terrestrial biosphere, where soils and forests sequester atmospheric C. Now, intensive forest management is viewed as a n important means for enhancing and sequestering atmospheric carbon dioxide (CO 2 ) (Lal , 2008). Abo veground productivity increases have been achieved with intensive forest management (Jokela et al., 2004), yet it is still poorly understood how these management systems will affect possible changes in soil organic carbon , specifically soil organic matter (SOM) and soil respiration (SR) ( Jandl et al., 2007 ) . Soil CO 2 efflux or soil respiration (SR) is the sum of autotrophic soil respiration (RA), CO 2 respired by plant roots and fungi, and heterotrophic soil respiration (RH), CO 2 respired by soil microorganisms (Luo and Zhou, 2006). Soil respiration is dependent on

PAGE 18

18 a variety of biotic and abiotic factors that control plant root respiration and microbial respiration including s oil temperature, soil moisture content, aboveground veg etation and belowground carbon allocation, soil carbon and nutrient content, disturbance regimes, soil porosity, CO 2 pressure gradients, and surface wind speed and air turbulence (Raich and Tufekciogul, 2000; Ryan and Law, 2005; Luo and Zhou, 2006). Loblol ly shortleaf (Pinus echinata) pine ecosystems are the second most widespread forest types in the southeastern United States (Mickler et al . , 2004). Almost twelve million hectares of loblolly pine had been planted at the end of the twentieth century in the United States (Fox et al. , 2007). The silvicultural regimes used to manage southern pine plantations are some of the most rigorous in the world (Fox et al. 2007). Silvicultural practices involve mechanical site preparation, the use of genetically elite fam ilies, weed control through herbicide use, and fertilization to correct nutrient limitations. Southern pine forests have been recognized as a regional sink of atmospheric carbon, annually sequestering about 76 Tg C, or the equivalent of 13% of greenhouse gas emissions (Richter et al. , 1995) . Southern tree improvement programs have mainly focused on improving volume growth, tree form, disease resistance, and wood quality (Zobel and Talbert , 1984). Genetic manipulation holds the key for potenti al advances in southern pine forestry (Fox et al. , 2007). What is unknown is the effect of genetic selection for these traits on belowground processes. However, it is known that coarse root biomass: total tree biomass is nearly constant for loblolly pine (Retzlaff et al. , 2001; Albaugh et al. 2006 ; Coyle et al. , 2008). Therefore, it is likely that if there are genetic changes in C allocation,

PAGE 19

19 it would be evident in fine roots of loblolly pine instead of coarse roots (Adegbidi et al ., 2002). The source of g enetic materials being deployed in reforestation has had varying effects on belowground C components. This has been demonstrated with Scots pine ( Pinus sylvestris L.) , with fine root biomass varying with geographic origin of the germplasm (Oleksyn et al. , 1999). For this study, it is significant to note that soil aggregation has varied with genetically unique families of loblolly pine (Sarkhot et al. 2007a, b, 2008). Also, i n a 32 year old black spruce ( Picea mariana (Mill.)) stand , Major et al. (2012) fou nd that belowground growth followed aboveground growth and that genetics impacted belowground C storage. Similarly, Lee and Jose (2003) documented that the effects of nitrogen (N) fertilization may be species specific; a seven y ear old Populus deltoid e s (M stand exhibited no change . In loblolly pine, a greenhouse study showed that SR and root exudates differ ed between closely related clones (Stovall et al., 2013). Conversely, another study dem on strated that no significant difference in SR were due to genetics (Tyree et al., 2008). However, in the two first years of development with two contrasting loblolly pine clones, Tyree et al. (2013) found differences in SR and RH. Therefore, genotypic vari ation in belowg round C processes appears possible with loblolly pine. Competition from hardwoods and other vegetation has been an obstacle for souther n pine plantation forestry (Clason , 1978 ; Cain and Mann , 1980). Herbicides are used to control hardwoods and other competing vegetation like grasses . Eliminating hardwood competition has been shown to increase southern pine growth by 100% or

PAGE 20

20 more (Burkhart and Sprinz , 1984 ; Minogue et al. , 1991 ; Borders and Bailey , 2001 ; Amishev and Fox , 2006). Controlling herbaceous weeds with herbicides such as hexazinone, imazapyr, and sulfometuron methyl has also significantly improved the growth of pines (Holt et al. , 1973 ; Fitzgerald , 1976; Nelson et al. , 1981). Herbaceous weed control has been shown to inc rease pine growth in the long term (Glover et al. , 1989). Thus, during the first growing season, it is now a common practice to control hardwoods and herbaceous competition (Minogue et al. , 1991). Weed control has been documented to affect factors that are tied to SR, like SOM, microbial biomass and composition, and the soil environment (Ibell et al., 2010; Busse et al., 1996; Ratcliff et al., 2006; Devine and Harrington, 2007; Parker et al., 2009; Curiel Yuste et al., 2007). In addition, d ecreases in SOM h ave been associated with the application of herbicides in southern pine stands (Shan et al., 2001; Sarkhot et al., 2007; Vogel et al., 2011). For example, w ith a decrease in competing vegetation, a decrease in soil organic substrates could lead to a declin e in microbial activity (Blazier et al., 2005), directly affecting RH and therefore SR. Despite the ties of herbicide use to these factors, changes in SR due to the application of herbicides are still poorly understood, especially in managed southern pine ecosystems (Busse et al., 2006). Like herbicide use, f ertilization is a significant component of southern plantation forest management. The North Carolina State Forest Fertilization Cooperative and the Cooperative Research in Forest Fertilization (CRIFF) program at the University of Florida found that slash (Pinus elliottii var. elliottii Engelm ) and loblolly pine growth is often nutrient limited by both phosphorus (P) and N (Fisher and Garbett 1980;

PAGE 21

21 Comerford et al ., 1983 ; Gent et al. , 1986 ; Allen , 1987 ; Jokela and Stearns Smith , 1993 ; Hynynen et al. , 1998 ; Amateis et al. , 2000). The same studies demonstrated that mid rotation fertilization result ed in consistent and sizeable growth responses with the application of N (1 70 220 kg.h a 1 of N) and P ( 30 60 k g.ha 1 of P). Combined N and P fertilization has result ed in 25% greater growth in loblolly pine plantations growing on deficient soils (Fox et al. , 2007). Fertilization to amend N and P site limitations has , therefore, increased significantly throughout t he South. Mid rotation fertilization of southern pine plantations has grown from 6000 hect acres annual ly in 1988 to approximately 485,000 567,000 hectares. year 1 since 2000 (Fox et al. , 2007). In southern pine plantations alone, about 6,475,000 hecta res ha ve received various inputs of fertilize r s over time (Albaugh et al. , 2007). Management, especially fertilization, in plantation forestry may obscure the possible responses of SOM to silvicultur al treatments because fertilization can cause conflicting bel owground responses. Increased nutrient availability from fertilization may favor the allocation of net primary productivity (NPP) to aboveground components (Haynes and Gower , 1995). Therefore, it is expected that fertilization c ould cause a decrease in C a l location to plant roots, which would cause a decrease in SOM (Haynes and Gower , 19 95 ; Albaugh et al. , 1998 ; Palmroth et al. , 200 6 ). In contrast , decreases in microbial heterotrophic respiration have been observed following fertiliz er additions , which c ould lead to an increase in SOM (Wallenstein et al. , 2006 ; Treseder , 2008). Some studies have shown that the type of SOM pool matters, with N fertilization decomposing labile pools of SOM faster than recalcitrant pools (Neff et al. , 2002 ; Torn

PAGE 22

22 et al. , 2005 ). Moreover, SOM pools could remain the same if inputs and outputs were balanced. Problem Assessing the ability of forests to actively sequester CO 2 is a crucial step in determining if increased management intensity is a viable option for enhanc ing ecosy stem C sequestration (Lal , 2008). In forest ecosystems, aboveground productivity and C sequestration have been studied at multiple scales. In contrast, studying belowground productivity and C sequestration processes may be limited by difficulties in meas urement techniques , which undoubtedly has contributed to the fewer number of published belowground versus aboveground studies (Norby and Jackson , 2000). The amount of SOM in forest ecosystems worldwide, approximately 787 Pg, is greater than the atmospheric C pool (Dixon et al. , 1994 ; Smith et al. , 2008). A change in belowground processes could result in gains or loss es in SOM and substantially affect the rate of atmospheric CO 2 accumulation (Trumbore , 2000). For example, l oss of SOM not only mean s less C sequestration, but could include increased SR and CO 2 evolution. Consequently, it is essential to determine how SOM turnover will be affected by forest management decisions that may alter belowground processes. Such information will be important for predicting C fluxes and developing regional and global C management strategies. Objectives In north Florida, two replicated experiments located on Spodosols were established each using different genetic sources of loblolly pine (fast and slow grower s) , and two levels of silvicultural intensity, high and operational. The genetics of the fast growing family was similar between study sites, but the slow growers were different.

PAGE 23

23 Measurements and data were collected over three growing seasons at the Austi n Cary Forest site and two growing seasons at Sanderson site. Specifically, this study examined the impacts of fertilization and weed control treatments and pine genotype s on 1. autotrophic and heterotrophic soil respiration through time and root biomass (Ch apter 2) ; and 2. carbon content of aboveground bi omass, litterfall, and forest floor, as well as the C allocated to soil respiration and total belowground carbon flux of the pine components (Chapter 3). Results from this study should reduce the ambiguity of belowground processes under different levels of fertilization, weed control, and genetic selection. Furthermore, the results should add to the understanding of management strategies and inform development of sustainable management practices used for loblo lly pine plantations.

PAGE 24

24 CHAPTER 2 EFFECTS OF FERTILIZATION AND WEED CONTROL TREATMENTS AND GENETIC S ON SOIL CO 2 EFFLUX, ROOT BIOMASS, AND ABOVEGROUND BIOMASS IN PINUS TAEDA (L.) PLANTATIONS Soil respiration (SR) is a key flux of carbon dioxide from soils to the atmosphere and is the second major terrestrial flux of carbon in the global carbon cycle after gross primary productivity (Raich and Schlesinger, 1992; Raich et al., 2002; Ryan and Law, 2005) .The two main components of SR are heterotrophic respiration and autotrophic respiration. He terotrophic soil respiration (RH ) is the result of microbial respiration due to the decomposition of soil organic matter (S OM). Autotrophic soil respiration (RA) is the result of root respiration, and the respiration coming from the heterotrophs closely associated with roots e.g. , with mycorrhizal fungi, generally being classified as part of root respiration. Variability in SR has been attributed to a number of sources, including soil temperature and moisture , the effects of plant C allocation patterns to roots , and other C inputs, such as detritus and litterfall (Bowden et al. 1993; Crow et al. 2009; Chen et al., 20 11; Raich and Tufekciogul, 2000; Ryan and Law, 2005 ). Physical factors can also influence soil CO 2 efflux , including surface wind speed and air turbulence, soil porosity, and CO 2 pressure gradients (Luo and Zhuo, 2006). In the southeastern United States, loblolly p ine ( Pinus taeda L.) forests are intensively managed for wood production (Jokela et al. , 2004). Managers routinely use improved genotypes, fertilizers, and herbicides to increase forest yield. As the future demand for forest products increases, it is like ly that the intensity of management activities will also increase (Mickler et al. , 2004). Advances in genetics have contributed to enhancing aboveground productivity by generating fast growing and disease resistant families and clones (McKeand et al. , 200 3). When coupled with appropriate silvicultural

PAGE 25

25 systems, treatment x family interactions may lead to rank changes in aboveground growth and quality traits among gene tic sources (McKeand et al., 2006 ; Roth et al. , 2006). Although the effects of integrated m anagement practices on aboveground growth are well defined (Jokela et al. , 2004) , few studies have examined the interaction between genetics and silvicultural practices in the context of belowground processes. For example, it is unclear if these increases in aboveground productivity translate into belowground C gains (Litton et al. , 2007) . Major et al. (2012) demonstrated that belowground growth mirrored aboveground growth and that genetics played an important role in the determination of belowground C stor age in a 32 year old black spruce ( Picea mariana (Mill.)) stand. In southeastern U.S. forests, soil C pool dynamics has been identified as one of the limitations in assessing the C stores and fluxes of the region (Raich and Schlesinger, 1992; Johnson et a l., 2001; Luo, 2007). Soil respiration makes up 50 60% of total ecosystem C cycling through temperate forests (Post et al., 1982; Raich and Schlesinger, 1992; Lal, 2005; Noormets et al. , 2010; Bracho et al., 2012). Although significant drivers of SR have been identified, the response of SR to different land management practices are uncertain (Schlesinger and Andrews, 2000; Ryan and Law, 2005; Luo and Zhuo, 2006). With fertilization , SR has increase d (Tyree et al., 2006; Contosta et al., 2011), decrea se d (Maier and Kress, 2000; Olsson et al., 2005), or has not changed (Pangle and Seiler, 200 2 ; Samuelson et al., 2009). Little information exists on the influence of genetic deployment on SR. However, Stovall et al. (2013) demonstrated in a short term gre enhouse study that SR and root

PAGE 26

26 exudates were different between two closely related loblolly pine clones. Another clonal study showed no difference in SR due to genetics (Tyree et al., 2008). Conversely, Tyree et al. (2013) found differences in SR and RH be tween two contrasting loblolly pine genotypes (narrow crown versus broad crown clones) in the first two years of stand development. Across genera the effects of nitrogen (N) fertilization have been species specific, where seven year old Populus deltiodes added N , while no change s in SR were documented for loblolly pine growing on sandy loams in northern Florida (Lee and Jose, 2003). Controlling understory competition with herbicides is often used in conjunction with fert ilizer additions for increasing southern pine forest productivity. Limited evidence in managed southern pine ecosystems suggest that the use of herbicides has been associated with decreases in SOM (Shan et al., 2001; Sarkhot et al., 2007; Vogel et al., 20 11). By controlling competing vegetation, microbial activity may decline due to a decrease in soil organic substrates (Blazier et al., 2005), thereby reducing RH and thus total SR. Just as SR has been shown to be affected by factors like SOM, microbial bio mass and composition, and the soil environment, weed control has also been shown to affect those factors as well (Ibell et al., 2010; Busse et al., 1996; Ratcliff et al., 2006; Devine and Harrington, 2007; Parker et al., 2007; Curiel Yuste et al., 2007). H owever, alterations in SR due to herbicide use are still generally poorly understood (Busse et al., 2006). This research was undertaken to examine the effects of varying levels of intensive forest management, through genetic deployment, fertilization and herbicide application , on SR. This study provided the unique opportunity to examine genotype x

PAGE 27

27 environment interactions on belowground processes such as SR . The following objectives were addressed: 1. E stimate contributions of autotrophic and heterotrophic respiration to soil CO 2 efflux as influenced by forest management intensity and genetic s ; 2. D etermine the influence of fertilization on soil respiration rates, both heterotrophic and autotrophic ; and 3. D etermine how aboveground stand characteristics, environ mental factors, and root biomass relate to the components of soil respiration . These objectives were addressed by examining SR and aboveground C accumulation in two replicated, long term field experiments in north central Florida. These studies included the deployment of different sources of genetic materials for loblolly pine and two levels of manag ement intensity, as affected by weed control and fertilizer applications. Methods Study Areas Two managed loblolly pine forests in north central Florida were utilized in this study. orest (ACF) (29 , 82 a teaching and research forest, approximately 14 km northeast of Gainesville, FL, USA. The elevation at ACF is approximately 44 m above sea level. The ACF site was established to examine the effects of silviculture and taxa differences on plantation health and productivity of loblolly pine . The ACF site include d t wo levels of silviculture that demonstrate d the evolution of management practices from the 1980s and 2000s. The treatments include d high intensity (2000s, high fertilization, weed control, double pass bed) and operational intensity (1980s, operational

PAGE 28

28 fertilization, operational weed control, single pass bed) treatments , two full sib loblolly pine families, and one spacing (1.8 m x 3 m). The second site was located near Sanderson, FL (30 . In 2000, the Forest Biology Research Cooperative (FBRC) from the University of Florida, established a trial series to examine the interactions of full sib loblolly and slash pine ( Pinus elliotii var . elliottii Engelm. ) families under different environmental factors (Roth et al. 2007). The study , referred to as PPINES (Pine Productivity Interactions on Experimental Sites) , is a regional experiment that examine d the effects of species, genotype, silvic ultur al intensity , and planting density across a range of sites in the s outheastern USA (Roth et al. 2007). Both study locations shared a subtropical and humid climate with long hot wet summers and mild dry winters. Both studies were established on poorly drained Spodosols ( Pomona and Leon series for the ACF and Sanderson sites , respectively ) . The Pomona series is classified as sandy, siliceous, hyperthermic Ultic Alaquods , while the Leon series are sandy, siliceous, thermic Aeric Alaquods . The Pomona series has a Bh (spodic) horizon within 74 cm of the surface and a Btg (argillic) horizon at a depth of 130 cm. The Leon series has a Bh horizon within 38 cm of the surface , but does not contain an argillic horizon. Understory vegetation was sp arse relative to natural areas in all plots , but where found it was typical of Coastal Plain flatwoods sites, and included: sawtooth palmetto ( Serenoa repens (B.)), wax myrtle ( Myrica ceriferea L.), runner oak ( Quercus pumila W.), blueberries ( Vaccinium sp p.), gallberry ( Ilex glabra (L.)), bluestem grasses ( Andropogon spp.), panic grasses ( Panicum spp.), and sedges ( Carex spp.).

PAGE 29

29 Study Designs The t wo field installations were planted in 2000 by the Forest Biology Research Cooperative (FBRC) , with the ACF s ite being planted in December 2000 and the Sanderson site in January 2000. At ACF, four blocks with four replicates of treatment and family plots were ins talled, with plot sizes of 0.03 ha and 60 trees per plot. The three way factorial design (2x2x3) cons isted of silvicultural treatment intensity (high and operational intensity), genetic families (slow and fast growers), and SR measurement position s (bed, inter bed, and root exclusion). At Sanderson, four replicate blocks were arranged as a randomized com plete block, split split plot design. The three way factorial design (2 x 2 x 3) consisted of silvicultural treatment intensity (high and operational ), genetic families (slow and fast growers), and SR sample position s (bed, inter bed, and root exclusion) . Each treatment plot (exclu ding the border trees) was 0.02 ha , with eight rows per plot and 48 measurement trees per plot. At Sanderson, all plots (high and operational treatments) were double bedded and chemically treated prior to planting. Arsenal (Imazapyr, 28.7% active ingredient, 1.02 L ha 1 ) and Garlon (Triclopyr, 60% active ingredient, 7.02 L ha 1 ) were applied to eliminate herbaceous and woody competition (Roth et al., 2007). Containerized seedlings were planted at an initial spacing of 1.22 m x 2.75 m. The operational intensity plots included management practices commonly used by forest industry in the southeastern U.S. and consisted of an application of 50 kg.ha 1 N and 60 kg.ha 1 P as diammonium phosphate (DAP) at the time of planting (Roth et al., 2007). In 2006 the operational intensity plots were again fertilized with 170 kg.ha 1 N and 28 kg.ha 1 P (as

PAGE 30

30 DAP and urea). The contrasting high intensity treatment was driven by early and complete vegetation control, along with fertilizer additi ons. Competing vegetation for the high treatment was controlled until canopy closure (age 5 y ea rs) using directed spray applications of Arsenal at 0.28 L . ha 1 and Oust (sulfometuron methyl) at 0.14 L . ha 1 . Directed spray applications of Roundup (glyphosa te; 2% solution) were also used to inhibit the development of herbaceous competition. The fertilizer regime for the high treatment included 6 60 kg.ha 1 of 10 10 10 plus micronutrients at time of planting, followed by annual applications of macro and micr o nutrient fertilizers using prescriptions based on foliar analyses (Table 2 1) . The last fertilizer treatment applied to the high intensity plots was in 2008 (225 kg.ha 1 N and 28 kg.ha 1 P applied as DAP and urea). Cumulative fertilizer N and P addition rates at Sanderson were 76 0 kg.ha 1 N and 1 80 kg.ha 1 P for the high intensity treatment and 220 kg.ha 1 N and 80 kg.ha 1 P for the operational intensity treatment . At ACF, bareroot seedlings (1 0 stock) were planted at 1.8 m x 3 m spacing. Prior to pla nting the operational intensity treatment received a single bedding pass, initial understory competition control, and a time of planting fertilizer application consisting of 50 kg.ha 1 N and 60 kg.ha 1 P , applied as DAP (Table 2 1 ). The high intensity treatment was double bedded, received complete understory vegetation control from establishment until crown closure, and multiple fertilizer additions (10 10 10 plus micronutrients) based on foliar analyses (Table 2 1). Similar to Sander son, herbicides were applied at labeled rates using Chopper (imazapyr) and Garlon (triclopyr) across both levels of treatment intensity. Directed spray applications of Roundup (2%) were used in the high treatment to control herbaceous and wood y competitio n until canopy

PAGE 31

31 closure (age 5 y ea rs). Cumulative fertilization rates for N and P at ACF through 2008 were 448 kg.ha 1 N and 104 kg.ha 1 P for the high intensity treatment and 50 kg.ha 1 N and 60 kg.ha 1 P for the operational intensity treatment. The high intensity treatment at ACF was re fertilized in March 2011 to be equivalent to the high intensity treatment at Sanderson. The resulting cumulative N and P fertilization rates for the Sanderson and ACF high intensity treatments were 76 0 kg.ha 1 N and 1 80 kg .ha 1 P . For this research, two full sib families of loblolly pine were contrasted and compared at each site; they were selected a priori from regional progeny test s. Each site included ing ing family . At the ACF, the fast growing family originated from the eastern Coastal Plain and the drought hardy, slow growing family came from the Western Gulf region. At Sanderson, the fast growing and slow growing famil ies were eastern Coastal Plain selections and were previously examined by Sarkhot et al. (2007) in a soil organic carbon ( SOC ) study . The fast growing family was in common between the ACF and Sanderson experiments . Soil collars, made from PVC pipe (diameter 10.16 cm, height 8 cm), were permanently installed 3 cm i nto the soil within each plot on the bed and inter bed position in December 2009. In order to separate autotrophic and heterotrophic respiration, root exclusion cores were installed on the beds using PVC pipes to sever roots and establish an impenetrable root barrier to a 75 cm depth in September 2010. Soil collars were also installed within each root exclusion core to measure SR . Once installed, the root exclusions were allowed to equilibrate for six weeks in order for root respiration to cease (J. Seiler , unpublished). .

PAGE 32

32 Soil Respiration Repeated measures of SR were made monthly , except during the growing season (May September) when measurements were made approximately every two weeks. Instantaneous total soil CO 2 efflux was measured using a Li COR 6400 infrared gas analyzer (IRGA) (Li Cor Inc., Lincoln, NE) , with the 6400 09 Soil CO 2 Flux Chamber attachment at both sites. For each SR measurement, the soil CO 2 efflux chamber was placed on the permanently installed soil collars (three per plot for bed, inter bed, and root exclusion positions ). Accompanying the soil respiration measurements , temperature of the soil surface ( approximately 5 15 cm ) was measured using a soil probe thermocouple within 5 cm of the soil col lar. In addition, soil moisture was measured following soil respiration measurements at the three position s within 10 cm of the soil collar using time domain reflectometry (Hydrosense Soil Water Measurement System (CS 620), Campbell Scientific, Inc. Logan, UT). At ACF, the measurements were made from April 2009 to April 2012 (3 years collected) and at Sanderson from November 2009 to May 2012 (two and half years collected). Root exclusion measurements began at both sites in October 2010 and were followed for 20 months. Carbon Components for Loblolly Pine A boveground biomass for each treatment and plot was estimated at ages 9 11 years at ACF and ages 11 12 years at Sanderson using annual inventory data (height and diameter at breast height) and general allometric equations developed for loblolly pine ( Gonzalez Benecke et al., In review ). Carbon pools were estimated from biomass components as follows: foliage C was estimated as 0.45 of litterfall biomass , and bole, branch, and coarse root biomass were estimated as 0.48 of biomass (Vogt et al., 1986) .

PAGE 33

33 Root Measurement A total of six soil cores (7.6 cm diameter x horizon thickness) were randomly collected from each plot, three in each inter bed and bed posi tion , in June 2009 and August 2010 at the ACF site and August 2010 at the Sanderson site. For an individual plot, the cores were sampled, mixed, and homogenized to a depth of 30 cm. The cores were then weighed and subsampled to 10% of the total field weig ht. The subsampled soil was picked for live and dead roots and separated into size classes. The r oot diameter classes that were sampled included : >5 mm, 2 5 mm, 1 2 mm, and <1 mm. R oot biomass was summed across depths and position s sampled and expressed o n a C basis (Vogt et al., 1986) . Data Analyses To assess the effects of family, intensity, position and their interactions on C, a linear mixed model based on a s plit plot design with blocks was fitted. The following 4 way factorial model, without covariates, was used: y = P + MI + MF + M P + IF + I P + FL + MIF + MI P + MF P + MIF P + w + s + e where, is the fixed effect of measurement date; I is the fixed effect of management intensity; F is the fixed effect of family; P is the fixed effect of collar position ; MI, MF, ML, IF, IL, FL, MIF, MIL, MFL and MIFL are second, third and fourth order fixed effect interactions between measurement, intensity, family, and position factors; w is the random effect of whole plot within block, with w ~ N(0, w ) ; s is the random effect of subplot within plot, with s ~ N(0, s ) ; and e is the random error, with e ~ N(0, e ).

PAGE 34

34 The inclusion of the random effect of subplot in this model incorpora te d a correlation between repeated measurements over time for the same collar. The significance of each model term was evaluated using an approximate F test with a significance level set at 5%. The above model was obtained for each of the sites (ACF and Sa nderson) individually, and, in addition, two datasets were considered . Of the two datasets, t he first had all data for the bed and inter bed, and the second included the bed and root exclusion data . This distinction was made as the status of the plots chan ged over the course of this study. All models were fitted with the MIXED procedure as implemented in SAS 9.3 (SAS Institute, 2009). Results Aboveground Loblolly Pine Carbon Accumulation Aboveground stand C accumulation for loblolly pine was assessed on an average basis over multiple years at both sites to establish differences between the two families and intensities of management. At ACF, the family x age and intensity x age interactions were significant for total aboveground C accumulation (Table 2 2, 2 3). The fast growing family accumulated nearly 30% more C than the slow growing family when averaged across the years sampled (60.4 MgC.ha 1 versus 41.7 MgC.ha 1 ) (Table 2 3). Similarly, the high intensity management treatment accumulated nearly twice as much C as the operational silvicultural treatment (67.0 MgC.ha 1 versus 35.1 MgC.ha 1 ) (Table 2 3). At Sanderson, both intensity and family were significant factors affecting total aboveground loblolly pine C accumulation (Table 2 4 , 2 5 ). The slow growi ng family accumulated less total C (43.2 MgC.ha 1 ) across years than the fast growing family (49.5 MgC.ha 1 ) (Figure 2 1 ). The high intensity management treatment accumulated

PAGE 35

35 more total C (56.6 MgC.ha 1 ) than the operational intensity management treatment (36.1 MgC.ha 1 ) (Figure 2 2 ). Mean Soil Respiration : Bed and Inter bed Comparison At the ACF site, the effects of date, position , position x date, family x date, position x family x date, and intensity x date on SR were significant (Table 2 6 ). D aily mean SR rates ranged from about 2 .s 1 2 .s 1 in August 2009. T he position x family x date interaction was significant (p=0.0173), with the bed position normally having the greatest SR rate through time (Figure 2 3 ). T he intensity x date interaction (p<0.0001 ) suggest ed that the operational silvicultural intensity had greater SR rates than the high intensity treatment (Figure 2 4 ). T here was a significant difference between the bed (5.0 2 .s 1 ) and inter bed (4.3 2 .s 1 ) position s (p=0.0002), with the bed position having greater SR rates (Figure 2 5 ). T he family x date interaction (p=0.0347) (Figure 2 6 ) indicated that the SR for the slow growing family was greater than the fast growing family from April 2009 to August 2010 (e.g., 6.64 2 .s 1 versus 5.82 2 .s 1 , respectively) . Thereafter, the families switch ed positions, and the fast growing family ha d greater or similar SR rates than the slow growing family until April 2012 (e.g., 5.55 2 .s 1 versus 4.66 2 .s 1 , respectively) (Figure 2 6 ). At the Sanderson experiment, SR rates were affected by date, position , date x position , family, date x family, intensity, date x intensity, position x intensity, date x position x intensity, and family x position x intensity (Table 2 7 ). Mean daily SR rates 2 .s 1 2 .s 1 in August 2011. T hrough time, SR rates for the bed and inter bed positions follow ed the same pattern ,

PAGE 36

36 but th e bed position had greater SR rates overall, which was similar t o the ACF experiment (Figure 2 5 ). The fast growing family had a greater mean SR rate than the slow growing family (4.2 2 .s 1 2 .s 1 for the fast and slow growing fam ily , respectively ). Similarly, t hrough time, the slow and fast growing families followed the same pattern of SR, but the fast grower had greater SR rates for the majority of the dates sampled (Figure 2 6 ). The operational management intensity treatment res ulted in greater mean SR rates overall than the high intensity treatment (4.3 2 .s 1 versus 2 .s 1 for the low and high treatment s, respectively ) . At Sanderson, t hrough time, both treatment intensities followed the same seasonal SR pattern (Figure 2 4 ). However, the operational treatment resulted in greater SR rates across a ll but several dates (Figure 2 4 ). The interaction of position x intensity demonstrated that the operational management intensity resulted in greater SR rates when bed position s were compared (Figure 2 7 ). However, the inter bed positions for both treatments were not significantly different (Figure 2 7 ). The management intensity x position interaction suggested that the operational management treatment in the bed position had the greatest SR rates through time, while the high intensity treatment inter bed position had the lowest rate s of SR through time (Figure 2 8 ). At Sanderson, t he interaction of family x management intensity x position indicated that the fast growing family under the operational management regime in the 2 .s 1 ) (Figure 2 9 ). The slow growing family under the high intensity management regime in the inter bed position had the lowest mean SR 2 .s 1 ) (Figure 2 9 ). Under the high intensity regime ,

PAGE 37

37 the positions and families were not significantly different, whereas with the operational intensity treatment differences between the bed po sitions exist ed betwe en the two families (Figure 2 9 ). Mean Soil Respiration: Bed and Root Exclusion Comparison The bed and root exclusion position s were compared to discern differences in RH and total SR. The difference between SR and RH yielded estimates of RA. In the ACF comparison of bed and root exclusion positions, the main eff ects of date, position , and family were highly significant (Table 2 8 ). In addition, the interactive effects of position x date, and intensity x date were significant (Table 2 8 ). The bed position had a significantly greater SR rate (4.98 2 .s 1 ) than the root exclusion position (3.97 2 .s 1 ; p=0.0012) , making RH and RA approximately 80% and 20% of total SR, respectively . The fast growing family (4.74 2 .s 1 ) had a significantly greater SR rate than the slow growing family (4.21 2 .s 1 ; p=0.0477) . The bed position ha d a significantly greater (p <0.0001 ) SR rate through time compared to the root exclusion position (Figure 2 10 ). The operational intensity treatment had greater SR rates through time compared to the high intensity management treatment (p=0.0002) (Figure 2 11 ). In the Sanderson comparison of bed and root exclusion position s, the mai n effects of date and position were significant (Table 2 9 ). Also, the interactive effects of date x position , family x position , date x family x position , date x intensity, and position x intensity were significant (Table 2 9 ). Through time, the bed and r oot exclusion follow ed the same pattern of SR rates , but the bed position ha d greater mean daily SR rates overall (4.57 2 .s 1 versus 4.03 2 .s 1 , respectively) and RH accounted for approximately 88% of total SR (Figure 2 10 ).

PAGE 38

38 At Sanderson, t he interactive effect of position x family showed that t he fast growing family had a greater mean SR rate in the bed position than the slow growing family (Figure 2 12 ). However, the fast growing family had a lower SR rate in the root exclusion position co mpared to the slow growing family (3.85 2 .s 1 versus 4.22 2 .s 1 ) (Figure 2 12 ). The difference between bed and root exclusion position s for the fast growing family was greater than the difference in the slow growing family (0.90 2 .s 1 versus 0.19 2 .s 1 ) , although there was no major difference between the root exclusions of the families (Figure 2 12 ). T he interactive effect of position x family x date at Sanderson demonstrated that the bed position of the fast growing family had the greatest SR rate s throughout time (Figure 2 13 ). The root exclusion position s of the fast and slow growing families had the lowest SR rates through time (Figure 2 13 ). For the duration of the Sanderson experiment, the high intensity treatment resulte d in a lower mean SR rate, except for several winter measurement periods in December 2011 to February 2012 (Figure 2 1 1 ). The h igh intensity management regime did not result in a significant difference in SR between the bed and root exclusion positions. However, the operational management regime had greater SR efflux rates in the bed position (Figure 2 1 4 ). There was no significant difference in SR between the root exclusion positions for either management intensity (Figu re 2 1 4 ). In addition, t he interaction of date x family x intensity at Sanderson demonstrated that t he fast growing family had the greatest difference in SR rates, with the operational intensity and fast growing family combination resulting in the greates t SR rates (Figure 2 1 5 ). Under the high intensity management regime , both families behave d similarly,

PAGE 39

39 with significant differences in SR only occurring between Octobe r 2010 and May 2011 (Figure 2 1 5 ). The same pattern occur red for families managed under t he operational management regime (Figure 2 1 5 ). Fertilization of Bed and Inter bed at ACF Time periods of pre fertilization (PreFert) and post fertilization (PostFert) were compared for the bed and inter bed position s at the ACF site to assess short term changes, if any, in SR due to the fertilizer treatment. Position , period, position x period, and intensity x period were all significant effects (Table 2 10 ). The position effect wa s significant (p=0.0017), with the b ed SR 2 .s 1 ) being higher than the inter bed position 2 .s 1 ) (Table 2 11 ) . Across all treatments, the post fertilization period had a greater mean SR rate ( 2 .s 1 ) than the pre fertilization period ( 2 .s 1 ). The interactive effect of position and period was si gnificant (p=0.0182), with the post fertilization period having a greater SR rate than the pre fertilization period, and the bed having greater efflux than the inter bed position across periods (Table 2 11 ). The p re fertilization SR rates for the high and operational intensity treatments were not significantly different (Figure 2 1 6 ) . However, post fertili zation, the operational mean SR 2 .s 1 ) was signif icantly greater than the high SR 2 .s 1 ) (Figure 2 1 6 ) . Root Carbon At ACF, date was a significant effect (p<0.05) across all root diameter classes, with the June 2009 sampling date accumulating more root C than the August 2010 sampling date (Table 2 12 ; 2 13 ). The date x intensity effect was significant for the 1 2mm medium fine root diameter class, with the operational intensity accumulating more

PAGE 40

40 C across both dates (Table 2 12 ; 2 13 ). For the 1 2mm diameter class, intensity and family were also significant effects (Table 2 12 ; 2 13 ). Under the operational management intensity, more root C was accumulated than in the high intensity management treatment for the 1 2mm diameter class (53.7 gC.m 2 versus 36.2 gC.m 2 ). Similarly for the 1 2mm diameter class, the slow growing family accumulated more root C than the fast growing family (50.4 gC.m 2 versus 39.5 gC.m 2 ). For the 2 5mm diameter class, intensity was significant , with the operationa l intensity management regime accumulating more root C than the high intensity management treatment (107.7 gC.m 2 versus 69.9 gC.m 2 ). At Sanderson, no significant treatment effects were detected for root C accumulation in the <1mm fine root diameter clas s (Table 2 14 ). For the 1 2mm medium fine root diameter class, intensity and family were significant (p<0.05) (Table 2 14 ). The operational management intensity accumulated more root C than the high intensity management regime (22.4 gC.m 2 versus 10.4 gC.m 2 ) (Table 2 1 5 ). The slow growing family accumulated more root C than the fast growing family (20.3 gC.m 2 versus 12.5 gC.m 2 ) (Table 2 11). Temperature Response Temperature response curves were developed based on the Arrhenius equation for both sites (Fi gure 2 17, 2 18). At ACF, Q10 values ranged from 1.3 to 1.4 ( Table 2 16 ). At Sanderson, Q10 values ranged from 1.3 to 1.6 ( Table 2 17 ). Discussion Understanding the influence of intensive forest management and deployment of different families on SR is relevant in light of global climate change concerns (Rustad et al ., 2000; Woodbury et al ., 2007). Through the examination of varying levels of forest

PAGE 41

41 management intensity using fertilization and weed control treatment s , and different families of loblolly pine, a unique opportunity was presented to study genotype x environment interactions on SR and belowground processes in two long term (two and a half to three years), replicated field experiments. In this study, mean daily SR rates (1.2 9.8 2 .s 1 ) were similar to reported averages, but ranged higher than those reported by Maier and Kress (2000) in an eleven year old loblolly pine plantation i n North Carolina (0.5 6.0 2 .s 1 ) and Samuelson et al. (2004) in a six year old loblolly pine plantation in Georgia (1.3 5.6 2 .s 1 ). Similarly, Samuelson et al. (2009) found SR rates of 2.0 6.0 2 .s 1 in a seven year old loblolly p ine plantation in South Carolina. Samuelson (2012) also reported SR rates of 1.6 to 6.4 2 .s 1 in a 50 year old longleaf pine forest with varying levels of basal area. Overall, the high intensity fertilization and weed control treatment regime resu lted in suppressed SR rates at both ACF and Sanderson when compared to the operational treatment for most dates. This trend was carried over in the short term, with the high intensity treatment resulting in less SR than the operational intensity treatment. However, in the same analysis of pre and post fertilization at ACF, greater SR rates were found post fertilization for both treatments than pre fertilization. Since the operational intensity treatment was not fertilized with the high intensity treatment, this increase in SR rates was likely due to the growth of trees throughout the second year of comparison since the Palmer Drought Severity Index (PDSI) values were more negative during the post fertilization period.

PAGE 42

42 In this study, three position s were sampled: bed, inter bed, and root exclusion . The bed and root exclusion were compared to estimate the RH and RA components. The bed position had the greatest SR rates and the root exclusion position had the lowest. Samuelson et al. (2004) found a di fference in SR rates based on the sampling distance from the tree, with higher rates associated with closer distance to the trees in a loblolly pine plantation. Our results are consistent with that finding, with bed SR rates being greater than inter bed ra tes since more roots are centered on the bed, or close to the tree (Adegbidi et al. , 2004) . Heterotrophic respiration can account for a large percentage of total SR (Maier and Kress, 2000; Kim et al., 2012). At ACF, there was no interaction between samp ling position (bed and root exclusion) and management intensity . However, at Sanderson, the position x intensity interaction was significant and showed that under the high intensity treatment, there was no difference between the bed and root exclusion posi tions. Under operational intensity, the bed position had a greater efflux than the root exclusion, which was indicative of greater autotrophic respiration. At Sanderson, greater root mass in the 1 2mm diameter class was found in the operational treatment than in the high treatment (46.7 gC.m 2 versus 21.6 gC.m 2 ) . G reater root mass is one explanation for the greater SR rates in the operational treatment than in the high intensity treatment. The application of herbicides and the reduction of understory biom ass and root inputs could cause a decrease in soil aggregation (Sarkhot et al. , 2007) and SOM in loblolly pine forests (Shan et al. , 2001 ; Echeverria et al. , 2004), both of which could affect SR rates. At Sanderson, there was no difference between the two

PAGE 43

43 root exclusions under the different intensities of management, demonstrating that in this case, fertilization and weed control did not apparently suppress RH rates. In an 11 year old loblolly pine plantation, Maier and Kress (2000) found that roots to a depth of 15 cm accounted for 20 50% of total SR. Therefore, RH accounted for an estimated 50 80% of SR. In a 9 and 29 year old slash pine plantation, Ewel et al. (1987) estimated RH at 49 and 38% of SR Kim et al. (2012) found that RH accounted for about 62% in unfertilized and 78% in fertilized plots in a Korean red pine ( Pinus densiflora S. et Z.) forest to a depth of 30 cm. In this study, under the high intensity management regime, there was no significant d ifference between the bed and RE positions. Therefore, the contribution of RH to total SR was large r than previously reported values ; RH was approximately 80% of SR at ACF and approximately 88% at Sanderson. It was probable that soil moisture within the ro ot exclusion affected the RH measurements, since there was an increase in soil moisture due to reductions in transpiration caused by root severing (Bond Lamberty et al., 2011). G reater root exclusion soil moisture was evident at Sanderson (Figure 2 1 9 ) . Gr eater efflux within the root exclusion may also be due to changes in the microbial populations within the collar (Moyano et al., 2013). At Sanderson, the difference in SR between the high and operational management regime was not due to RH, but to RA since there was no difference between RH for the two management i ntensities ; just a difference in proportional contribution of RH to SR (high 97% and operational 81%). The effects of fertilization on SR has varied among studies (Maier and Kress, 2000; Butnor et al., 2003; Tyree et al., 2006). Some have shown that the short term

PAGE 44

44 effect of fertilization contributed to a suppression in RH (Gough and Seiler, 2004; Tyree et a l., 2008). However, our study demonstrated that between the two management intensities, RH was not suppressed. Similarly, other studies have shown increases or no effect of fertilizer additions on RH (Samuelson et al., 2009; Kim et al., 2012). These variable responses of RH to fertilization may be site specific. For example, at A CF, soil temperature varied by intensity through time, but with no pattern other than seasonal fluctuations for the bed and inter bed position s (Figure 2 20 ) . However, at Sanderson, intensity was significant, with the operational intensity having a slightl y higher soil temperature than the high intensity treatment for the bed and inter bed positions (22.8 ° C versus 22.4 ° C); the same pattern was found for the bed and root exclusion comparison (operational, 23.3 ° C versus high, 23.0 ° C) (Figure 2 21 ) . The differ ence in temperature between intensities could be related to canopy leaf area, as our foliage C estimates indicate d that the high intensity plots had greater C mass than the operational intensity plots (Table 2 5 ) . In order to separate RA and RH from tota l SR, root exclusion plot s are normally used since trenching and girdling trees results in large disturbances and the destruction of trees (Hanson et al., 2000; H g berg et al. 2001; Kutsch, 2009). Root exclusions made of PVC pipes, like those deployed in t his study, were used by Vogel and Valentine (2005) to separate RH from SR by isolating tree roots from new photosynthate. Vogel and Valentine (2005) also found that respiration within the PVC root exclusion fell to that of trenched root exclusions in one t o three weeks. The r oot exclusions in this study were installed in late September and differences between the bed and root exclusion SR values were apparent in November (1.5 months later) at

PAGE 45

45 ACF and in March (5.5 months later) at Sanderson (Figure 2 10 ). Early differences were detected at ACF, but large differences were not clearly evident until February (Figure 2 10 ). The lack of strong differences early on in the installation of the root exclusion s may have be en an artifact of the season, since SR rates decline in the winter. Similarly, greater SR or no difference in SR between the bed and root exclusion may be the result of elevated soil moisture within the root exclusion collar, which could affect the microbial populations (Moyano et al. 2013). For exam ple, at ACF, soil moisture was measured only from October 2011 to April 2012 and the soil moisture was modestly greater in the root exclusion than the bed (5% versus 2%) position . At Sanderson, the root exclusion also had greater soil moisture (4.4%) compa red to the bed (2.6%) (Figure 2 1 9 ). Another possibility is that gas diffusion in the bed collar could have decreased the measured SR rates (Davidson and Trumbore, 1995). Observed differences in mean SR rates among family treatments could be related to to tal aboveground C accumulation . In this experiment, at both position s, intensity and family significantly affected total aboveground C accumulation, but there were no genotype x silviculture interactions detected. Overall, the fast growing family accumulat ed more aboveground C than the slow growing family and the high intensity management treatment had greater C accumulation than the operational management regime. Other studies have found that SR varied with stand age, with greater mean SR rates found in ol der stands with greater aboveground biomass (Ewel et al., 1987; Amiro et al., 2010). Although there was no interaction between intensity and family in terms of aboveground C , it was clear that individually the effects of increased management

PAGE 46

46 intensity and superior, full sib families resulted in significant ly greater gains in C accumulation . The effects of cultural treatments and genetics were also apparent for SR , with higher SR rates associated with the full sib fast growing family over the slow growing f amily. Another study found that SR rates were correlated with stand basal area (Luan et al., 2011). Subke et al. (2006) and Kuzyakov and Gavrichkova (2010) suggested that with greater biomass, an increase in total belowground C allocation could be the sour ce of a proportional decrease of RH to SR. However, our root data indicated that the slow growing family produced more medium fine roots (1 2mm) than the fast growing family. Stovall et al. (2013) found that two full sib loblolly pine clones differed in SR and responded to fertilization differently. Similarly , family effects on SOC pools have been demonstrated (Sarkhot et al., 2008). In this study, greater aboveground tree C due to genotype corresponded to a larger rate of SR, but the difference was not due to greater fine root biomass. T he higher rates of SR observed at both sites in this study were likely due to the length of the growing season, high annual precipitation, and warm annual temperatures. The Palmer Drought Severity Index (PDSI) was used as a metric to determine the status of the soil moisture and temperature. For nearly the entire duration of the experiment (80% of the duration of the experiments at both sites), the north central Florida region was under drought conditions according to the PDSI (values below 0) (Figure 2 2 2 ). Furthermore, several severe drops in SR, e.g. in August 2011, align with the lowest values of PDSI ( 4) and, thus, the most extreme drought periods. Similarly,

PAGE 47

47 severa l upward spikes in SR, e.g. April May 2012, align with large increases in PDSI, indicating a decrease in drought conditions (change from 4 to 0 PDSI). Correlations of efflux with soil temperature indicated that soil temperature was a major driver of SR , accounting for 23% of the variation for the ACF bed and inter bed position s and 25% of the variation for the bed and root exclusion comparison. Similarly, at Sanderson, soil temperature accounted for 49% of the variation in SR in the bed and inter bed po sition , and 30% in the bed and root exclusion comparison. Using the Arrhenius exponential equation, the temperature dependence of soil respiration was examined. Q10 values (1.3 1.6) were lower than that of Maier and Kress (2000) , which ranged from 2.2 to 2 .4. Although soil moisture was a significant correlate at Sanderson for both comparisons, soil moisture accounted for a lesser extent in SR variability than soil temperature, with 10% for the bed and inter bed comparison and 2% for the bed and root exclu sion comparison. At ACF, soil moisture only accounted for about 13% of the variation in SR for the bed and inter bed position s and no variation in SR for the bed and root exclusion comparison. Hence, t here were widespread drought conditions throughout muc h of the study that likely resulted in short term reduced tree photosynthetic activity and reduced belowground C allocation and root respiration. This could also occur for heterotrophic microbial activity and respiration due to limited soil moisture (Ryan and Law, 2005; Cisneros Dozal et al. , 2007). Bracho et al. (2012) found in a multi year eddy covariance study of a slash pine plantation that long periods of drought stress reduce d aboveground C assimilation and total ecosystem respiration.

PAGE 48

48 Summary The r esults from this study indicate d that management intensity and genetics influence d SR rates. In these loblolly pine plantations , RH was the major component of SR. RH as an indicator of belowground C storage could increase belowground storage, with a decrease in RH, or alternatively could decrease belowground storage with an increase in RH. In this study, however, RH did not vary b etween management intensities, indicating no change in belowground C storage due to microbial activity. Along the se same lines, the RH contribution to total SR was generally greater than previously reported for Pinus species (Ewel et al., 1987; Kim et al., 2012; Maier and Kress, 2000) . The fast and slow growing families differed in soil respiration and root C accumulation. The fast grower had greater SR than the slow growing family; however, the slow growing family had greater medium fine root C than the fast growing family. No differences were found between the root exclusions for either of these families, suggesting that this difference in SR was likely due to RA. Alternatively, this difference could be due to the greater aboveground C accumulation for the fast growing family. Future studies should examine the microbial communities found within and outside the root exclusion plots . This would assist in determining if the relatively large proportion of SR that was RH was due to differences in the microbi al community or if there was greater microbial biomass within the root exclusion collar than in the bed. In addition, a fertilization study should examine fine roots before and after fertilization , as well as root exudates and other labile forms of C to determine how those sources affect ed C storage and the partitioning of RH and RA .

PAGE 49

49 In this study, there were clear effects of temporal variation in SR with genotype and intensity, as well as differences in root C mass in the medium fine root category . However, it remains unclear if C storage was different between the two families and intensities since fertilization suppresse d overall SR but not RH , and since the fast growing family had greater SR rates but less root C mass. Future studies could elucidate the magnitude of these trade offs.

PAGE 50

50 Table 2 1 . Cumulative fertilization rates for the loblolly pine experimental sites, Austin Cary Forest (ACF) and Sanderson in north Florida. Cumulative application rates (kg. ha 1 ) Position Time Period Intensity N P ACF 2000 2011 High 450 100 2011 2012 High 760 180 2000 2012 Operational 50 60 Sanderson 2000 2012 High 760 180 Operational 220 80 Table 2 2 . Analysis of variance for aboveground carbon mass by tree component for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) across ages (A, 10 12 years) were included in the analysis. Effect TOTAL BOLE BRANCH FO LIAGE Block x A 0.1621 0.1235 0.2751 0.3566 A < 0 .0001 <0.0001 <0.0001 0.2521 I <0.0001 <0.0001 <0.0001 <0.0001 I x A 0.0008 <0.0001 0.2797 0.1847 F 0.0004 0.0004 0.0004 0.0007 F x A <0.0001 <0.0001 0.0015 0.1914 I x F 0.0582 0.0587 0.061 0.0576 I x F x A 0.701 0.5584 0.934 0.6523

PAGE 51

51 Table 2 3. Means of aboveground carbon mass by tree component for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) across ages (A, 10 12 years) were included in the analysis. Standard errors are in parentheses. Effect Intensity Family Age Total Bole Branch Foliage A 10 43.0 (1.7) 32.3 (1.3) 7.4 (0.3) 3.2 (0.1) A 11 46.7 (1.7) 35.7 (1.3) 7.7 (0.3) 3.2 (0.1) A 12 50.9 (1.7) 39.6 (1.3) 8.1 (0.3) 3.2 (0.1) I High 61.5 (2.4) 47.1 (1.8) 10.3 (0.4) 4.1 (0.2) I Oper 32.2 (2.1) 24.6 (1.6) 5.3 (0.4) 2.4 (0.1) I x A High 10 56.7 (2.4) 42.7 (1.8) 9.9 (0.4) 4.1 (0.2) I x A High 11 61.3 (2.4) 47.0 (1.8) 10.2 (0.4) 4.1 (0.2) I x A High 12 66.4 (2.4) 51.7 (1.8) 10.7 (0.4) 4.0 (0.2) I x A Oper 10 29.2 (2.1) 21.9 (1.6) 5.0 (0.4) 2.4 (0.1) I x A Oper 11 32.1 (2.1) 24.4 (1.6) 5.3 (0.4) 2.4 (0.1) I x A Oper 12 35.4 (2.1) 27.4 (1.6) 5.6 (0.4) 2.4 (0.1) F Fast 55.4 (2.4) 42.5 (1.8) 9.2 (0.4) 3.7 (0.2) F Slow 38.3 (2.1) 29.2 (1.6) 6.3 (0.4) 2.7 (0.1) F x A Fast 10 50.3 (2.4) 37.9 (1.8) 8.7 (0.4) 3.7 (0.2) F x A Fast 11 55.3 (2.4) 42.3 (1.8) 9.2 (0.4) 3.7 (0.2) F x A Fast 12 60.7 (2.4) 47.2 (1.8) 9.7 (0.4) 3.7 (0.2) F x A Slow 10 35.6 (2.1) 26.7 (1.6) 6.1 (0.4) 2.8 (0.1) F x A Slow 11 38.1 (2.1) 29.1 (1.6) 6.3 (0.4) 2.7 (0.1) F x A Slow 12 41.1 (2.1) 31.9 (1.6) 6.5 (0.4) 2.7 (0.1) I x F High Fast 66.8 (3.1) 51.2 (2.4) 11.2 (0.5) 4.4 (0.2) I x F High Slow 56.2 (3.1) 43.0 (2.4) 9.4 (0.5) 3.8 (0.2) I x F Oper Fast 44.1 (3.1) 33.7 (2.4) 7.3 (0.5) 3.1 (0.2) I x F Oper Slow 20.4 (2.9) 15.4 (2.3) 3.3 (0.5) 1.7 (0.2)

PAGE 52

52 Table 2 4. Analysis of variance for aboveground carbon mass by tree component for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) across ages (A, 10 12 years) were included in the analysis. Effect TOTAL BOLE BRANCH FOLIAGE Block x A 0.2876 0.2961 0.2891 0.3141 A <0.0001 <0.0001 0.0004 0.0112 I 0.0074 0.0077 0.0075 0.0099 I x A 0.5316 0.6567 0.4112 0.6367 F 0.0006 0.0008 0.0004 0.0003 F x A 0.9752 0.9535 1 .0000 0.8124 I x F 0.2636 0.5084 0.111 0 0.0563 I x F x A 0.7023 0.6846 0.6969 0.6367

PAGE 53

53 Table 2 5. Means of aboveground carbon mass by tree component for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) across ages (A, 10 12 years) were included in the analysis. Standard errors are in parentheses. Effect Intensity Family Age Total Bole Branch Foliage A 11 44.2 (1.6) 32.5 (0.9) 10.1 (0.6) 1.6 (0.1) A 12 48.4 (1.6) 35.4 (0.9) 11.3 (0.6) 1.8 (0.1) I High 56.6 (2.2) 39.9 (1.3) 14.3 (0.8) 2.4 (0.2) I Oper 36.1 (2.2) 28.1 (1.3) 7.1 (0.8) 1.0 (0.2) F Slow 43.1 (1.6) 32.2 (1.0) 9.5 (0.6) 1.5 (0.1) F Fast 49.5 (1.6) 35.8 (1.0) 11.8 (0.6) 1.9 (0.1) I x F High Slow 2.1 (0.2) I x F High Fast 2.8 (0.2) I x F Oper Slow 0.8 (0.2) I x F Oper Fast 1.1 (0.2) Table 2 6 . Analysis of variance for soil respiration rates for bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the analysis. Effect Num DF Den DF F Value P Value D 51 942 54.69 <0. 0001 P 1 12.6 17.86 0.0011 P x D 51 942 1.89 0.0002 F 1 12.2 0.15 0.7074 F x D 45 948 1.43 0.0347 P x F 1 12.5 2.15 0.1669 P x F x D 45 949 1.51 0.0173 I 1 12.2 2.17 0.1664 I x D 48 937 2.38 <0.0001 P x I 1 12.6 2.86 0.1152 P x I x D 48 937 1.01 0.4528 F x I 1 12.1 0.38 0.5501 F x I x D 43 944 1.19 0.1869 P x F x I 1 12.4 0.7 0.4176 P x F x I x D 43 944 1.03 0.4151

PAGE 54

54 Table 2 7 . Analysis of variance for soil respiration rates for bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the anal ysis. Effect Num DF Den DF F Value p Value B lock 3 8.93 1.26 0.3455 D 33 1221 141.62 <0.0001 P 1 26.4 56.45 <0.0001 D x P 33 1221 3.02 <0.0001 F 1 9.86 5.88 0.0362 D x F 33 1221 2.73 <0.0001 F x P 1 26.4 0.01 0.9286 D x F x P 33 1221 0.81 0.7723 I 1 9.86 10.05 0.0102 D x I 33 1221 2.15 0.0002 P x I 1 26.4 9.75 0.0043 D x P x I 33 1221 1.58 0.0206 F x I 1 9.86 0.46 0.5124 D x F x I 33 1221 1.29 0.1244 F x P x I 1 26.4 7.98 0.0089 D x F x P x I 33 1221 1.26 0.1525

PAGE 55

55 Table 2 8 . Analysis of variance for soil respiration rates for bed and root exclusion positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the analysis. Effect Num DF Den DF F Value P value D 21 411 84.42 <0.0001 P 1 12 17.69 0.0012 P x D 21 411 4.41 <0.0001 F 1 12.1 4.85 0.0477 F x D 21 411 0.88 0.6201 P x F 1 12 0.21 0.6575 P x F x D 21 411 1.37 0.1273 I 1 12.1 4.36 0.0587 I x D 21 411 2.56 0.0002 P x I 1 12 1.41 0.2574 P x I x D 21 411 1.43 0.0972 F x I 1 12.1 0.86 0.3731 F x I x D 21 411 1.1 0.346 P x F x I 1 12 0.27 0.6132 P x F x I x D 21 411 0.59 0.9269

PAGE 56

56 Table 2 9 . Analysis of variance for soil respiration rates for bed and root exclusion positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the analysis. Effect Num DF Den DF F Value P v alue B lock 3 9.01 0.88 0.4876 D 20 702 127.69 <0.0001 P 1 12.3 14.07 0.0026 D x P 20 702 2.12 0.003 F 1 9.1 0 0.9591 D x F 20 702 0.78 0.7358 F x P 1 12.3 5.96 0.0306 D x F x P 20 702 2.18 0.0022 I 1 9.1 3.05 0.1145 D x I 20 702 3.66 <0.0001 P x I 1 12.3 7.77 0.016 D x P x I 20 702 1.01 0.4497 F x I 1 9.1 0.27 0.6189 D x F x I 20 702 1.7 0.0285 F x P x I 1 12.3 3.86 0.0723 D x F x P x I 20 702 0.99 0.4768

PAGE 57

57 Table 2 10 . Analysis of variance for soil respiration rates for bed and inter bed positions during pre fertilization and post fertilization periods at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and period (pre fert and post fert) included in the analysis. Effect Num DF Den DF F Value P value P 1 15.8 14.15 0.0017 F 1 13.3 1.15 0.3029 P x F 1 15.8 0.94 0.3478 I 1 13.3 1.74 0.2095 P x I 1 15.8 1.28 0.2744 F x I 1 13.3 0 0.9539 P x F x I 1 15.8 0.95 0.345 P eriod 3 1374 182.69 <0.0001 L x P eriod 3 1374 3.36 0.0182 F x P eriod 3 1374 0.72 0.5412 P x F x P eriod 3 1374 1.97 0.1173 I x P eriod 3 1374 4.51 0.0037 P x I x P eriod 3 1374 2.43 0.0638 F x I x P eriod 3 1374 1.94 0.1205 P x F x I x P eriod 3 1374 1.18 0.3147

PAGE 58

58 Table 2 11 . Means of soil respiration rates for bed and inter bed positions during pre and post fertilization periods at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P ) and period (pre fert and post fert) included in the analysis. Effect Position Intensity Period Estimate P bed 4.68 (0.16) P inter bed 4.07 (0.16) P x P eriod bed PostFert 5.76 (0.18) P x P eriod bed PreFert 4.10 (0.17) P x P eriod inter bed PostFert 5.30 (0.18) P x P eriod inter bed PreFert 3.52 (0.17) I x P eriod I x P eriod I x P eriod I x P eriod High High Oper Oper PostFert PreFert PostFert PreFert 5.09 (0.21) 3.73 (0.20) 5.97 (0.21) 3.89 (0.20)

PAGE 59

59 Table 2 12 . Analysis of variance for root carbon mass by root diameter class for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with date of measurement (D) included in the analysis . Root diameter class Effect Num DF Den DF F Value P Value <1mm D 1 18 40.64 <0.0001 <1mm I 1 6 2.19 0.1892 <1mm D x I 1 18 2.88 0.107 <1mm F 1 18 0.22 0.6417 <1mm D x F 1 18 0.12 0.7287 <1mm I x F 1 18 1.23 0.2817 <1mm D x I x F 1 18 0.82 0.378 1 2mm D 1 18 76.95 <0.0001 1 2mm I 1 6 11.28 0.0153 1 2mm D x I 1 18 6 0.0247 1 2mm F 1 18 5.47 0.031 1 2mm D x F 1 18 4.07 0.0589 1 2mm I x F 1 18 0.31 0.5844 1 2mm D x I x F 1 18 1.46 0.2432 2 5mm D 1 24 26.25 <0.0001 2 5mm I 1 24 4.31 0.0488 2 5mm D x I 1 24 2.67 0.1154 2 5mm F 1 24 0.35 0.5602 2 5mm D x F 1 24 0 0.9575 2 5mm I x F 1 24 0.27 0.6071 2 5mm D x I x F 1 24 0.07 0.7963 Table 2 13 . Means of root carbon mass by diameter class for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with date of

PAGE 60

60 measurement (D) were included in the ana lysis. Root carbon mass means units are gC.m 2 and their standard errors are given in parentheses. Root diameter class Effect Date Intensity Family Mean <1mm D 2009 216.2 (23.5) <1mm D 2010 33.4 (23.5) 1 2mm D 2009 65.4 (3.5) 1 2mm D 2010 24.6 (3.5) 1 2mm I High 36.2 (3.7) 1 2mm I Oper 53.7 (3.7) 1 2mm F Fast 39.5 (3.5) 1 2mm F Slow 50.4 (3.5) 2 5mm D 2009 135.8 (13.0) 2 5mm D 2010 41.5 (13.0) 2 5mm I High 69.6 (13.0) 2 5mm I Oper 107.7 (13.0)

PAGE 61

61 Table 2 1 4. Analysis of variance for root carbon mass by root diameter class for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) were included in the analysis. Root diameter class Effect Num DF Den DF F Value P Value <1mm I 1 3.11 3.44 0.1577 <1mm F 1 6.11 0.14 0.7215 <1mm I x F 1 6.11 0.33 0.585 1 2mm I 1 5.74 12.12 0.0141 1 2mm F 1 5.31 10.49 0.0211 1 2mm I x F 1 5.31 4.18 0.0931 2 5mm I 1 8.3 3.03 0.1184 2 5mm F 1 8.3 0 0.9533 2 5mm I x F 1 8.3 4.98 0.0549 Table 2 1 5 . Means of root carbon mass by diameter class for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) were included in the analysis. Root carbon mass means units are gC.m 2 and their standard errors are given in parentheses. Root diameter class Effect Intensity Family Mean 1 2mm I High 10.4 (2.5) 1 2mm I Oper 22.4 (2.4) 1 2mm F Slow 20.3 (2.0) 1 2mm F Fast 12.5 (2.2)

PAGE 62

62 Table 2 16. Arrhenius temperature response parameters and Q10 values for the Austin Cary Forest in north Florida. The Arrhenius equation in the form of Efflux = A* e^ (B*Soil Temperature) was used. The groups are given as family (fast or slow growing), location (Bed, inter bed, or root exclusion (RE)), and intensity (high or operational). Soil CO 2 2 .s 1 . Family Location Intensity A B Efflux at 25 o C Efflux at 15 o C Q10 Fast Bed High 2.6619 0.0264 5.2 4.0 1.3 Fast Bed Oper 2.2543 0.0349 5.4 3.8 1.4 Fast Inter High 2.2731 0.0252 4.3 3.3 1.3 Fast Inter Oper 2.6659 0.027 5.2 4.0 1.3 Fast RE High 2.5794 0.0236 4.7 3.7 1.3 Fast RE Oper 1.9554 0.034 4.6 3.3 1.4 Slow Bed High 2.6392 0.0285 5.4 4.0 1.3 Slow Bed Oper 2.5877 0.0308 5.6 4.1 1.4 Slow Inter High 1.8311 0.0348 4.4 3.1 1.4 Slow Inter Oper 2.0477 0.0341 4.8 3.4 1.4 Slow RE High 1.5439 0.0349 3.7 2.6 1.4 Slow RE Oper 1.9342 0.0316 4.3 3.1 1.4 Table 2 17. Arrhenius temperature response parameters and Q10 values for the Sanderson site in north Florida. The Arrhenius equation in the form of Efflux = A* e^ (B*Soil Temperature) was used. The groups are given as family (fast or slow growing), location (Bed, inter bed, or root exclusion (RE)), and intensity (high or operational). Soil CO 2 2 .s 1 . Family Location Intensity A B Efflux at 25 o C Efflux at 15 o C Q10 Fast Bed High 1.32 0.0482 4.4 2.7 1.6 Fast Bed Oper 1.819 0.0446 5.5 3.6 1.6 Fast Inter High 1.231 0.0493 4.2 2.6 1.6 Fast Inter Oper 1.3365 0.0443 4.0 2.6 1.6 Fast RE High 1.4347 0.0419 4.1 2.7 1.5 Fast RE Oper 1.9114 0.0293 4.0 3.0 1.3 Slow Bed High 1.4811 0.0433 4.4 2.8 1.5 Slow Bed Oper 1.6626 0.0418 4.7 3.1 1.5 Slow Inter High 1.2505 0.041 3.5 2.3 1.5 Slow Inter Oper 1.4678 0.0374 3.7 2.6 1.5 Slow RE High 1.8994 0.0334 4.4 3.1 1.4 Slow RE Oper 1.6775 0.0386 4.4 3.0 1.5

PAGE 63

63 Figure 2 1. Mean for total aboveground loblolly pine carbon accumulation at the Sanderson site in north Florida across two years of measurement for slow and fast growers. Two loblolly pine families (slow and fast growing) growing under operational and high intensity managemen t included in the analysis. Ages examined included 11 and 12 years. Units of total aboveground carbon mass are MgC.ha 1 and error bars show the standard error of the mean.

PAGE 64

64 Figure 2 2 . Mean of total aboveground loblolly pine carbon accumulation at the Sanderson site in north Florida across two years of measurement for high and operational intensity management. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management included in the analysis. Ages examine d included 11 and 12 years. Units of total aboveground carbon mass are MgC.ha 1 and error bars show the standard error of the mean.

PAGE 65

65 Figure 2 3 . Means for soil respiration rates for bed and inter bed positions and families at the Austin Cary Forest and the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the analysis. Th e Austin Cary Forest (A) and the Sanderson site (B) are given and error bars show the standard error 2 .s 1 ).

PAGE 66

66 Figure 2 4 . Means for soil respiration rates by intensity at the Austin Cary Forest and the Sanderson site in north Flori da. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P; bed and inter bed) and measurement date (D) included in the analysis. The Austin Cary Fores t (A) and the Sanderson site (B) are given and error bars show the standard error 2 .s 1 ).

PAGE 67

67 Figure 2 5 . Means for soil respiration rates for bed and inter bed positions at the Austin Cary Forest and the Sanderson site in north Fl orida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and measurement date (D) included in the analysis. The Austin Cary Forest (A) and the Sa nderson site (B) are given and error bars show the standard error 2 .s 1 ).

PAGE 68

68 F igure 2 6 . Means for soil respiration rates by family at the Austin Cary Forest and the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P; bed and inter bed) and measurement date (D) included in the analysis. The Austin Cary Forest (A) and the Sanderson site (B) ar e given and error bars show the standard error 2 .s 1 ).

PAGE 69

69 Figure 2 7 . Means for soil respiration rates for bed and inter bed positions and intensities at the Sanderson site in north Florida. Two loblolly pine families (slow and fast grow ing) growing under operational and high intensity management with soil respiration measurement position (bed and inter bed) included in the analysis. Error bars show the standard error of the mean 2 .s 1 ).

PAGE 70

70 Figure 2 8 . Means for soil respiration rates for the interaction of management intensity and position at the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position (be d and inter bed) and measurement date included in the analysis. Error bars show the standard 2 .s 1 ).

PAGE 71

71 Figure 2 9 . Means for soil respiration rates for the interaction of management intensity, family, and position at the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position (bed and inter bed) and measurement date included in the analysis. Error bars show the standard 2 .s 1 ).

PAGE 72

72 Figure 2 10 . Means for soil respiration rates for bed and root exclusion positions at the Austin Cary Forest and the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position and measurement date included in the analysis. The Austin Cary Forest (A) and the Sanderson site (B) are given and error bars show the standard erro r of the mean 2 .s 1 ).

PAGE 73

73 Figure 2 11 . Means for soil respiration rates for bed and root exclusion positions and intensity at the Austin Cary Forest and the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position and measurement date included in the analysis. The Austin Cary Forest (A) and the Sanderson site (B) are given and error bars show the standard error mol.m 2 .s 1 ).

PAGE 74

74 Figure 2 12 . Means for soil respiration rates for bed and root exclusion positions and family at the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity managemen t with soil respiration measurement position included in the analysis. Error bars 2 .s 1 ).

PAGE 75

75 Figure 2 1 3 . Means for soil respiration rates for the interaction of bed and root exclusion positions, family, and measurement date at the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position included in the analysis. Error bars show the sta ndard error of the 2 .s 1 ).

PAGE 76

76 Figure 2 1 4 . Means for soil respiration rates for the interaction of bed and root exclusion positions and intensity at the Sanderson site in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position included in the 2 .s 1 ).

PAGE 77

77 Figure 2 1 5 . Means for soil respiration rates for bed and root ex clusion positions with an intensity x family interaction at the Austin Cary Forest in north Florida. Two loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position and m easurement date included in the analysis. Error bars show the standard 2 .s 1 ).

PAGE 78

78 Figure 2 1 6 . Means for soil respiration rates for pre and post fertilization periods and intensity at the Austin Cary Forest in north Florida. Tw o loblolly pine families (slow and fast growing) growing under operational and high intensity management with soil respiration measurement position and period included 2 .s 1 ).

PAGE 79

79 Figure 2 17. Temperature response curves for soil respiration at the Austin Cary Forest in north Florida. Curves are based on the Arrhenius equation, with soil temperature ( o 2 .s 1 ).

PAGE 80

80 Figure 2 18. Temperature response curves for soil respiration at the Sanderson site in north Florida. Curves are based on the Arrhenius equation, with soil temperature ( o 2 .s 1 ).

PAGE 81

81 Figure 2 1 9 . Means for soil moisture (%) through time at the Sanderson site fo r the bed and root exclusion positions of measurement. Error bars show the standard error of the mean.

PAGE 82

82 Figure 2 20 . intensities, high and operational, at the Austin Cary Forest for three measurement positions: the bed and root exclusion positions (A) and bed and inter bed positions (B). Note that the x axes of the graphs have different start dates. Error bars show the standard error of the mean.

PAGE 83

83 Figure 2 21 . Means of soil temper Flori d a for three measurement positions: the bed and root exclusion positions (A) and bed and inter bed positions (B). Note that the x axes of the graphs have different start dates. Error bars show the standard error of the mean.

PAGE 84

84 Figure 2 2 2 . Palmer drought severity index over the study period for north central Florida. Palmer drought severity index ranges from 6 to 6 with more negative values indicating drought conditions and positive values indicating wetter conditions.

PAGE 85

85 CHAPTER 3 EFFECTS OF FERTILIZATION AND WEED CONTROL TREATMENTS AND GENETIC S ON ANNUAL SOIL RESPIRATION , LITTERFALL , AND TOTAL BELOWGROUND CARBON FLUX IN PINUS TAEDA (L.) PLANTATIONS The largest terrestrial carbon (C) pool is found in soil, with soil respiration (SR) comprising the second largest flux in the C cycle (Schlesinger 1997, Luo and Zhuo, 2006). T he main reservoirs of terrestrial C are forests and forest soils, which contain 70% of the global terrestrial C poo l (Post et al., 1982; Luo and Zhuo, 2006). The size s of the global forest and forest soils C pool s are such that even small alterations of the C fluxes into and out of the soil pool could affect the global C cycle. In light of their global significance and the need to slow the rise of atmospheric CO 2 concentrations, forests have been identified as a means for enhancing CO 2 sequestration (McKinley et al., 2011; Maier et al., 2012). Globally, forests are an important C sink (Raich and Schlesinger, 1992). Pin e plantations span nearly 16 million ha in the southeastern United States, making them an important asset in the effort to reduce the effects of CO 2 emissions (Fox et al., 2007; Johnsen et al., 2001 ; Wear and Greis, 2013 ). In the southeastern US, loblolly pine ( Pinus taeda L.) is the most important commercial pine species (Schultz, 1997). Fertilizer, weed control, and improved genotypes are used regularly to improve forest production in these systems (Jokela et al., 2004). The production of fast growing and disease resistant families due to advances in genetics has enhanced aboveground productivity (McKeand et al., 2003). Similarly, silvicultural treatments, specifically fertilization and weed control, result in increase d southern pine productivity (Jokel a et al., 2004). Silvicultural treatments (fertilization and weed control) and improved families may interact and result in rank changes in the productivity and quality characteristics of

PAGE 86

86 the genotypes being deployed (McKeand et al., 2006; Roth et al, 2006 ). While it is evident that the combination of silvicultural treatments and improved genotypes can lead to aboveground gains in productivity, it is still uncertain if these aboveground gains result in increased belowground C sequestration. Soil respiratio n is linked to the climate, plant productivity, and site fertility (Luo and Zhou, 2006). These factors may differentially affect the components of SR: autotrophic respiration (RA) and heterotrophic respiration (RH). Autotrophic soil respiration is the resu lt of plant roots and associated fungi , whereas heterotrophic soil respiration is the result of microbial soil organisms. Partitioning the components of soil respiration, RA and RH, is a challenge generally recognized as being a major limitation to ecosyst em science and the ability to predict soil carbon fluxes (Hanson et al., 2000; Kuzyakov, 2006; Subke et al., 2006). Experiments that allow for the partitioning of RA and RH with fertilization and weed control treatments and different genotypes are importan t since their results could i nform how forest management affect s C fluxes and pools. Soil respiration rates are also known to vary in response to biotic factors , including : plant C allocation to roots, detritus, litterfall, and the microbial and fungal co mmunities present (Bowden et al. 2993; Crow et al. 2009; Chen et al., 2011; Raich and Tufekciogul, 2000; Ryan and Law, 2005; Luo and Zhou, 2006). Abiotic factors affecting SR may include soil temperature moisture, surface wind speed and air turbulence, soi l porosity, and CO 2 pressure gradients (Luo and Zhou, 2006). Carbon allocated to fine roots, root exudates, and mycorrhizae is referred to as total belowground carbon flux (TBCF); and is a key flux in forest ecosystems (Davidson et al., 2002; Giardinia et al., 2005). TBCF is the result of biologically mediated C

PAGE 87

87 movement from that fixed by photosynthesis to the soil. Although TBCF is a substantial and important C flux, it is one of the most difficult to understand, measure, and predict (Giardina et al., 20 05). Estimating TBCF assum es a near steady state condition with no changes in belowground C storage and is calculated by utilizing the difference in annual SR and annual litterfall measurements (Raich and Nadelhoffer, 1989). This method is supported in ma ture forests lacking disturbance, but there is controversy on its application to young, fertilized, or irrigated stands (Gower et al., 1996). Nonetheless, the approach does allow for the identification of the bounds of TBCF estimates in forest ecosystems ( Raich and Nadelhoffer, 1989; Davidson et al., 2002; Giardina and Ryan, 2002; Vogel et al. 2008). Fifty percent of the C stored in managed loblolly pine forests is found in the soil and some soil C pools are sensitive to weed control and fertiliz er applicat ions (Vogel et al., 2011). In general, forest management has had varied effects on soil C pools and SR rates (Johnson, 2001; Johnson et al., 2002; Nave et al., 2010), leaving the effects of genetic selection, fert ilization and understory competition contro l on belowground C processes poorly understood (Johnson , 1992; Post and Kwon , 2000). For example, d ecreased root and mycorrhizae C allocation have been demonstrated with the application of fertilizers (Haynes and Gower , 199 5 ; Giardina et al. , 2004; Treseder , 2004). Lee and Jose (2003) found that the effect of nutrient additions on SR may be species specific. Sarkhot et al. (2008) documented significant changes in SOC pools due to family genetic differences. Similarly, Major et al. (2012) found that genetics influe nced belowground C storage in a 32 year old black spruce ( Picea mariana (Mill.))

PAGE 88

88 stand . The larger issue to be addressed is how different management approaches used to enhance aboveground productivity (fertilization, weed control, gene tic selection, and combination) will affect the flux of C through the soil system and, specifically, the loss or accumulation of soil organic matter ( SOM ) . This research was conducted to assess the effects of intensive forest management, through fertilization and herbicide application, and slow and fast growing families of loblolly pine on total soil respiration, heterotrophic respiration, litterfall, and TBC F. An overarching objective was to determine if these processes would vary with components of net primary productivity (NPP) in a predictable manner, as NPP is directly related to the objectives of forest management and is a significant component of the ec osystem C balance. The following questions were addressed to determine the effects of intensive management and pine genotype deployment options on SR and TBCF: 1. H ow do genotypes and silvicultural treatments affect sources of C (litter, SR, and belowground a llocation) to the soil ? ; and 2. Do these sources of C (SR and belowground allocation) vary in a predictable way with aboveground growth? T hese questions were addressed by examining SR, litterfall, forest floor, and NPP across silvicultural management intensit ies and full sib loblolly pine families. The study consisted of two replicated, long term field experiments in north central Florida. These studies included the deployment of different sources of genetic materials for loblolly pine and two levels of silvicultural management intensity, as affected by weed control and fertilizer applications.

PAGE 89

89 Methods Study Areas This study included two sites; both sites were located in north central Florida and were comprised of managed loblolly pine forests. The fir st study location was the , 82 approximately 14 km no rtheast of Gainesville, FL, USA and 44 m above sea level. This site was established in 2000 to determine the effects of silvicult ure and genetics on plantation health and productivity of loblolly pine. Two intensities of silviculture were implemented, including a high intensity treatment (high fertilization, weed control, double pass bedding) and an operational intensity treatment ( operational fertilization, medium weed control, single pass bedding). Two full sib loblolly pine families were deployed on site with one spacing (1.8 m x 3 m). The second study location is near Sanderson, FL (30 . This site was es tablished in 2000 by the Forest Biology Research Cooperative (FBRC), located at the University of Florida, as part of a regional experiment, Pine Productivity Interactions on Experimental Sites (PPINES) . These studies ha ve examined the interactions of ful l sib loblolly and slash pine families to different management actions , including silvicultural intensity and planting density (Roth et al., 2007). This study used one PPINES experiment and focused on one planting density, two intensities of silviculture, and two genotypes of loblolly pine. One genotype, the fast growing family, is shared with the ACF forest. A subtropical and humid climate with long, hot, wet summers and mild, dry winters characterized both study sites in north central Florida. The two s ites were

PAGE 90

90 established on poorly drained Spodosols, but differed in regard to the soil series. The ACF site was classified as the Pomona soil series ( sandy, silice ous, hyperthermic Ultic Alaquods) , with a Bh (spodic) horizon within 74 cm of the surface and a Btg (argillic) horizon at a depth of 130 cm. The Sanderson site was classified as the Leon soil series ( sandy, siliceous, thermic Aeric Alaquods ), with a Bh horizon within 38 cm of the surface , but the argillic horizon was not present . As both were typical Coastal plain flatwoods sites , the understory vegetation was similar and included: sawtooth palmetto ( Serenoa repens (B.)), wax myrtle ( Myrica ceriferea L.), runner oak ( Quercus pumila W.), blueberries ( Vaccinium spp.), gallberry ( Ile x glabra (L.)), bluestem grasses ( Andropogon spp.), panic grasses ( Panicum spp.), and sedges ( Carex spp.). Study Designs Both sites were planted in 2000 by the FBRC; ACF was planted in December and Sanderson was planted in January. The ACF site had four blocks, and each block had a replicate of treatment and family plots. Plot sizes were 0.03 ha and included 60 trees per plot. The Sanderson site had four replicate blocks organized as a randomized complete block, split split plot design. Plot sizes were 0 .02 ha and included 48 measurement trees per plot. At each site, a three way factorial design (2x2x3) was implemented, which included silvicultural treatment intensity (high and operational intensity), genetic families (fast and slow growers), and SR measu rement position (bed, inter bed, and root exclusion). At Sanderson, all plots (high and operational treatments) were double bedded and chemically treated prior to planting. Arsenal (Imazapyr, 28.7% active ingredient, 1.02 L ha 1 ) and Garlon (Triclopyr, 60 % active ingredient, 7 . 02 L ha 1 ) were applied to eliminate herbaceous and woody competition (Roth et al., 2007). Containerized

PAGE 91

91 seedlings were planted at an initial spacing of 1.22 m x 2.75 m. The operational intensity plots included management practices commonly used by forest industry in the southeastern U.S. and consisted of an application of 50 kg.ha 1 N and 60 kg.ha 1 P as diammonium phosphate (DAP) at the time of planting (Roth et al., 2007). In 2006 the operational intensity plots were again fertili zed with 170 kg.ha 1 N and 28 kg.ha 1 P (as DAP and urea). The contrasting high intensity treatment was driven by early and complete vegetation control, along with fertilizer additions. Competing vegetation for the high treatment was controlled until canopy closure (age 5 yrs) using directed spray applications of Arsenal at 0.28 L/ha and Oust (sulfometuron methyl) at 0.14 L/ha. Directed spray applications of Roundup (glyphosate; 2% solution) were also used to inhibit the development of herbaceous competition. The fertilizer regime for the h igh treatment included 6 60 kg.ha 1 of 10 10 10 plus micronutrients at time of planting, followed by annual applications of macro and micro nutrient fertilizers using prescriptions based on foliar analyses (Table 2 1) . The last fertilizer treatment applied to the high intensity plots was in 2008 (225 kg.ha 1 N and 28 kg.ha 1 P applied as DAP and urea). Cumulative fertilizer N and P addition rates at Sanderson were 76 0 kg.ha 1 N and 1 80 kg.ha 1 P for the high intensity treatment and 220 kg.ha 1 N and 80 kg.h a 1 P for the operational intensity treatment . At ACF, bare root seedlings (1 0 stock) were planted at 1.8 m x 3 m spacing. Prior to planting the operational intensity treatment received a single bedding pass, initial understory competition control, and a fertilizer application consisting of 50 kg.ha 1 N and 60 kg.ha 1 P , applied as DAP, at time of planting (Table 2 1). The high intensity treatment was double bedded, received complete understory vegetation control from

PAGE 92

92 establishment until crown closure, and multiple fertilizer additions (10 10 10 plus micronutrients) based on foliar analyses (Table 2 1). Similar to Sanderson, herbicides were applied at labeled rates for Chopper (imazapyr) and Garlon (triclopyr) across both levels of treatment intensity. Directed spray applications of Roundup (2%) were used in the high treatment to control herbaceous and wood y competition until canopy closure (age 5 yrs). Cumulative fertilization rates for N and P at ACF through 2008 were 448 kg.ha 1 N and 104 kg.ha 1 P for the high intensity treatment and 50 kg.ha 1 N and 60 kg.ha 1 P for the operational intensity treatment. The high intensity treatment at ACF was re fertilized in March 2011 to be equivalent to the high intensity treatment at Sanderson. The resulting cum ulative N and P fertilization rates for the Sanderson and ACF high intensity treatments were 76 0 kg.ha 1 N and 1 80 kg.ha 1 P . At each site, two families of loblolly pine, selected through progeny testing, were wer was deployed at each site, with both ing , full family was a drought hardy family from east Texas. At Sanderson, a different slow growing full sib family was used that originated from the eastern lower Coastal Plain . The two families deployed at Sanderson were previously studied by Sarkhot et al. (2007) to assess their effects on C and N distributions among soil size fractions . In order to measure soil respiratio n from the same position, soil collars made of PVC pipe ( diameter 10.16 cm, height 8 cm) , were permanently installed 3 cm into the soil in the bed and inter bed position within each plot in December 2009 (three per plot for bed, inter bed, and root exclusi on) . The bed and inter bed positions were selected since the stands were spatially variable, with trees planted on the elevated bed position.

PAGE 93

93 A total of 16 collars were installed for each position at both study sites. In September 2010, root exclusion core s made of PVC pipes were installed in the bed positions within each plot to sever roots and create a solid root barrier to 75 cm. After installation, root exclusions were allowed to equilibrate for six weeks so that root decomposition could occur. Soil Re spiration At both sites, i nstantaneous total soil CO 2 efflux was measured using a Li COR 6400 infrared gas analyzer (IRGA) (Li Cor Inc., Lincoln, NE) with the 6400 09 Soil CO 2 Flux Chamber attachm ent. S oil CO 2 efflux was repeatedly measured monthly between December 2009 and June 2012. However, during the growing season (May September), SR was measured every two weeks. At ACF, SR measurements were made from April 2009 to April 2012 (3 years collected) and at Sanderson from November 2009 to May 2012 (two and half years collected). Root exclusions measurements began at both sites in October 2010 and ended in April 2012 at ACF and May 2012 at Sanderson (one year and six months collected). Litterfall Litterfall was collected every 30 60 days for each treatmen t and replication in circular litterfall traps (0.7 m 2 ). Four litterfall traps were randomly installed in each plot, two in the bed position and two in the inter bed position. Collected litterfall was dried (65 o C), separated into component parts (e.g., needles, twigs, cones), and weighed. Aboveground C Pools Annual inventory data (diameter at breast height) were collected for each treatment plot from ages 9 11 years at ACF and 11 12 years at Sanderson. Allometric equations generated for loblolly pine ( Gonzalez Benecke et al., in press) were used .

PAGE 94

94 Biomass estimates were converted to a C basis using a factor of 0.48 for the bole and branches and 0.45 for foliage (Vogt et al. , 1986 ) . Components were then summed to estimate total aboveground loblolly pine C mass. Net primary production N et primary production included biomass increment and annual litterfall. Annual fine root net primary production (< 2 mm ) was not included. Coarse root biomass, stemwood, branch and foliage increment was estimated from the difference in biomass estimated for two different years from annual measurements of tree diameter and height. Forest Floor Forest floor samples were collected using a 30.2 cm diamet er cutting ring. Bed and inter bed areas were sampled randomly three times each in every plot ( n= 6 samples (Vogel et al., 2011) C and weighed. The Oe + Oa horizon was sieved t o remove the mineral soil . Oi and Oe + Oa horizons were ground using a Wiley Mill to pass a 20 mm mesh screen , then mixed, and subsampled . The separated mineral soil was mortar and pestle ground until homogenized. From these samples, two samples from each horizon were analyzed for C and N concentrations on a CNS analyzer ( Costech ECS 4010 CHNS O Elemental Analyzer (Valencia, CA)) . Turnover time was calculated as forest floor C divided by average litterfall C. Soil Respiration Models D aily estimate s of eff lux for each of the treatment combination s (based on family, intensity and collar position) at each site were obtained using a multiple linear regression model . The model that was fitted estimated efflux for every half hour period

PAGE 95

95 based on available meteorological data , which had a slightly different set of variables ( e.g., Austin Cary Forest Ameriflux station and the Florida Automated Weather Network at the Macclenny Station). Several independent variables that were expected to affect efflux were pre selected for the ACF and Sanderson sites C), soil temperature (TS C), relative humidity (H, %), vapor pressure deficit (VPD, kPa), and solar radiation (R, W.m 2 ) were considered for ACF. At Sanderson, TA, TS, H, R and air temperatu re at 10 meters (TA10, C) were selected. Data obtained from the stations were paired with available field measurements. A multiple linear regression model was developed for each of the treatment combinations (k = 12 for each site) by individually fitting a model that considered as the dependent variable the log transformed efflux, ln(Efflux 1), and as independent variables those described above, together with their differences with respect to their inal models were fitted using the REG procedure as implemented in SAS 9.3 (SAS Institute, 2009) and contained only variables that were significant at a 5% level and that presented low collinearity levels (VIF < 5). The R 2 empirical was a coefficient of dete rmination obtained over the original (back transformed) response variable (efflux). The final fitted model for each treatment combination was later used to estimate efflux every half hour based on the metrological data available from the weather stations and then summarized on an annual basis over the span of the study ( e .g., three years at ACF, two years at Sanderson). Finally, adjustments were made to those predictions at the subplot level by year to obtain the same coefficient of variation as the observ ed mean efflux data.

PAGE 96

96 Litterfall, Soil Respiration and TBCF Analysis A repeated measures analysis of variance was implemented using yearly information on litterfall, soil respiration and total below ground carbon flux (TBCF) estimates, corresponding to three and two years of data collection at the ACF and Sanderson sites , respectively. For each response variable, available sub subplot measurements were a ccumulated and scaled over to an annual hectare basis (Mg.C.ha 1 .yr 1 ). Annual root exclusion estimates were adjusted for root decomposition (15% annually) using root C mass data presented in Chapter 2. The following linear mixed model, based on a Split s plit plot design with blocks was fitted: y = b + M + I + F + P + MI + MF + M P + IF + I P + F P + MIF + MI P + MF P + MIF P + w + s + e where, b is the random effect of block, with b ~ N(0, b ) ; M is the fixed effect of year of measur ement; I is the fixed effect of management intensity; F is the fixed effect of family; P is the fixed effect of collar position ; MI, MF, M P , IF, I P , F P , MIF, MI P , MF P and MIF P are second, third and fourth order fixed effect interactions between measurement , intensity, family, and position factors; w is the random effect of whole plot within block, with w ~ N(0, w ) ; s is the random effect of subplot within plot, with s ~ N(0, s ) ; and e is the random error, with e ~ N(0, R e ) and R is a matrix of that models correlations between repeated measurements of the same sub subplot, which was specified as an autoregressive of order 1 error structure.

PAGE 97

97 The significance of all model terms was evaluated using an approximate F test with a signi ficance level set at 10%. All models were fitted with the MIXED procedure as implemented in SAS 9.3 (SAS Institute, 2009). Results Net Primary Production At ACF, net primary production varied with management intensity and family (p<0.1) (Table 3 1). Net p rimary production for the high intensity management regime was 8.4 MgC.ha 1 y 1 , while the operational treatment was 5.6 MgC.ha 1 y 1 (p=0.00 82 ) (Table 3 2). Similarly, NPP for the fast growing family was greater than the slow growing family (8. 6 MgC.ha 1 versus 5.4 MgC.ha 1 , respectively) (Table 3 2). At Sanderson, NPP only varied with management intensity (Table 3 3 ). The high intensity treatment produced more C than the operational treatment (1 0.9 MgC.ha 1 y 1 versus 7. 5 MgC.ha 1 y 1 , respectively) (Table 3 4 ). Soil Respiration Models Soil respiration m odels selected at ACF included the variables : soil temperature (TS), relative humidity (RH), the difference of RH from the daily mean , and the difference of measured air temperature (TA) from th e daily mean . At Sanderson, The models selected at ACF for each treatment combination had the following general structur e: ln(Efflux 0 1 2 3 4 0 1 2 3 4 are coefficients of the independent variables, and e is the random error.

PAGE 98

98 The models selected at Sanderson for each treatment combination had the follo wing general structure: ln(Efflux 0 1 2 3 4 5 0 1 2 3 4 5 are coefficients of the independent variables, and e is the random error. Model coefficient estimates, standard errors, and R 2 emp are provided in Table 3 5 and Table 3 6 for ACF and Sanderson, respectively. Environmental Variables and Soil CO 2 Efflux The variables with the most common coefficients in the above models, soil temp erature and relative humidity at ACF, and soil temperature at Sanderson, were used to plot soil CO 2 efflux as a regression with those environmental factors. This was done to understand the amount of variability in soil CO 2 efflux that those environmental f actors could account for. The regression model at ACF had the following general structure: 0 1 2 RH 0 1 2 are coefficients of the independent variables. The regression model at Sanderson had the followi ng general structure: 0 1 TS 0 1 is the coefficient of the independent variable. At ACF, the soil temperature and relative humidity account ed for 32% of the variability in SR across all treatments (R 2 =0.3181) (Fi gure 3 1 ). At Sanderson, soil temperature account ed for 46% of the variability in SR (R 2 =0.4588) (Figure 3 2 ). The coefficients for these regression equations are presented in Table 3 7 .

PAGE 99

99 Annual Soil Respiration MgC.ha 1 y 1 ) (Table 3.1; 3.2). The operational management intensity had greater annual respiration than the high management intensity (p=0.0049) (18.3 versus 16.4 MgC.ha 1 y 1 ) (Table 3 9 ). The bed position emitted more C than the int er bed position (18.0 versus 16.6 MgC.ha 1 y 1 ) (Table 3 9 ). The interactive effect of intensity x position (p=0.033) demonstrated that under the operational regime, there were no difference s between position s, but under the high management regime the bed respired more C annually (17.8 MgC.ha 1 y 1 ) than the inter bed position (15.1 MgC.ha 1 y 1 ) , likely because roots were concentrated on the bed (Figure 3 3 ). The interactive effect of intensity x family MgC.ha 1 y 1 ) than the slow grower (17.2 MgC.ha 1 y 1 ) (Figure 3 4 ). At Sanderson, the effects of family, intensity x family, position , and intensity x position were significant (p<0.1) (Table 3 10 ). The effect of family demonstrated that the fast grower emitted more C (19.6 MgC.ha 1 y 1 ) than the slow grower (16.7 MgC.ha 1 y 1 )) (p=0.0001) (Table 3 10 , 3 11 ). The bed position respired significantly (p=0.0009) more C (19.4 MgC.ha 1 y 1 ) than the inter bed position (17.0 MgC.ha 1 y 1 ) (Table 3 10 , 3 11 ). The interaction of intensity x position (p=0.0002) showed differences betwee n the bed (20.7 MgC.ha 1 y 1 ) and inter bed (15.6 MgC.ha 1 y 1 ) position s under the operational intensity treatment, but no differences were apparent under the high intensity treatment

PAGE 100

100 (Table 3 10 , Figure 3 3 ). The operational intensity bed position also had the greatest flux, and all soil respiratory fluxes generally followed the pattern : operational, bed > high, inter bed (Figure 3 3 ). The interaction of intensity x family demonstrated differences (p=0.0361) between fam ilies under both the operational and high intensity management regimes ; the two families differed the most under the operational intensity treatment (Table 3 10 , Figure 3 4 ). For the intensity x family interaction, the combination of the fast growing famil y and operational intensity had the greatest SR (Figure 3 4 ). For this interaction, the respiratory fluxes followed the pattern: fast, operational > fast, high > slow, high > slow, operational (Figure 3 4 ). Annual Litterfall At ACF, intensity, family and y ear significantly affected litterfall (p<0.1) (Table 3 12 ). The high intensity management regime produced more litterfall (p=0.0492) (1.3 MgC.ha 1 y 1 ) than the operational treatment (1.0 MgC.ha 1 y 1 ) (Table 3 12 , 3 13 ). The fast growing family produced more litterfall (1.3 MgC.ha 1 y 1 ) than the slow growing family (1.1 MgC.ha 1 y 1 ) (p=0.0603) (Table 3 12 , 3 13 ). Year had a significant effect (p= .0372), with the second year of study producing more litterfall (1.6 MgC.ha 1 y 1 ) than the first (0.8 MgC.ha 1 y 1 ) or third year (1.1 MgC.ha 1 y 1 ) , and was likely due to drought in the first and third year of study (Table 3 12 , 3 13 ). The intensity x position and intensity x family interactions are displayed in Figure 3 3 and 3 4 , respectively, for comparison with the annual SR estimates. At Sanderson, litterfall was significant for intensity, position , intensity x position , year x intensity, and year x intensity x family (p<0.1) (Table 3 14 ). The high intensity than the operational treatment (Table 3 14 , 3 15 ). The inter bed position captured more litterfall (3.9 MgC.ha 1 y 1 ) than the bed

PAGE 101

101 position (3.3 MgC.ha 1 y 1 , p=0.0073) (Table 3 14 , 3 15 ). For both the year x intensity and year x intensity x family interac tions (p=0.0225 and p=0.0155, respectively), the second year had greater litterfall under the high intensity treatment, with the fast growing family producing more than the slow growing family (Table 3 14 , 3 15 ). For the year x intensity x family interaction, there was no difference under the operational management treatment (Table 3 15 ). The intensity x position interaction indicated that there was no difference s in litterfall between the operational intensity position s (Figure 3 3 ). The intensity x family interaction was not significant; however, it is displayed for comparison with the annual SR estimates (Figure 3 4 ). Total Belowground Carbon Flux At ACF , treatment effects on TBCF were significant for intensity, intensity x family, position , and intensity x position (p<0.1) (Table 3 16 ). Trees in t he high intensity management regime allocated less C belowground than those in the operational intensity (15.1 MgC.ha 1 y 1 versus 17.3 MgC.ha 1 y 1 ) (Table 3 16 , 3 1 7 ). The interaction of intensity x fam ily demonstrated differences in TBCF under the operational management treatment, with the fast grower being greater than the slow grower (18.3 MgC.ha 1 y 1 versus 16.3 MgC.ha 1 y 1 ); however, this was not the case under the high intensity treatment (approxim ately 15 MgC.ha 1 y 1 ) (Figure 3 4 ). Therefore, the TBCF was greater under the operational management regime (Figure 3 4 ). The bed position TBCF was greater than the inter bed (16.9 MgC.ha 1 y 1 versus 15.5 MgC.ha 1 y 1 ) (Table 3 16 , 3 17 ). In addition, the intensity x position interaction (p=0.0332) indicated that there were no differences between position s under operational management, but the bed and inter bed position s differed under the high intensity management regime (16.4 MgC.ha 1 y 1 versus 13.8 MgC.ha 1 y 1 , bed and inter bed, respectively) (Figure 3 3 ).

PAGE 102

102 At Sanderson, the following effects were significant for TBCF (p<0.1): intensity, family, intensity x family, position , and intensity x position (Table 3 18 ). Similar to ACF, u nder operational management, more C was allocated belowground than under the high intensity management regime (15.7 MgC.ha 1 y 1 versus 13.5 MgC.ha 1 y 1 , p=0.0002) (Table 3 1 9 ). The fast growing family allocated more C belowground than the slow growing family ( 15.9 MgC.ha 1 y 1 versus 13.4 MgC.ha 1 y 1 , p<0.0001) (Table 3 1 9 ). The intensity x family interaction (p=0.0001) indicated no difference between the two families in TBCF under the high intensity treatment. However, under the operational treatment, the fast grower had a greater TBCF (17.7 MgC.ha 1 y 1 ) than the slow grower (13.8 MgC.ha 1 y 1 ) (Figure 3 4 ). The pattern that TBCF followed was : fast, operational > fast, 4 ). Similar to ACF, t he bed position at Sanderson resulted in a greater (p<0.0001) TBCF than the inter bed position (16.1 MgC.ha 1 y 1 versus 13.1 MgC.ha 1 y 1 ) (Table 3 19 ). The intensity x position interaction was high ly significant (p=0.0001) and demonstrated that no differences occurred un der high management between the bed and inter bed position s. However, under the operational management treatment, the bed position had greater TBCF (20.7 MgC.ha 1 y 1 ) than the inter bed position (15.6 MgC.ha 1 y 1 ) (Figure 3 3 ). The pattern that TBCF follow ed was operational, bed > high, bed > high, inter bed (Figure 3 3 ). Bed and Root Exclusion Soil Respiration Comparison At ACF, t he intensity x family x position interaction was significant for the analysis of the bed and root exc lusion positions (p=0.0331) (Table 3 20 ). The slow grower, high intensity root exclusion combination had the lowest annual respiratory flux (13.8 MgC.ha 1 y 1 ) (Figure 3 5 ). The combination of the f ast grower under high intensity

PAGE 103

103 management in the root exclusion position had the greatest annual SR ( 19.1 MgC.ha 1 y 1 ) (Figure 3 5 ). Under operational management, there was no difference between families and positions (Figure 3 5 ). Under the high management intensity, the fast grower did not disp lay a difference between positions (Figure 3 5 ). At Sanderson, the interactions of intensity x family and family x position were significant (p<0.1) (Table 3 21 ). For the slow growing family, the root exclusion treatment respired more C annually than the bed position (Figure 3 6 ). For the fast growing family, the bed position respired more C annually than the root exclusion position (Figure 3 6 ). The slow growing family exhibited no differences in annual C SR under either management intensity (Figure 3 7 ) . The fast growing family responded to the high intensity treatment with less annual respiration than the operational treatment (17.8 MgC.ha 1 y 1 versus 21.0 MgC.ha 1 y 1 ) (Figure 3 7 ). Forest Floor At ACF, forest floor C accumulation was significantly af fected (p<0.1) by intensity, position , intensity x position and position x horizon (Table 3 22 ). The high intensity management regime accumulated more forest floor C (13.2 MgC.ha 1 ) than the operational treatment (7.9 MgC.ha 1 ) (Table 3 23 ). The inter bed position accumulated more forest floor C than the bed position (11.6 MgC.ha 1 versus 9.5 MgC.ha 1 ) , which wa s no t consistent with the litterfall findings since there was no difference between the two positions for litterfall C accumulation (Table 3 23 ). The intensity x position interaction (p=0.0276), suggested that there were no difference s in forest floor C content under the operational management regime between position s (Figure 3 8 ). However, there was greater forest floor C stored under the high t reatment than the operational treatment across position s (Figure 3 8 ). For the horizon x position

PAGE 104

104 interaction (p=0.0036), there was no difference between the Oi content for the two positions ( ~11 MgC.ha 1 ) (Figure 3 9 ). There was greater C content in the O a horizon of the inter bed position than the bed position (12.5 MgC.ha 1 versus 7.9 MgC.ha 1 ) (Figure 3 9 ) . At Sanderson, the forest floor C content was significantly affected by i ntensity, position , horizon, family x horizon, and family x intensity x p osition interactions (p<0.1) (Table 3 24 ). The high intensity management treatment resulted in a greater accumulation of forest floor C than the operational management treatment (13.1 MgC.ha 1 versus 8.9 MgC.ha 1 ) (p=0.0463) (Table 3 25 ). The inter bed pos ition accumulated more forest floor C than the bed position , which was consistent with litterfall inputs for this site (13.1 MgC.ha 1 versus 8.9 MgC.ha 1 ) (Table 3 25 ). The family x horizon interaction (p=0.0616) suggest that the fast grower had a greater C content in the Oa horizon than the slow grower, but no difference s existed between the families in the Oi horizon (Figure 3 10 ). The family x intensity x position x horizon interaction as displayed in Table 3 25 was weakly significant (p=0.0864) . Carbon Turnover Time At ACF, the C turnover time for the forest floor was significant for the position and intensity x position factors. The inter bed position had a longer turnover time than the bed position (8.8 years versus 7.4 years). The intensity x position interaction suggested that the turnover times were: High, inter bed (10.7 years) > operational, bed (7.5 years) high, bed (7.4 years) operational, inter bed (7.0 years). At Sanderson, the intensity, position , and inten sity x position effects were significant (p<0.1) for the forest floor C turnover time. In contrast to ACF, the h igh intensity management regime resulted in a shortened turnover time compared to the

PAGE 105

105 operat ional treatment (5.6 years versus 7.3 years). Simila r to ACF, t he inter bed position had a longer turnover time than the bed position (7.3 years versus 5.7 years). The intensity x position interaction resulted in the following pattern for turnover time: operational, inter bed (8.6 years) > high, inter bed (5.9 high, bed (5.3 years). Overall, the Sanderson site had faster turnover time than the ACF site. However, they did not follow the same pattern for the intensity x position interaction. Discussion Alterations in belowground C allocation and RH have the potential to modify patterns of forest productivity and soil C dynamics . By investigating the effects of varying intensities of silvicultural management (fertilization and weed control) and different families of loblolly pine on the belowground processes of SR and TBCF, the distinctive opportunity to examine genotype x environment interactions on belowground processes was presented at two long term (two and half to three years), replicated f ield experiments in north Florida . Through the pairing of long term SR data and environmental variables, model predictions were developed that enabled the comparison of ecosystem C fluxes as impacted by silvicultural treatments and genotype deployment dec isions . At ACF, NPP was much less than at Sanderson for the high intensity treatment, despite only being offset by one growing season (8.4 MgC.ha 1 y 1 versus 10.9 MgC.ha 1 y 1 ). This may account for the differences in litterfall between the two sites, si nce was approximately 5 MgC.ha 1 y 1 compared to ACF with approximately 3 MgC.ha 1 y 1 . Litterfall was less at ACF despite having the fast growing family in common with the Sanderson site. However, the fast growing fam

PAGE 106

106 litterfall did not respond the same at both sites under the high intensity treatment and in fact did not change drastically at ACF between management intensities. Total belowground C flux represents the second largest flux of C in forests after canopy C assimilation and represents the integration of SR and litterfall data ( Raich and Nadelhoffer, 1989; Davidson et al. , 2002). In mature forests, TBCF is usually two times the aboveground litterfall (Raich and Nadel hoffer, 1989; Davidson et al. , 2002). However, this ratio can be larger if litterfall rates are low, like in young forests, indicating a greater TBCF (Davidson et al. , 2002). Our study did not align with the ratio outlined by Raich and Nadelhoffer (1989), and was much larger. In this study annual SR ranged from 15.1 MgC.ha 1 .y 1 to 19.3 MgC.ha 1 .y 1 at ACF and 15.6 MgC.ha 1 .y 1 to 20.7 MgC.ha 1 .y 1 at Sanderson. Litterfall ranged from 1.8 MgC.ha 1 .y 1 to 3.6 MgC.ha 1 .y 1 at ACF and 2.3 MgC.ha 1 .y 1 and 5.4 MgC.ha 1 .y 1 at Sanderson. TBCF ranged from 12.2 MgC.ha 1 .y 1 to 17.0 MgC.ha 1 .y 1 at ACF and from 12.9 MgC.ha 1 .y 1 to 18.4 MgC.ha 1 .y 1 at Sanderson. Compared to Andrews and Schlesinger (2001) and Finzi et al. (2001), another loblolly pine forest at Duk e University had an annual SR of 9.94 MgC.ha 1 .y 1 and litterfall of 2.57 MgC.ha 1 .y 1 (Davidson et al., 2002). Our results are more comparable to rates of SR and litterfall at a Eucalyptus plantation in Hawaii, which ranged from 19.20 MgC.ha 1 .y 1 to 22.0 0 MgC.ha 1 .y 1 for annual SR and from 3.70 MgC.ha 1 .y 1 to 4.70 MgC.ha 1 .y 1 for litterfall (Giardina and Ryan, 2002; Davidson et al. , 2002). Similarly, our rates were close to those reported by Davidson et al. (2000) for a tropical evergreen old growth forest (20.00 MgC.ha 1 .y 1 for annual SR and 4.80 MgC.ha 1 .y 1 for litterfall) and a

PAGE 107

107 20 year old secondary forest (18.00 MgC.ha 1 .y 1 for annual SR and 4.75 MgC.ha 1 .y 1 for litterfall) in Paragominas, Brazil (Davidson et al. , 2002). In this study , annua l SR, litterfall, and TBCF were averaged over multiple years ; three years for ACF and two years for Sanderson. There was no significant difference across the years for TBCF ; therefore , the inter annual variation concerns expressed in several papers have be en addressed through averaging of more than one year of data (Davidson et al. , 2002; Pregitzer and Burton, 1991). Although the year effect was significant for litterfall at ACF and year x intensity, year x intensity x family, and year x position were signi ficant for litterfall at the Sanderson site, the interactions with year did not translate into temporal differences in TBCF. Annual estimates of SR were developed using a predictive, not explanatory, model in order to best fill gaps between sampling dates . The predictive SR model at ACF resulted in soil temperature and relative humidity as the significant factors for all of the model groups. At Sanderson, the predictive model resulted in soil temperature as the only significant factor for all of the model groups. Soil temperature has been recognized as one of the driving factors in SR rates (Lloyd and Taylor, 1994 ; Davidson et al., 1998 ). For t he significant factors for all models at each site, a correlation was performed to determine how much variation in SR could be accounted for by the factors. When the rates of SR were pooled and compared at ACF approximately 32% of the variation was accounted for by soil temperature and relative humidity (Figure 3 1 ). Similarly, at Sanderson, approximately 46% of the variability in SR was due to soil temperature (Figure 3 2 ). However, SR cannot be fully accounted for by these key environmental variables, suggesting that treatment effects and other environmental

PAGE 108

108 factors influenced the SR response. For example, differenc es in soil properties between sites may have influenced SR, since the ACF site had an argillic horizon (Btg) and the Sanderson site did not. Since soil texture can influence SR through moisture content and rooting systems, it is possible that some of the d ifferences between the two sites could be due to the water holding capacity of the soil s (Luo and Zhuo, 2006). Similarly, Sarkhot (2007) found that C storage in sandy soil depend ed on C protection by soil mic r o aggregates. The NPP of the fast growing f amily was greater than the slow growing family at ACF but not at Sanderson. However, overall aboveground C accumulation was greater at both sites for the fast growing family (Chapter 2). Greater rates of SR in the fast growing fa mily were likely due to gre ater stand biomass and productivity, which drove the increase in the autotrophic component of SR (Raich and Tufekciogul, 2000) (Figure 3 5 ). However, there was not a consistent pattern in SR that was tied to NPP in this study. Therefore, aboveground produc tivity at these two sites could not be used t o predict belowground responses ( r =0. 45620 , p =0. 2559 ), despite the findings of Chen et al. (2014) who described an allometric constraint between aboveground and belowground allocation; i.e., they allowed for oth er controls on belowground allocation like fine root production versus root respiration and above and belowground C trade off. The lack of relationship between NPP and SR could also be linked to the sources of C being utilized in SR. For example, RA is rel ated to recent photosynthate, and ultimately NPP (Hogberg and Read, 2006; Subke et al., 2011), whereas RH sources of C range from recent, labile forms of root and leaf litter to older, recalcitrant SOM (Schmidt et al.,

PAGE 109

109 2011). Since RA in this system was mi nimal compared to RH (Chapter 2), the majority of SR could not be attributed to aboveground production. In the short term, recent photosynthate and fine root turnover could be main sources of C for SR (Epron et al., 2012; Warren et al., 2012). Fine roots allocate C to the soil through their turnover and exudates that deliver usable substrate for RH (Wiant , 1967). Therefore, more fine roots can both increase SR through greater RA or increase SR through greater RH, as the roots both respire and lead to great er RH by providing valuable substrate. At the A C F site, RH rates responded the same under the operational treatment for both families, but responded differentially under the high intensity treatment, with the slow growing family producing less RH than the fast growing family (Figure 3 5 ). At Sanderson, the families behaved differently, with the (Figure 3 6 ). Despite these differences in responses of RH, fine root C was consisten t between the two sites; the slow growing family produced more fine root C than the fast growing family and the operational management regime produced more fine root C than the high intensity management treatment (Chapter 2). Therefore, other factors , in addition to fine root C , were contributing to the differences in SR and RH. For example, ACF was fertilized the year root exclusions were installed, and fertilization effects may take time to become apparent (Albaugh et al., 1998). Similarly, the root s ampling date was prior to fertilization at ACF ; therefore, belowground non quantified changes may have occurred in fine root C production , which affected SR (Burton et al., 201 2 ).

PAGE 110

110 Multiple factors interact to alter SR. For example, even though SR was less in the high intensity treatment, increased litterfall in the high intensity vers u s operational treatment could have led to an increase in SR rates, despite the high intensity SR being less that the operational treatment (Sulzman et al., 20 05 ). In addition, litter manipulations have demonstrated that RH rates can be influenced by the addition or exclusion of litter (Bowden et al., 1993; Sayer, 2006; Chemidlin Prevost Boure et al., 2010; Sulzman et al., 20 05 ). Fertilization effects on SR and RH tend to be variable (Maier and Kress 2000; Butnor et al., 2003; Tyree, 2006). In this study, the high intensity treatment reduced SR for the fast growing family , but it did not change it for the slow growing family at either site (Figure 3 4 ). Fertilizatio n has been shown to reduce microbial respiration in the short term, which could carry over to total SR rates (Thirukkuran and Parkinson, 2000; Gough and Seiler, 2004; Tyree et al., 2008). However, other studies have demonstrated increases , or no effect of fertilization on RH . (Gallardo and Schlesinger, 1994; Samuelson et al., 2009; Kim et al., 2012). This may explain the differential response of the families under the high intensity management regime at ACF and the intensity x family interaction at Sanderso n (Figure 3 5 , 3 6 ). Other possibilit ies were that the root exclusion estimates were er soil moisture contents (Chapter 2), thereby altering the microbial populations of the collar (Moyano et al., 2013) . A ltern atively , the predictive models may have overestimated the root . Belowground C allocation cannot be easily measured ; therefore , it is often estimated as the difference between SR and litterfall, assuming steady state conditions

PAGE 111

111 (Raich and Nadelhoffer, 1989). When TBCF was calculated, changes in belowground allocation were apparent at both sites. The h igh intensity management treatment was associated with reduce d TBCF for both families at both study sites across several years (Fig ure 3 4 ). This wa s a consistent finding despite high spatial variability and imperfect sampling methods (Yanai et al., 1999; Yanai et al., 2003; Ryzhova and Podvezennaya 2008). The intensity x position interactions at both sites indicated that there were differences between position s of measurement and management intensities (Figure 3 3 ). Although the sampling positions behaved differently for SR, litterfall consistently increased with increasing management intensity and TBCF consistently decreased a s man agement intensity increase d (Figure 3 3 ). Similarly, i t wa s evident that int ensity x family interactions were a significant factor in the assessment of annual SR , litterfall, and TBCF rates (Figure 3 4 ) . Both families responded to intensive management with a decrease in TBCF ; this is similar to another study that demonstrated a decrease in belowground C cycling with fertilization (Giardina et al., 2004). The interactions of intensity x position and intensity x family demonstrated that intensive management d ecreased TBCF (Figure 3 3 and 3 4 ). However, for our study, there was no correlation between NPP and TBCF, although Litton et al. (2007) found that increased resource availability increased wood production and decreased TBCF. Litton et al. (2007) further a cknowledged that partitioning between aboveground and TBCF can be highly variable. Under the operational management treatment, families behaved differently; the family at both sites (Figure 3 4 ). Similarly, SR for the fast growing family was greater under operational

PAGE 112

112 management. The slow growing family responded with no change in SR between management intensities, but the fast growing family resulted in a decrease (Figure 3 4 ). Since the fast growing family used in this study is commonly deployed commercially by forest industry , this research has implications for C emissions under varying levels of management inputs . These findings are in line with those of Stovall et al. (2013), who reported that two full sib loblolly pine clones differed in their responses to fertilization. Furthermore, Tyree et al. (2013) documented differences in SR between two contrasting loblolly pine clones, but found no change in overall SR due to fertilization, althoug h there was a change in RH. However, belowground studies with a genetic component s are limited, making this study with two study sites involved especially unique. The differences in the families under operational intensity management may be due to the diff erences in nitrogen use efficiency of the families. Li et al. (1991) found family differences in nitrogen use efficiency at low nitrogen levels. Similarly, Li et al. (1991) also identified that nitrogen use efficiency had a moderate to high degree of genet ic control. Litter and SOM decomposition rates can be altered by the addition of nitrogen. For example, decomposition rates were higher when nutrient concentrations of litter or roots were high (Melillo et al., 1982; Silver and Miya, 2001). However, in th is study turnover time followed no strict pattern; at ACF the operational, inter bed position (7.0 years) and the Sanderson the high intensity, bed position had the shortest turnover times (5.3 years). Forest floor C storage was not negatively affected by the high management intensity at ACF and at Sanderson there was no significant difference between management intensities (Figure 3 8 ). However, at Sanderson, the fast growing

PAGE 113

113 family accumulated more C in the forest floor than the slow growing family (Figur e 3 10 ). Similarly, at Sanderson, Sarkhot et al. (2008) found significant changes in SOC pools due to family genetic differences. Also at the Sanderson site, Sarkhot et al. (2007) found that soil aggregation could be affected negatively when herbicides we re applied and understory biomass was reduced. However, this was not evident in the forest floor C at either the Sanderson or ACF site s. The multi year datasets at two study sites with different soil types demonstrated significant intensity x family inte ractions for SR and TBCF. These findings, therefore, lend support that genotype x environment interactions can occur not only aboveground but also belowground. These results also argue that genotype x environment should be considered in the southeastern Un ited States when estimating forest C budgets, especially for families that are widely deployed, such as the fast grower in this study. Summary The results of this two site, multi year study have shown that intensive forest management and the deployment of improved genotypes can significantly influence annual SR, litterfall, and TBCF in managed loblolly pine forests. However, these differences could not be reliably predicted by aboveground productivity alone. This research highlights the importance of rec ognizing differences among genotypes and management intensities when scaling both above and belowground C estimates since three of the four genotypes responded similarly to the management treatments . Future research regarding the mechanisms and controls behind the belowground intensity x family interactions would allow for a more comprehensive understanding of how management intensity and genetic deployment decisions interact to store C in the long ter m. In addition, an in depth examination of RH is needed that focuses on

PAGE 114

114 heterotrophic soil organism populations as a means to provide better insight as to why RH is such a dominant component of SR in these managed forest ecosystems.

PAGE 115

115 Table 3 1. Analysis of variance for net primary productivity for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) and age (A; years) were included in the analysis. Effect Num DF Den DF F Value P value I 1 5 18 0.00 82 F 1 11.6 41.05 < 0.000 1 I x F 1 17.8 0.1 9 0 .6672 A 1 3.68 0.06 0.8238 A x I 1 13.9 0.06 0.8038 A x F 1 11.7 0.21 0.6522 A x I x F 1 17.8 0.34 0.5645 Table 3 2. Means for net primary productivity for the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) were included in the analysis. Standard e rrors are in parentheses and the units are Mg C.ha 1 y 1 . Effect Intensity Family Estimate I High 8.4 (0. 4 ) I Oper 5.6 (0. 4 ) F Fast 8. 6 (0. 4 ) F Slow 5.4 (0. 4 ) Table 3 3. Analysis of variance for net primary productivity for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 6 6.48 0.0 438 F 1 6 0. 10 0.7 622 I x F 1 6 0.03 0.8 758

PAGE 116

116 Table 3 4. Means for net primary productivity for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) were included in the analysis. Standard errors are in parentheses and the units are Mg C.ha 1 y 1 . Effect Intensity Family Estimate I High 1 0.9 ( 0.9 ) I Oper 7.5 ( 0.9 ) F Slow 9.0 ( 0.9 ) F Fast 9. 4 ( 0.9 )

PAGE 117

117 Table 3 5 . Coefficient estimates and R 2 emp of soil respiration models for the Austin Cary Forest in north Florida. Coefficient milies examined are fast and slow growing, intensities examined are high and operational (oper), and positions examined are bed, inter bed (inter), and root exclusion (RE). The time scale of the model is half hourly. Standard errors are in parentheses. Fam ily Intensity Position Intercept TS H R 2 Fast High Bed 0.24580 0.03596 0.00947 0.00712 0.4907 (0.1159) (0.0051) (0.0020) (0.0025) Inter 0.48646 0.03682 0.00440 0.3255 (0.1286) (0.0060) (0.0018) RE 0.48850 0.02984 0.00921 0.01655 0.06616 0.5457 (0.1600) (0.0087) (0.0028) (0.0034) (0.0160) Oper Bed 0.35150 0.03342 0.00982 0.01366 0.02937 0.4383 (0.1472) (0.0057) (0.0024) (0.0037) (0.0168) Inter 0.36639 0.03007 0.01089 0.01566 0.03878 0.3881 (0.1759) (0.0068) (0.0025) (0.0038) (0.0181) RE 0.07359 0.03475 0.01433 0.01278 0.6203 (0.1732) (0.0076) (0.0023) (0.0033) Slow High Bed 0.46453 0.04609 0.00390 0.01223 0.06864 0.4927 (0.1210) (0.0054) (0.0021) (0.0036) (0.0164) Inter 0.34842 0.04604 0.00329 0.00777 0.03417 0.5535 (0.1096) (0.0048) (0.0019) (0.0030) (0.0141) RE 0.15906 0.04588 0.00519 0.01606 0.06605 0.6294 (0.1460) (0.0073) (0.0025) (0.0034) (0.0168) Oper Bed 0.14642 0.04756 0.00699 0.00952 0.4128 (0.1467) (0.0061) (0.0025) (0.0030) Inter 0.34200 0.04144 0.00599 0.00558 0.3621 (0.1466) (0.0062) (0.0023) (0.0028) RE 0.18460 0.03364 0.01087 0.01341 0.03935 0.6051 (0.1827) (0.0073) (0.0022) (0.0034) (0.0169)

PAGE 118

118 Table 3 6 . Coefficient estimates and R 2 emp of soil respiration models for the Sanderson site in north Florida. Coefficient es examined are fast and slow growing, intensities examined are high and operational (oper), and positions examined are bed, inter bed (inter), and root exclusion (RE). The time scale of the model is half hourly. Standard errors are in parentheses. Family Intensity Position Intercept TS TA R R 2 Fast High Bed 0.75343 0.04243 0.00026 0.01453 0.6453 (0.0607) (0.0023) (0.0001) (0.0056) Inter 0.80091 0.04314 0.00044 0.01359 0.6584 (0.0699) (0.0026) (0.0001) (0.0055) RE 0.76611 0.05079 0.01712 0.4668 (0.1034) (0.0067) (0.0067) Oper Bed 0.78959 0.04611 0.00038 0.6535 (0.0590) (0.0023) (0.0001) Inter 0.64475 0.04525 0.00037 0.6453 (0.0644) (0.0026) (0.0001) RE 0.87140 0.04834 0.01446 0.00040 0.4864 (0.0927) (0.0066) (0.0063) (0.0001) Slow High Bed 0.84471 0.03480 0.02170 0.5327 (0.0647) (0.0023) (0.0056) Inter 0.84202 0.03762 0.00047 0.01833 0.5446 (0.0742) (0.0029) (0.0001) (0.0060) RE 1.09154 0.03493 0.00047 0.01727 0.4579 (0.0933) (0.0039) (0.0001) (0.0063) Oper Bed 0.71498 0.04184 0.00020 0.6301 (0.0559) (0.0022) (0.0001) Inter 0.70080 0.03299 0.3999 (0.0787) (0.0032) RE 0.62086 0.05659 0.01340 0.5450 (0.0970) (0.0067) (0.0060)

PAGE 119

119 Table 3 7 . Soil respiration regression coefficients for the Austin Cary Forest and Sanderson site in north Florida. Soil temperature (TS) and relative humidity (H) are the coefficients. Standard errors are in parentheses. Site Intercept TS H R 2 P Value ACF 0.9752 (0.1895) 0.2052 (0.0097) 0.0188 (0.0025) 0.3181 <0.0001 Sanderson 0.3651 (0.1149) 0.1900 (0.0047) 0.4588 < 0.0001

PAGE 120

120 T able 3 8 . Repeated measures analysis of variance for annual soil respiration for bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 14.2 11.04 0.0049 F 1 20.2 2.77 0.1114 I x F 1 15.5 5.27 0.0361 P 1 12.2 6.58 0.0245 I x P 1 13.2 5.67 0.033 F x P 1 18.4 1.4 0.2514 I x F x P 1 16.9 0.16 0.6978 Y 2 8.44 0.11 0.898 Y x I 2 14.2 0.01 0.992 Y x F 2 20.2 0.03 0.9741 Y x I x F 2 15.5 0 0.9998 Y x P 2 12.2 0 0.9989 Y x I x P 2 13.2 0 0.9958 Y x F x P 2 18.4 0 0.9999 Y x I x F x P 2 16.9 0.01 0.9898 T able 3 9 . Mean annual soil respiration for the bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and year (Y) were included in the analysis. Annual soil respiration units are MgC.ha 1 .y 1 and standard errors are in parentheses. Effect Intensity Family Position Estimate I High 16.4 (0.6) I Oper 18.3 (0.6) F Fast 17.8 (0.6) F Slow 16.9 (0.6) P Bed 18.0 (0.5) P Inter 16.6 (0.5)

PAGE 121

121 Table 3 10 . Repeated measures analysis of variance for annual soil respiration for bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 21 0.02 0.8824 F 1 21 21.88 0.0001 I x F 1 21 4 0.0586 P 1 21 15 0.0009 I x P 1 21 20.13 0.0002 F x P 1 21 0 0.9747 I x F x P 1 21 0.68 0.418 Y 1 24 0.6 0.4459 Y x I 1 24 0 0.9773 Y x F 1 24 0 0.9573 Y x I x F 1 24 0 0.9829 Y x P 1 24 0.01 0.9432 Y x I x P 1 24 0 0.9551 Y x F x P 1 24 0 0.9984 Y x I x F x P 1 24 0 0.9847 Table 3 11 . Mean annual soil respiration for the bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with soil respiration measurement position (P) and year (Y) were included in the analysis. Annual soil respiration units are MgC.ha 1 .y 1 and standard errors ar e in parentheses. Effect Family Position Estimate F Slow 16.7 (0.5) F Fast 19.6 (0.5) P Bed 19.4 (0.5) P Inter 17.0 (0.5)

PAGE 122

122 Table 3 12 . Repeated measures analysis of variance for annual litterfall for bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 8.03 5.36 0.0497 F 1 18.9 3.99 0.0604 I x F 1 5.18 2.66 0.1625 P 1 2.18 0 0.9869 I x P 1 3.26 2.21 0.2273 F x P 1 12 0.1 0.7602 I x F x P 1 9.47 0 0.953 Y 2 1.65 43.76 0.0377 Y x I 2 5.09 0.4 0.6911 Y x F 2 19.6 1.35 0.2842 Y x I x F 2 8.36 0.17 0.8424 Y x P 2 2.23 0.07 0.933 Y x I x P 2 3.28 0.44 0.6758 Y x F x P 2 12.1 0.17 0.8459 Y x I x F x P 2 9.6 0.39 0.6889 Table 3 13 . Mean annual litterfall for the bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Annual litterfall units are MgC.ha 1 .y 1 and standard errors are in parentheses. Effect Year Intensity Family Estimate I High 2.9 (0.2 ) I Oper 2.3 (0.2 ) F Fast 2.8 (0.2 ) F Slow 2.4 (0.2 ) Y 1 1.8 (0.2 ) Y 2 3.6 (0.2 ) Y 3 2.4 (0.2 )

PAGE 123

123 Table 3 14 . Repeated measures analysis of variance for annual litterfall for bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 5.61 36.7 0.0012 F 1 17.8 2.68 0.1191 I x F 1 17.8 0.24 0.6311 P 1 17.2 9.26 0.0073 I x P 1 17.2 4.4 0.0512 F x P 1 17.2 2.88 0.1075 I x F x P 1 17.2 0.24 0.6307 Y 1 24 1.86 0.185 Y x I 1 24 5.95 0.0225 Y x F 1 24 2.52 0.1258 Y x I x F 1 24 6.79 0.0155 Y x P 1 24 6.05 0.0215 Y x I x P 1 24 0.69 0.4134 Y x F x P 1 24 1.61 0.2173 Y x I x F x P 1 24 2.91 0.1012

PAGE 124

124 Table 3 15 . Mean annual litterfall for the bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Annual litterfall units are Mg C.ha 1 .y 1 and standard errors are in parentheses. Effect Intensity Family Position Year Estimate I High 4.7 (0.3) I Oper 2.4 (0.3) P Bed 3.3 (0.2) P Inter 3.9 (0.2) Y x I High 1 4.5 (0.3) Y x I Oper 1 2.5 (0.3) Y x I High 2 4.9 (0.3) Y x I Oper 2 2.4 (0.3) Y x I x F High Slow 1 4.5 (0.3) Y x I x F High Fast 1 4.6 (0.4) Y x I x F Oper Slow 1 2.3 (0.3) Y x I x F Oper Fast 1 2.6 (0.4) Y x I x F High Slow 2 4.5 (0.3) Y x I x F High Fast 2 5.4 (0.4) Y x I x F Oper Slow 2 2.3 (0.3) Y x I x F Oper Fast 2 2.4 (0.4) Y x P Bed 1 3.3 (0.2) Y x P Inter 1 3.7 (0.2) Y x P Bed 2 3.2 (0.2) Y x P Inter 2 4.1 (0.2)

PAGE 125

125 T able 3 16 . Repeated measures analysis of variance for total belowground carbon flux for bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 15.1 17.74 <0.0001 F 1 20.8 2.32 0.2048 I x F 1 16.3 4.91 0.0545 P 1 13.1 7.53 0.015 I x P 1 14.1 5.58 0.0387 F x P 1 19.2 1.53 0.2175 I x F x P 1 17.7 0.15 0.7152 Y 2 5.56 0.41 0.3289 Y x I 2 15.1 0.01 0.9564 Y x F 2 20.8 0.06 0.8154 Y x I x F 2 16.3 0 0.9919 Y x P 2 13.1 0 0.99 Y x I x P 2 14.1 0.01 0.9354 Y x F x P 2 19.2 0 0.9942 Y x I x F x P 2 17.7 0.04 0.9033 Table 3 1 7 . Mean total belowground carbon flux for the bed and inter bed positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Total belowgro und carbon flux units are MgC.ha 1 .y 1 and standard errors are in parentheses. Effect Year Intensity Family Position Estimate I High 13.5 (0.6) I Oper 16.1 (0.6) F Fast 15.1 (0.6) F Slow 14.4 (0.6) Y 1 15.4 (0.9 ) Y 2 13.6 (0.9 ) Y 3 15.3 (0.9 ) P Bed 1 5 . 5 (0.6) P Inter 14 . 1 (0.6)

PAGE 126

126 Table 3 1 8 . Repeated measures analysis of variance for total belowground carbon flux for bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Effect Num DF Den DF F Value P Value I 1 21 20.25 0.0002 F 1 21 25.98 <0.0001 I x F 1 21 7.24 0.0137 P 1 21 36.71 <0.0001 I x P 1 21 22.09 0.0001 F x P 1 21 0.55 0.4676 I x F x P 1 21 0.68 0.4198 Y 1 24 0.2 0.6614 Y x I 1 24 0.4 0.5344 Y x F 1 24 0.11 0.7393 Y x I x F 1 24 0.44 0.5126 Y x P 1 24 0.46 0.5037 Y x I x P 1 24 0.02 0.8836 Y x F x P 1 24 0.1 0.7559 Y x I x F x P 1 24 0.16 0.6921 Table 3 1 9 . Mean total belowground carbon flux for the bed and inter bed positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and year (Y) were included in the analysis. Total belowground carbon flux units are MgC.ha 1 .y 1 and standard errors are in parentheses. Effect Intensity Family Position Estimate I High 13.5 (0.5) I Oper 15.7 (0.5) F Slow 13.4 (0.5) F Fast 15.9 (0.5) P Bed 16.1 (0.5) P Inter 13.1 (0.5)

PAGE 127

127 Table 3 20 . Repeated measures analysis of variance for annual soil respiration for bed and root exclusion positions at the Austin Cary Forest in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) included in the analysis. Effect Num DF Den DF F Value P Value I 1 21 0.26 0.6448 F 1 21 2.45 0.1346 I x F 1 21 0.28 0.6061 P 1 21 2.23 0.0699 I x P 1 21 0 0.9246 F x P 1 21 2.13 0.1621 I x F x P 1 21 5.2 0.033 Table 3 21 . Repeated measures analysis of variance for annual soil respiration for bed and root exclusion positions at the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) included in the analysis. Effect Num DF Den DF F Value P Value I 1 6 1.03 0.3916 F 1 18 0.1 0.6852 I x F 1 18 4.58 0.0553 P 1 18 0.25 0.4844 I x P 1 18 2.25 0.139 F x P 1 18 5.93 0.039 I x F x P 1 18 0.23 0.6525

PAGE 128

128 Table 3 22 . Analysis of variance for forest floor carbon for the Austin C ary in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and horizon (H; Oi and Oe+Oa) included in the analysis. Effect Num DF Den DF F Value P Value F 1 6 2.41 0.1713 I 1 3 9.78 0.0522 F x I 1 6 1.23 0.3102 P 1 36 6.98 0.0121 F x P 1 36 1.73 0.1971 I x P 1 36 5.27 0.0276 F x I x P 1 36 1.75 0.1945 H 1 36 0.63 0.4333 F x H 1 36 1.28 0.2654 I x H 1 36 0.22 0.6447 F x I x H 1 36 0.08 0.7812 L x H 1 36 9.71 0.0036 F x P x H 1 36 0.42 0.5201 I x P x H 1 36 0.03 0.863 F x I x P x H 1 36 0.01 0.9153 Table 3 23 . Mean forest floor carbon for the Austin Cary in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and horizon (H; Oi and Oe+Oa) included in the analysis. Forest floor units are Mg C.ha 1 and standard errors are in parentheses . Effect Intensity Position Estimate I High 13.2 (1.4) I Oper 7.9 (1.4) P Bed 9.5 (1.2) P Inter 11.6 (1.2)

PAGE 129

129 Table 3 24 . Analysis of variance for forest floor carbon for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and horizon (H; Oi and Oe+Oa) included in the analysis. Effect Num DF Den DF F Value P V alue F 1 3.7 4.31 0.1119 I 1 5.73 6.42 0.0463 F x I 1 2.54 3.29 0.1841 P 1 36 35.33 <0.0001 F x P 1 36 0.69 0.4101 I x P 1 36 1.26 0.2698 F x I x P 1 36 0.24 0.6305 H 1 36 1.11 0.2982 F x H 1 36 3.72 0.0616 I x H 1 36 1.33 0.2557 F x I x H 1 36 2.34 0.1345 L x H 1 36 1.37 0.25 F x P x H 1 36 0.51 0.4796 I x P x H 1 36 1.1 0.3007 F x I x P x H 1 36 3.11 0.0864

PAGE 130

130 Table 3 25 . Mean forest floor carbon for the Sanderson site in north Florida. Two loblolly pine families (F; slow and fast growing) growing under operational and high intensity management (I) with measurement position (P) and horizon (H; Oi and Oe+Oa) included in the analysis. Forest floor units are Mg C.ha 1 and standard errors are in parentheses . Effect Family Intensity Position Horizon Estimate I High 13.1 (1.2) I Oper 8.9 (1.2) P Bed 8.9 (0.9) P Inter 13.1 (0.9) F x I x P x H Slow High Bed Oe+Oa 8.8 (1.8) F x I x P x H Slow High Bed Oi 9.5 (1.8) F x I x P x H Slow High Inter Oe+Oa 11.8 (1.8) F x I x P x H Slow High Inter Oi 15.8 (1.8) F x I x P x H Slow Oper Bed Oe+Oa 5.4 (1.8) F x I x P x H Slow Oper Bed Oi 9.5 (1.8) F x I x P x H Slow Oper Inter Oe+Oa 10.1 (1.8) F x I x P x H Slow Oper Inter Oi 9.6 (1.8) F x I x P x H Fast High Bed Oe+Oa 12.5 (1.8) F x I x P x H Fast High Bed Oi 11.9 (1.8) F x I x P x H Fast High Inter Oe+Oa 19.5 (1.8) F x I x P x H Fast High Inter Oi 15.2 (1.8) F x I x P x H Fast Oper Bed Oe+Oa 5.9 (1.8) F x I x P x H Fast Oper Bed Oi 8.0 (1.8) F x I x P x H Fast Oper Inter Oe+Oa 11.0 (1.8) F x I x P x H Fast Oper Inter Oi 11.5 (1.8)

PAGE 131

131 Figure 3 1. The interaction of soil respiration, soil temperature, and relative humidity at the Austin Cary Forest in north Florida over three years of measurement. 2 .s 1

PAGE 132

132 Figure 3 2. The response of soil respiration to soil temperature at the Sanderson site in north Florida over two years of measurement. Units are, respect 2 .s 1

PAGE 133

133 Figure 3 3 . Effects of management treatments and positions on the mean fluxes of soil respiration, litterfall, and total belowground carbon flux at the Austin Cary Forest and Sanderson site. Note the scale for B and E differs from the rest. The fluxes at Austin Cary Forest (A through C, respectively) are the means over three years of measurement and the fluxes at Sanderson (D through F, respectively) are the means over two years of measurement.

PAGE 134

134 Figure 3 4 . Effects of management treatments and families on the mean fluxes of soil respiration, litterfall, and total belowground carbon flux at the Austin Cary Forest and Sanderson site. Note the scale for B and E differs from the rest. The fluxes at Austin Cary F orest (A through C, respectively) are the means over three years of measurement and the fluxes at Sanderson (D through F, respectively) are the means over two years of measurement.

PAGE 135

135 Figure 3 5 . Means for annual soil CO 2 efflux for the interaction of man agement intensity x family x measurement position at the Austin Cary Forest in north Florida. Position (bed and root exclusion (RE)), family (fast or slow grower), and management intensity (high or operational) were examined. Error bars show the standard e rror of the mean.

PAGE 136

136 Figure 3 6 . Means for annual soil CO 2 efflux for the interaction of measurement position and family at the Sanderson site in north Florida. Position (bed and root exclusion (RE)), family (fast or slow grower), and management intensity ( high or operational) were examined. Error bars show the standard error of the mean.

PAGE 137

137 Figure 3 7 . Means for annual soil CO 2 efflux for the interaction of measurement position and management intensity at the Sanderson site in north Florida. Position (bed and root exclusion), family (fast or slow grower), and management intensity (high or operational) were examined. Error bars show the standard error of the mean.

PAGE 138

138 Figure 3 8 . Means for forest floor C for the interaction of sampling position and management intensity at the Austin Cary Forest in north Florida. Horizon (Oi and Oe+Oa), position (bed and inter bed), family (fast and slow growing), and management intensity (high and operational) were examined. Error bars show the standard error of the mean.

PAGE 139

139 Figure 3 9 . Means for forest floor C for the interaction of sampling position and horizon at the Austin Cary Forest in north Florida. Horizon (Oi and Oe+Oa), position (bed and inter bed), family (fast and slow growing), and management intensity (high and o perational) were examined. Error bars show the standard error of the mean.

PAGE 140

140 Figure 3 10 . Means for forest floor C for the interaction of horizon and family at the Sanderson site in north Florida. Horizon (Oi and Oe+Oa), position (bed and inter bed), fami ly (fast and slow growing), and management intensity (high and operational) were examined. Error bars show the standard error of the mean.

PAGE 141

141 CHAPTER 4 CONCLUSIONS Two long term fores t productivity studies, with two , full sib loblolly pine ( Pinus taeda L.) families deployed at each location , were used to examine genotype x environment interactions on soil respiration (SR) and belowground allocation on Spodosols in north Florida. Both experiments advance d the understanding of the effects of intensive for est management and the deployment of contrasting genotypes, fast and slow growing families, on the components of SR, including: autotrophic respiration (RA), heterotrophic respiration (RH), and total belowground carbon flux (TBCF), a measure of belowground allocation. Both study designs were organized in a 2x2x3 factorial design of intensity x family x position of SR measurement (bed, inter bed, and root exclusion), and used family block plots. Although extensive research exists regarding the effects of fer tilization on SR, few studies have examined genotypic effects on SR and belowground allocation, especially in forests after canopy closure. In Chapter 2, the temporal effects of SR and root carbon (C) mass were examined at Austin Cary Forest (ACF) for thr ee years and at Sanderson for two and a half years. At both sites, RH was examined for one year and eight months. Management and genotype affected loblolly pine carbon accumulation at both sites. The high intensity management treatment accumulated more abo veground C in loblolly pine components than the operational management treatment over the years of study. Similarly, the fast growing family accumulated more aboveground C than the slow growing family. Temporal variation in SR was evident across treatment s, with less SR in the winter than the summer months. This variation in SR could partially be accounted for by

PAGE 142

142 soil temperature, which was a major driver of SR. At ACF, for all positions of measurement, soil temperature accounted for approximately 25% of t he variation in SR. At Sanderson, soil temperature accounted for 49% of the variation in SR in the bed and inter bed positions, and 30% in the bed and root exclusion positions. Although soil moisture is known to be a significant correlate of SR, it played less of a role than soil temperature, with less than ~10% of variation accounted for by soil moisture. Although the soils are poorly drained, the A and E horizons have rapid to moderate permeability, which c ould regulate the soil moisture at the surface. Several major significant factors were identified through time at both sites. In the comparison of bed and inter bed positions at both sites, the bed position had greater SR rates than the inter bed position through time. The high intensity management tre atment was associated with suppressed SR rates when compared to the operational family h since the root exclusion position accounted only for RH and the decomposition of roots. In the comparison of bed and root exclusion positions, RH was found to be approximately 80% of total SR at the ACF and approximately 88% at Sanderson. Although some studies have found that fertilization suppresses SR and RH, no difference in RH were found betwee n management intensities. However, at Sanderson, an interaction between position and intensity showed that under the operational

PAGE 143

143 treatment, there was more RA than in the high intensity treatment. This was in line with the finding of greater root C mass (1 2mm) in the operational treatment at Sanderson. Through time, the fast growing family had greater SR rates than the slow growing family. However, this was not due to greater fine root C mass, as the slow growing family accumulated a greater amount of fine root C than the fast growing family. Other studies have found a correlation with stand basal area, but there was no association between SR and NPP in th is stud y (Luan et al., 2011). In Chapter 3, the modeled annual estimates of SR and TBCF at both sites were studied over three years at ACF and over two years at Sanderson. At both sites, RH was examined for one year. At ACF the SR models accounted for the following significant factors (p < 0.05 ): soil temperature, relative humidity, the difference of rela tive humidity from the daily mean, and the difference of the air temperature from the daily mean. At Sanderson the SR models accounted for the following significant factors (p < 0.05 ): air temperature, radiation, difference of air temperature from the dail y mean, and the difference of radiation from the daily mean. Annual values of litterfall and annual estimates of SR were integrated to develop an estimate of TBCF. NPP was estimated as the amount of C assimilated in the tree components: needles, branches, bole, and coarse roots. At ACF, NPP varied with management intensity and family, with the high intensity treatment having a greater NPP than the operational treatment (8.4 MgC.ha 1 versus 5.6 MgC.ha 1 ). The fast growing family also had a greater NPP than the slow growing family (8. 6 MgC.ha 1 versus 5.4 MgC.ha 1 ). At Sanderson, NPP only changed with management intensity, with the high intensity treatment having a greater NPP than the

PAGE 144

144 operational treatment (1 0.9 MgC.ha 1 versus 7. 5 MgC.ha 1 ). This result rai ses the (9. 0 MgC.ha 1 versus 9. 4 MgC.ha 1 ) ? Despite this, Roth et al. (2007) showed for Sanderson that differences in standing crop biomass did exist between these two families at ag e five. In both experiments, the most notable interaction was with intensity x family, which was significant for annual SR and TBCF. Under the high intensity treatment, there was no difference between the two families for annual SR. than the slow grower. At Sanderson, t he intensity x family interaction resulted in differences between families under both the operational and high intensity management systems. A great er difference between the two families was found under the operational intensity treatment under the operational treatment. In that case, the fast grower had a greater TB CF than the slow grower (18.3 MgC.ha 1 y 1 versus 16.3 MgC.ha 1 y 1 ). However, at ACF, under the high intensity treatment, the families behaved similarly (approximately 15.0 MgC.ha 1 y 1 ). At Sanderson, TBCF followed the same pattern as ACF. Under the operat ional treatment, the fast grower had a greater TBCF (17.7 MgC.ha 1 y 1 ) than the slow grower (13.8 MgC.ha 1 y 1 ) . Therefore, at both sites, for both families, TBCF was greater under the operational management regime . This is in line with another study that d emonstrated a decrease in belowground C cycling with fertiliz er applications (Giardina et al., 2004). Despite the connection between increased SR and the fast growing family, this study indicated that there was no significant connection between NPP and SR or TBCF.

PAGE 145

145 The lack of relationship between NPP and SR or TBCF could be linked to the sources of C being utilized in SR since the RH comprised approximately 80% of the total SR across treatments. This suggests that the majority of the SR may not be solely d ependent on aboveground production, but on the contribution of RH. However, it should be noted that root exclusions that are impenetrable to moisture have known drawbacks, such as increased soil moisture within the root exclusion, which could artificially increase microbial activity, and thereby increasing RH. This research suggests the need to incorporate genotype x environment interactions into models of SR in managed forest systems. The high intensity management treatment resulted in less SR than the op erational treatment. However, this reduction in SR due to intensity was not caused by a decrease in RH. This study also provide d an understanding o f the effects of forest management intensity and genotype deployment on SR and belowground C allocation. Important q uestions still remain , however. For example, does root production decrease with a mid rotation application of fertilizer? Sampling after fertilization at the ACF site would have provided greater clarity on this question relative to the two gen otypes being studied . In addition, examining the sources of C for heterotrophic organisms responsible for RH could be inform ative on how plant belowground allocation affects RH. Directed research could focus on fine root turnover, root exudates, and determ ining the sources of the C being respired. Studies exploring these processes could better inform our understanding of those factors controlling SR and RH. Results from this study demonstrated that a decrease in SR and TBCF for loblolly pine was associated with the high intensity management regime, but was not

PAGE 146

146 dependent on the genotype being deployed. This study was conducted after canopy closure and the long term data collected in family block plots, added strength to t he results of this investigation. Thus, these estimates may be applicable to comparable silvicultural treatments when applied on similar soil types . This research further informs on is not necessarily predictable with aboveground production. More research is required to examine the C balance in these systems, as SR and TBCF are only two components of the C cycle in these managed forest ed ecosystems.

PAGE 147

147 LIST OF REFERENCES Adegbidi, H.G., E.J. Jokela, N.B. Comerford, and N.F. Barros. 2002. Biomass development for intensively managed loblolly pine plantations growing on Spodosols in the southeastern USA. For . Ecol . and Manage . 167(1 3): 91 102. Adegbidi, H.G., N.B. Comerford, E.J. Joke la, and N.F. Barros. 2004. Root Development of Young Loblolly Pine in Spodosols in Southeast Georgia. Soil Sci. Soc. Am. J. 68(2): 596. Albaugh, T.J., H.L. Allen, P.M. Dougherty, L.W. Kress, and J.S. King. 1998. Leaf Area and Above and Belowground Growth Responses of Loblolly Pine to Nutrient and Water Additions. For. Sci. 44(2): 317 328. Albaugh, T.J., H.L. Allen, and L.W. Kress. 2006. Root and stem partitioning of Pinus taeda. Trees 20(2): 176 185. Albaugh, T.J., H.L. Allen, and T.R. Fox. 2007. Historic al Patterns of Forest Fertilization in the Southeastern United States from 1969 to 2004. South. J. Appl. For. 31(3): 129 137. Allen, H.L., 1987. Forest fertilizers: nutrient amendment, stand productivity, and environmental impact. J. For. 85, 37 46. Amateis, R.L., J. Liu, M.J. Ducey, and H.L. Allen. 2000. Modeling Response to Midrotation Nitrogen and Phosphorus Fertilization in Loblolly Pine Plantations. South. J. Appl. For. 24(4): 207 212. Amishev, D.Y., and T.R. Fox. 2006. The effect of weed control and fertilization on survival and growth of four pine species in the Virginia Piedmont. For. Ecol. and Manage. 236(1): 93 101. Andrews, J.A., and W.H. Schlesinger. 2001. Soil CO2 dynamics, acidification, and chemical weathering in a temperate forest with experimental CO2 enrichment. Global Biogeochem . Cy . 15(1): 149 162. Blazier, M.A., T.C. Hennessey, and S. Deng. 2005. Effects of Fertilization and Vegetation Control on Microbial Biomass Carbon and Dehydrogenase Activity in a Juvenile Loblolly Pine Plantat ion. For. Sci. 51(5): 449 459. Bond Lamberty, B., D. Bronson, E. Bladyka, and S.T. Gower. 2011. A comparison of trenched plot techniques for partitioning soil respiration. Soil Biol . Biochem . 43(10): 2108 2114. Borders, B.E., and R.L. Bailey. 2001. Loblol ly Pine Pushing the Limits of Growth. South . J. Appl. For. 25(2): 69 74.

PAGE 148

148 Bowden, R.D., K.J. Nadelhoffer, R.D. Boone, J.M. Melillo, and J.B. Garrison. 1993. Contributions of aboveground litter, belowground litter, and root respiration to total soil respirat ion in a temperate mixed hardwood forest. Can. J. For. Res. 23(7): 1402 1407. Bracho, R., G. Starr, H.L. Gholz, T.A. Martin, W.P. Cropper, and H.W. Loescher. 2012. Controls on carbon dynamics by ecosystem structure and climate for southeastern US slash pin e plantations. Ecol . Monogr . 82(1): 101 128. Burkhart, H.E., and P.T. Sprinz. 1984. Compatible Cubic Volume and Basal Area Projection Equations for Thinned Old Field Loblolly Pine Plantations. For. Sci. 30(1): 86 93. Burton, A.J., J.M. Melillo, and S.D. F rey. 2008. Adjustment of Forest Ecosystem Root Respiration as Temperature Warms. J . Integr . Plant Biol . 50(11): 1467 1483. Burton, A.J., J.C. Jarvey, M.P. Jarvi, D.R. Zak, and K.S. Pregitzer. 2012. Chronic N deposition alters root respiration tissue N rela tionship in northern hardwood forests. Glob. Change Biol. 18(1): 258 266. Butnor, J.R., K.H. Johnsen, R. Oren, and G.G. Katul. 2003. Reduction of forest floor respiration by fertilization on both carbon dioxide enriched and reference 17 year old loblolly p ine stands. Glob . Change Biol . 9(6): 849 861. Cain, M.D., and W.F. Mann. 1980. Annual Brush Control Increases Early Growth of Loblolly Pine. South . J. Appl. For. 4(2): 67 70. Chen, G., Y. Yang, and D. Robinson. 2014. Allometric constraints on, and trade of fs in, belowground carbon allocation and their control of soil respiration across global forest ecosystems. Glob . Change . Biol . 20(5): 1674 1684. Chen, G., Yang, Y., Guo, J., Xie, J., Yang, Z., 2011. Relationships between carbon allocation and partitionin g of soil respiration across world mature forests. Plant Ecol . 212, 195 206. Chemidlin Prévost Bouré, N., K. Soudani, C. Damesin, D. Berveiller, J. C. Lata, and E. Dufrêne. 2010. Increase in aboveground fresh litter quantity over stimulates soil respiratio n in a temperate deciduous forest. Appl . Soil Ecol . 46(1): 26 34. Cisneros Dozal, L.M., S.E. Trumbore, and P.J. Hanson. 2007. Effect of moisture on leaf litter decomposition and its contribution to soil respiration in a temperate forest. J. Geophys. Res. 1 12(G1): G01013. Crow, S.E., K. Lajtha, T.R. Filley, C.W. Swanston, R.D. Bowden, and B.A. Caldwell. 2009. Sources of plant derived carbon and stability of organic matter in soil: implications for global change. Glob . Change Biol . 15(8): 2003 2019.

PAGE 149

149 Curiel Yu ste, J., D.D. Baldocchi, A. Gershenson, A. Goldstein, L. Misson, and S. Wong. 2007. Microbial soil respiration and its dependency on carbon inputs, soil temperature and moisture. Glob . Change Biol . 13(9): 2018 2035. Davidson, E.A., K. Savage, P. Bolstad, D .A. Clark, P.S. Curtis, D.S. Ellsworth, P.J. Hanson, B.E. Law, Y. Luo, K.S. Pregitzer, J.C. Randolph, and D. Zak. 2002. Belowground carbon allocation in forests estimated from litterfall and IRGA based soil respiration measurements. Agr . Forest Meteorol . 1 13(1 4): 39 51. Davidson, E.A., L.V. Verchot, J.H. Cattânio, I.L. Ackerman, and J.E.M. Carvalho. 2000. Effects of soil water content on soil respiration in forests and cattle pastures of eastern Amazonia. Biogeochemistry 48(1): 53 69. Davidson, E.A., and S .E. Trumbore. 1995. Gas diffusivity and production of CO2 in deep soils of the eastern Amazon. Tellus B . 47(5): 550 565. Davidson, E.A., E. Belk, and R.D. Boone. 1998. Soil water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest. Global Change Biology 4(2): 217 227. Dixon, R.K., A.M. Solomon, S. Brown, R.A. Houghton, M.C. Trexier, and J. Wisniewski. 1994. Carbon Pools and Flux of Global Forest Ecosystems. Science 263(5144): 185 190 . Echeverría, M.E., Markewitz, D., Morris, L.A., Hendrick, R.L., 2004. Soil organic matter fractions under managed pine plantations of the southeastern USA. Soil Sci. Soc. Am. J. 68, 950 958. Epron, D., M. Bahn, D. Derrien, F.A. Lattanzi, J. Pumpanen, A. Gessler, P. Högberg, P. Maillard, M. Dannoura, D. Gérant, and N. Buchmann. 2012. Pulse labelling trees to study carbon allocation dynamics: a review of methods, current knowledge and future prospects. Tree Phys . 32(6): 776 798. Ewel, K.C., W.P. Cropper.Jr. , and H.L. Gholz. 1987. Soil CO2 evolution in Florida slash pine plantations. II. Importance of root respiration. Can. J. For. Res. 17(4): 330 333. Finzi, A.C., A.S. Allen, E.H. DeLucia, D.S. Ellsworth, and W.H. Schlesinger. 2001. Forest litter production, chemistry, and decomposition following two years of free air CO2 enrichment. Ecology 82(2): 470 484. Fox, T.R., E.J. Jokela, and H.L. Allen. 2007. The development of pine plantation silviculture in the southern United States. J. For. 105(7): 337 347. Gallardo, A., and W.H. Schlesinger. 1994. Factors limiting microbial biomass in the mineral soil and forest floor of a warm temperate forest. Soil Biol . Biochem . 26(10): 1409 1415.

PAGE 150

150 Giardina, C.P., D. Binkley, M.G. Ryan, J.H. Fownes, and R.S. Senock. 2004. Belowground carbon cycling in a humid tropical forest decreases with fertilization. Oecologia 139(4): 545 550. Giardina, C.P., and M.G. Ryan. 2002. Total Belowground Carbon Allocation in a Fast growing Eucalyptus Plantation Estimated Using a Carbon Balance Approach. Ecosystems 5(5): 487 499. Giardina, C.P., M.D. Coleman, J.E. Hancock, J.S. King, E.A. Lilleskov, W.M. Loya, K.S. Pregitzer, M.G. Ryan, and C.C. Trettin. 2005. The Response of Belowground Carbon Allocation in Forests to Global Change. p. 119 154. In Binkley, D., Menyailo, O. (eds.), Tree Species Effects on Soils: Implications for Global Change. Springer Netherlands. Gonzalez Benecke, C.A., E.J. Jokela, and T.A. Martin. 2012. Modeling the Effects of Stand Development, Site Quality, and Silviculture on Leaf Area Index, Litterfall, and Forest Floor Accumulations in Loblolly and Slash Pine Plantations. For. Sci. 58(5): 457 471. Gonzalez Benecke, C.A., T.A. Martin, E.J. Jokela, and R.D.L. Torre. 2011. A Flexible Hybrid Model of Life Cycle Carbon Balanc e for Loblolly Pine (Pinus taeda L.) Management Systems. Forests 2(4): 749 776. Gonzalez Benecke, C.A. , S.A. Gezan, T.J. Albaugh, H.L. Allen, H.E. Burkhart, E.J. Jokela, C.A. Maier, T.A. Martin, R.A. Rubilar and L.J. Samuelson. Local and general above stum p biomass functions for loblolly and slash pine trees. In review . Gough, C.M., and J.R. Seiler. 2004. Belowground carbon dynamics in loblolly pine (Pinus taeda) immediately following diammonium phosphate fertilization. Tree Physiol. 24(7): 845 851. Gower, S.T., S. Pongracic, and J.J. Landsberg. 1996. A Global Trend in Belowground Carbon Allocation: Can We Use the Relationship at Smaller Scales? Ecology 77(6): 1750 1755. Hanson, P.J., N.T. Edwards, C.T. Garten, and J.A. Andrews. 2000. Separating root and soi l microbial contributions to soil respiration: A review of methods and observations. Biogeochemistry 48(1): 115 146. Haynes, B.E., and S.T. Gower. 1995. Belowground carbon allocation in unfertilized and fertilized red pine plantations in northern Wisconsin . Tree Physiol 15(5): 317 325. Högberg, P., A. Nordgren, N. Buchmann, A.F.S. Taylor, A. Ekblad, M.N. Högberg, G. Nyberg, M. Ottosson Löfvenius, and D.J. Read. 2001. Large scale forest girdling shows that current photosynthesis drives soil respiration. Natu re 411(6839): 789 792.

PAGE 151

151 Holt, H., J. Voller, and J. Young. 1973. Vegetation control in newly established pine plantations. Southern Weed Science Society. Jandl, R., M. Lindner, L. Vesterdal, B. Bauwens, R. Baritz, F. Hagedorn, D.W. Johnson, K. Minkkinen, an d K.A. Byrne. 2007. How strongly can forest management influence soil carbon sequestration? Geoderma 137(3 4): 253 268. Johnson, D.W., J.D. Knoepp, W.T. Swank, J. Shan, L.A. Morris, D.H. Van Lear, and P.R. Kapeluck. 2002. Effects of forest management on s oil carbon: results of some long term resampling studies. Environ . Pollut . 116, Supplement 1: S201 S208. Jokela, E.J., P.M. Dougherty, and T.A. Martin. 2004. Production dynamics of intensively managed loblolly pine stands in the southern United States: a s ynthesis of seven long term experiments. For. Ecol. and Manage. 192(1): 117 130. Kim, C., J. Jeong, N. S. Bolan, and R. Naidu. 2012. Short term effects of fertilizer application on soil respiration in red pine stands. J . Ecol . and F ield B iol . 35:307 311. Lal, R. 2008. Carbon sequestration. Phil. Trans. R. Soc. B 363(1492): 815 830. Lee, K. H., and S. Jose. 2003. Soil respiration, fine root production, and microbial biomass in cottonwood and loblolly pine plantations along a nitrogen fertilization gradient . For. Ecol. and Manage. 185(3): 263 273. Litton, C.M., J.W. Raich, and M.G. Ryan. 2007. Carbon allocation in forest ecosystems. Glob . Change Biol . 13(10): 2089 2109. Lloyd, J., and J.A. Taylor. 1994. On the Temperature Dependence of Soil Respiration. Functional Ecology 8(3): 315 323. Luo, Y. 2007. Terrestrial Carbon Cycle Feedback to Climate Warming. Annu . Rev . of Ecol . Evol . S . 38: 683 712. Luo, Y. and X. Zhou. 2006. Soil Respiration and the Environment. Academic Press. Maier, C.A., and L.W. Kress. 20 00. Soil CO2 evolution and root respiration in 11 year old loblolly pine (Pinus taeda) plantations as affected by moisture and nutrient availability. Can. J. For. Res. 30(3): 347 359. Major, J.E., K.H. Johnsen, D.C. Barsi, and M. Campbell. 2012. Total belo wground carbon and nitrogen partitioning of mature black spruce displaying genetic × soil moisture interaction in growth. Can. J. For. Res. 42(11): 1939 1952. McKeand, S., T. Mullin, T. Byram, and T. White. 2003. Deployment of Genetically Improved Loblolly and Slash Pines in the South. J. For. 101(3): 32 37.

PAGE 152

152 McKeand, S.E., E.J. Jokela, D.A. Huber, T.D. Byram, H.L. Allen, B. Li, and T.J. Mullin. 2006. Performance of improved genotypes of loblolly pine across different soils, climates, and silvicultural input s. For. Ecol. and Manage. 227(1 2): 178 184. McKinley, D.C., M.G. Ryan, R.A. Birdsey, C.P. Giardina, M.E. Harmon, L.S. Heath, R.A. Houghton, R.B. Jackson, J.F. Morrison, and B.C. Murray. 2011. A synthesis of current knowledge on forests and carbon storage in the United States. Ecol . Appl . 21(6): 1902 1924. Melillo, J.M., J.D. Aber, and J.F. Muratore. 1982. Nitrogen and Lignin Control of Hardwood Leaf Litter Decomposition Dynamics. Ecology 63(3): 621 626. Minogue, P.J., Cantrell, R.L., Griswold, H.C., 1991. Vegetation management after plantation establishment. In: Duryea, M.L., Dougherty, P.M. (eds.) Forest regeneration manual. Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 335 358. Mickler, R.A., J.E. Smith, and L.S. Heath. Forest carbon trends in the Southern United States. Available at http://www.treesearch.fs.fed.us/pubs/viewpub.jsp?index=9739 (verified 22 June 2014). Moyano, F.E., S. Manzoni, and C. Chenu. 2013. Respo nses of soil heterotrophic respiration to moisture availability: An exploration of processes and models. Soil Biol . Biochem . 59: 72 85. Nave, L.E., E.D. Vance, C.W. Swanston, and P.S. Curtis. 2010. Harvest impacts on soil carbon storage in temperate forest s. For. Ecol. and Manage. 259(5): 857 866. Neff, J.C., A.R. Townsend, G. Gleixner, S.J. Lehman, J. Turnbull, and W.D. Bowman. 2002. Variable effects of nitrogen additions on the stability and turnover of soil carbon. Nature 419(6910): 915 917. Nelson, L.R. , R.C. Pedersen, L.L. Autry, S. Dudley, and J.D. Walstad. 1981. Impacts of Herbaceous Weeds in Young Loblolly Pine Plantations. South. J. Appl. For. 5(3): 153 158. Norby, R.J., and R.B. Jackson. 2000. Root dynamics and global change: seeking an ecosystem p erspective. New Phytol . 147(1): 3 12. Noormets, A., S.G. McNulty, J. C. Domec, M. Gavazzi, G. Sun, and J.S. King. 2012. The role of harvest residue in rotation cycle carbon balance in loblolly pine plantations. Respiration partitioning approach. Glob . Chan ge Biol . 18(10): 3186 3201. Oleksyn, J., P.B. Reich, W. Chalupka, and M.G. Tjoelker. 1999. Differential Above and Below ground Biomass Accumulation of European Pinus sylvestris Populations in a 12 year old Provenance Experiment. Scand . J . of Forest Res . 1 4(1): 7 17.

PAGE 153

153 Olsson, P., Linder, S., Giesler, R., Högberg, P., 2005. Fertilization of boreal forest reduces both autotrophic and heterotrophic soil respiration. Glob . Change Biol . 11, 1745 1753. Palmroth, S., R. Oren, H.R. McCarthy, K.H. Johnsen, A.C. Fi nzi, J.R. Butnor, M.G. Ryan, and W.H. Schlesinger. 2006. Aboveground sink strength in forests controls the allocation of carbon below ground and its [CO2] induced enhancement. PNAS 103(51): 19362 19367. Pangle, R.E., and J. Seiler. 2002. Influence of seedl ing roots, environmental factors and soil characteristics on soil CO2 efflux rates in a 2 year old loblolly pine (Pinus taeda L.) plantation in the Virginia Piedmont. Environ. Pollut. 116 Suppl 1: S85 96. Parker, W.C., Pitt, D.G., Morneault, A.E., 2009. In fluence of woody and herbaceous competition on microclimate and growth of eastern white pine (Pinus strobus L.) seedlings planted in a central Ontario clearcut. For. Ecol. and Manage. 258, 2013 2025. Post, W.M., and K.C. Kwon. 2000. Soil carbon sequestrati on and land use change: processes and potential. Glob . Change Biol . 6(3): 317 327. Post, W.M., W.R. Emanuel, P.J. Zinke, and A.G. Stangenberger. 1982. Soil carbon pools and world life zones. Nature 298(5870): 156 159. Raich, J.W., Potter, C.S., Bhagawati, D., 2002. Interannual variability in global soil respiration, 1980 94. Glob . Change Biol . 8, 800 812. Raich, J.W., and K.J. Nadelhoffer. 1989. Belowground Carbon Allocation in Forest Ecosystems: Global Trends. Ecology 70(5): 1346 1354. Raich, J.W., and W. H. Schlesinger. 1992. The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus B . 44(2): 81 99. Raich, J.W., and A. Tufekciogul. 2000. Vegetation and soil respiration: Correlations and controls. Biogeochemis try 48(1): 71 90. Ratcliff, A.W., Busse, M.D., Shestak, C.J., 2006. Changes in microbial community structure following herbicide (glyphosate) additions to forest soils. Appl . Soil Ecol . 34, 114 124. d, and M.A. Topa. 2001. Whole tree biomass and carbon allocation of juvenile trees of loblolly pine: influence of genetics and fertilization. Can. J. For. Res. 31(6): 960 970. Richter, D.D., J.M. Kelly, W.M. McFee, D. Markewitz, J.K. Dunsomb, P.R. Heine, C.G. Wells, A. Stuanes, H.L. Allen, B. Urrego, K. Harrison, and G. Bonani. 1995. Carbon Cycling in a Loblolly Pine Forest: Implications for the Missing Carbon

PAGE 154

154 Sink and for the C oncept of Soil. In ACSESS publications. Soil Science Society of America. Roth, B.E., X. Li, D.A. Huber, and G.F. Peter. 2007. Effects of management intensity, genetics and planting density on wood stiffness in a plantation of juvenile loblolly pine in the southeastern USA. For. Ecol. and Manage. 246(2 3): 155 162. Rustad, L.E., Huntington, T.G., Boone, R.D., 2000. Controls on soil respiration: implications for climate change. Biogeochemistry 48, 1 6. R yan, M.G., and B.E. Law. 2005. Interpreting, measuring, and modeling soil respiration. Biogeochemistry 73(1): 3 27. Samuelson, L.J., K. Johnsen, T. Stokes, and W. Lu. 2004. Intensive management modifies soil CO2 efflux in 6 year old Pinus taeda L. stands. For. Ecol. and Manage. 200(1 3): 335 345. Samuelson, L. J., and W.B. Whitaker. 2012. Relationships between Soil CO2 Efflux and Forest Structure in 50 Year Old Longleaf Pine. For. Sci. 58(5): 472 484. Samuelson, L., R. Mathew, T. Stokes, Y. Feng, D. Aubrey, and M. Coleman. 2009. Soil and microbial respiration in a loblolly pine plantation in response to seven years of irrigation and fertilization. For. Ecol. and Manage. 258(11): 2431 2438. Sarkhot, D.V. , E.J. Jokela, and N.B. Comerford. 2008. Surface soil carbon size density fractions altered by loblolly pine families and forest management intensity for a Spodosol in the southeastern US. Plant Soil 307(1 2): 99 111. Sarkhot, D.V., Comerford, N.B., Jokel a, E.J., Reeves, J.B., 2007. Effects of forest management intensity on carbon and nitrogen content in different soil size fractions of a North Florida Spodosol. Plant S oil 294, 291 303. SAS Institute Inc., 2009. SAS OnlineDoc® 9.2. Cary, NC, USA. Schlesin ger, W.H., and J.A. Andrews. 2000. Soil respiration and the global carbon cycle. Biogeochemistry 48(1): 7 20. Schlesinger, W. H. 1997. Biogeochemistry : an analysis of global change. Academic Press, San Diego, Calif. Schmidt, M.W.I., M.S. Torn, S. Abiven, T. Dittmar, G. Guggenberger, I.A. Janssens, M. Kleber, I. Kögel Knabner, J. Lehmann, D.A.C. Manning, P. Nannipieri, D.P. Rasse, S. Weiner, and S.E. Trumbore. 2011. Persistence of soil organic matter as an ecosystem property. Nature 478(7367): 49 56. Schul tz, R.P. 1997. Loblolly pine: the ecology and culture of loblolly pine (Pinus taeda L.). (713): xiv + 494 pp.

PAGE 155

155 Shan, J., Morris, L.A., Hendrick, R.L., 2001. The effects of management on soil and plant carbon sequestration in slash pine plantations. J . Appl . Ecol . 38, 932 941. Smith, P., C. Fang, J.J.C. Dawson, and J.B. Moncrieff. 2008. Impact of Global Warming on Soil Organic Carbon. p. 1 43. In Donald L. Sparks (ed.), Adv . Agron . . Academic Press. Stovall, J.P., J.R. Seiler, and T.R. Fox. 2013. Respiratory C fluxes and root exudation differ in two full sib clones of Pinus taeda (L.) under contrasting fertilizer regimes in a greenhouse. Plant Soil 363(1 2): 257 271. Subke, J. A., I. Inglima, and M. Francesca Cotrufo. 2006. Trends and methodological impacts in soil CO2 efflux partitioning: A metaanalytical review. Glob. Change Biol. 12(6): 921 943. Subke, J. A., N.R. Voke, V. Leronni, M.H. Garnett, and P. Ineson. 2011. Dynamics and pathways of autotrophic and heterotrophic soil CO2 efflux revealed by forest gir dling. J . Ecol . 99(1): 186 193. Sulzman, E.W., J.B. Brant, R.D. Bowden, and K. Lajtha. 2005. Contribution of aboveground litter, belowground litter, and rhizosphere respiration to total soil CO2 efflux in an old growth coniferous forest. Biogeochemistry 73 (1): 231 256. Thirukkumaran, C.M., and D. Parkinson. 2000. Microbial respiration, biomass, metabolic quotient and litter decomposition in a lodgepole pine forest floor amended with nitrogen and phosphorous fertilizers. Soil Biol . Biochem . 32(1): 59 66. Torn , M.S., P.M. Vitousek, and S.E. Trumbore. 2005. The Influence of Nutrient Availability on Soil Organic Matter Turnover Estimated by Incubations and Radiocarbon Modeling. Ecosystems 8(4): 352 372. Treseder, K.K. 2004. A meta analysis of mycorrhizal response s to nitrogen, phosphorus, and atmospheric CO2 in field studies. New Phytol . 164(2): 347 355. Treseder, K.K. 2008. Nitrogen additions and microbial biomass: a meta analysis of ecosystem studies. Ecol . Lett . 11(10): 1111 1120. Trumbore, S. 2000. Age of Soil Organic Matter and Soil Respiration: Radiocarbon Constraints on Belowground C Dynamics. Ecol . Appl . 10(2): 399 411. Tyree, M.C., J.R. Seiler, W.M. Aust, D.A. Sampson, and T.R. Fox. 2006. Long term effects of site preparation and fertilization on total so il CO2 efflux and heterotrophic respiration in a 33 year old Pinus taeda L. plantation on the wet flats of the Virginia Lower Coastal Plain. For. Ecol. and Manage. 234(1 3): 363 369. Tyree, M.C., J.R. Seiler, and T.R. Fox. 2008. The Effects of Fertilizatio n on Soil Respiration in 2 Year Old Pinus taeda L. Clones. For. Sci. 54(1): 21 30.

PAGE 156

156 Tyree, M.C., J.R. Seiler, and C.A. Maier. 2014. Contrasting genotypes, soil amendments, and their interactive effects on short term total soil CO2 efflux in a 3 year old Pi nus taeda L. plantation. Soil Biol . Biochem . 69: 93 100. Vogel, J.G., L.J. Suau, T.A. Martin, and E.J. Jokela. 2011. Long term effects of weed control and fertilization on the carbon and nitrogen pools of a slash and loblolly pine forest in north central F lorida. Can. J. For. Res. 41(3): 552 567. Vogel, J.G., B.P. Bond Lamberty, E. a. G. Schuur, S.T. Gower, M.C. Mack, K.E.B. black spruce forests across regions varying in soil temp erature and precipitation. Glob. Change Biol. 14(7): 1503 1516. Wear, D.N., and J.G. Greis. 2013. The Southern Forest Futures Project: technical report. 178: 1 542. Zobel, B. and Talbert, J. 1984. Applied Forest Tree Improvement. John Wiley & Sons, NY.

PAGE 157

157 BIOGRAPHICAL SKETCH Chelsea Gill Drum was born in Santa Cruz, California in 1987, but grew up on Galveston Island, Texas. While visiting her grandparents on the Olympic Peninsula in Washington State, she developed an appreciation for the woods at a young age. After graduating from high school, Chelsea moved to Seattle, Washington to pursue an undergraduate degree in 2006. In 2010, Chelsea received her Bachelor of Science degree in f orest r esources from the University of Washington at Seattle. In 2011, she joined the School of Forest Resources and Conservation at the University of Florida to pursue her m degree. Chelsea is married to Michael Drum of Sequim, Washington. They currently reside in Gainesville, Florida, with their chocolate Labrador , Teddy. They have plans to move to Blacksburg, Virginia for Chelsea to pursue a PhD at Virginia Tech.