HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Burner, D. M. Articles by Belesky, D. P. Search for Related Content PubMed Articles by Burner, D. M. Articles by Belesky, D. P. Agricola Articles by Burner, D. M. Articles by Belesky, D. P. Related Collections Other Forage Crops Agroforestry Forage quality Crop Systems Plant and Environment Interactions Tillage

Effect of Loblolly Pine Root Pruning on Alley Cropped Herbage Production and Tree Growth

^a Dale Bumpers Small Farms Research Center, USDA-ARS, 6883 S. State Hwy. 23, Booneville, AR 72927
^b USDA-ARS, Appalachian Farming Systems Research Center, 1224 Airport Rd., Beaver, WV 25813

^* Corresponding author (David.Burner{at}ars.usda.gov).

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Tillage to disrupt (prune) tree roots is an intensive practice which could improve herbage productivity at the crop–tree interface by reducing competition for water. We compared tillage effects on 9- to 11-yr-old loblolly pine (Pinus taeda L.) growth and herbage yields of annual ryegrass (Lolium multiflorum Lam.) and pearl millet [Pennisetum glaucum (L.) R. Br.] on a fragipan soil in Arkansas. Alley crops were rotationally grown in a 9.7-m wide alley (main plot) between bordering trees on one of three tillage treatments: control (surface tillage), rip followed by surface tillage, and trench plus root barrier followed by surface tillage. Topsoil water in May through September, herbage mass, and nutritive value were measured for each crop for 2 or 3 yr in three subplots systematically arrayed (north, middle, and south) across the alley. Diameter at breast height (DBH, measured 1.3 m above soil surface) and height of border trees were measured annually. Trenching resulted in a more uniform distribution of topsoil water among subplots compared to the other tillage treatments. Annual ryegrass yield did not show a tillage response, but pearl millet yielded more herbage in the rip (6760 kg ha^–1 in 2003) and trench (3300 kg ha^–1 in 2005) than the control treatment (4990 and 1260 kg ha^–1 for 2003 and 2005, respectively). Ripping and trenching significantly reduced loblolly pine DBH and height compared to the control. Similarly configured alley cropping practices probably have little potential for annual herbage production even with root pruning.

Abbreviations: DBH, diameter breast height • IVDMD, in vitro dry matter digestibility • PAR, photosynthetically active radiation • TNC, total nonstructural carbohydrates

Received for publication May 28, 2008.

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
NUMEROUS SPATIAL and temporal factors associated with plantation tree-row spacing impact forage management considerations in alley cropping practices. The functional width for alley cropping decreases with time as trees become increasingly greater sinks for solar radiation, soil water, and soil nutrients. Plantation alleys 4.9 m wide in rectangular configurations of loblolly pine are minimally acceptable for equipment access and herbage production 5 to 6 yr after tree planting (Burner and Brauer, 2003; Pearson et al., 1995). In Australia, herbage biomass decreased significantly <4 yr after planting rose gum (Eucalyptus grandis W. Hill ex Maid.) at stand densities >300 stems ha^–1 (Cameron et al., 1989). Sward botanical composition also changes with time in tree alleys. Yield of a mixed species, warm-season grass meadow increased quadratically with alley width in a loblolly pine plantation (Burner and Brauer, 2003). The percentage of bermudagrass (Cynodon dactylon L.) in a mixed species sward was inversely related to loblolly pine canopy cover (Brauer et al., 2004), but tree density had little effect on herbaceous species composition of a botanically complex warm-season sward (Cameron et al., 1989).

An understanding of the spatial extent of the competitive zone aids in the design of alley crop plantations. Microenvironmental alley modifications induced by the tree crop vary across short spatial scales, creating an edge effect that affects alley crop productivity (Ssekabembe et al., 1997). In Australia, the competition zone for maritime pine (P. pinaster Soland., non Ait.) extends about 2.5 times the tree height (H), and cropping is not economical within 1H of the tree (Sudmeyer et al., 2002). If the same constraints applied to loblolly pine practices, crop production would be uneconomical in a 10-m wide alley when tree height is 5 m. Therefore, silvopasture could be an alternative land use with livestock grazed on some perennial herbage providing short-term income, and wood fiber providing long-term income (Husak and Grado, 2002).

Competition for water often constrains crop yields more than competition for light at the tree–crop interface, suggesting that the economical area of crop production could be extended by pruning the tree roots (Jose et al., 2000; Gillespie et al., 2000; Sudmeyer et al., 2002). Similarly, the effective canopy cover of a pearl millet agroforestry practice in an arid region could be increased with tree root pruning (Payne, 2000). Conversely, low solar irradiance under a loblolly pine canopy was a greater constraint than low soil water to tall fescue herbage specific leaf weight, leaf extension rate, tillers plant^–1, mass tiller^–1, mass plant^–1, and total nonstructural carbohydrate (TNC) concentration (Burner and Belesky, 2008), suggesting that herbage productivity might benefit little from tree root pruning. It might be difficult to rate the relative magnitudes of shade and soil water constraints in any given agroforestry practice because of confounding site-specific factors such as tree species and management, crop species and management, annual rainfall amount and distribution, and soil type, depth, and drainage.

The Environmental Quality Incentives Program (EQIP) of 2002 Farm Bills authorized a cost share incentive of U.S.$5.00 per 30 m of root pruning along tree and shrub rows and woodland edges associated with cropland and grassland fields (USDA-Natural Resources Conservation Service, 2002). This cost presumably could be achieved by ripping (or subsoiling), an agricultural practice commonly used for disrupting fragipans (Tyler and McCutchen, 1980). Trench-root barriers and irrigation are costlier alternatives to ripping, justifiable only for high value alley crops. Eliminating tree roots from the alley by root barriers was economical only after 5 to 6 yr of continuous row crop and fodder production on degraded soils in a temperate region of India (Dadhwal and Tomar, 1999). Ripping could emulate the trench-root barrier by eliminating crop–tree root competition, although its temporal effectiveness would be limited by the time required for tree roots to recolonize the alley. Passive approaches to improving alley crop productivity at the crop–tree interface include using shade- and drought-tolerant herbage species (Burner, 2003; Lin et al., 1999), design (e.g., wide crop alleys), or cultural practices (e.g., tree pruning or thinning, and addition of any limiting growth factor). The goal of active mechanical or passive interventions is to reduce water stress and competition for soil nutrients or solar irradiance, thereby optimizing productivity of the tree–crop components of the system.

The objective of this experiment was to compare tillage effects on loblolly pine growth and rotational yields of annual ryegrass and pearl millet in an alley cropping practice. We hypothesized that disruption of tree roots at the crop–tree interface would improve herbage productivity more than it reduced tree growth. We did not examine the cost-benefit economics of these cultural interventions.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Site History and Description
The experiment was located near Booneville, AR, at 35° N, 94° W, and 130 m above sea level. In spring 1994, 1-yr-old loblolly pine seedlings were planted in east–west (70–80° from north) orientation (Fig. 1 ) in three multiple-row sets (hedgerows) per 0.4 ha replication (three tree rows with 1.2 and 2.4 m spacing within and between rows, respectively, and an alley of 9.7 m). There were three 9.7-m alleys per replication, and three replications. There were about 970 trees ha^–1 in January 2003. Alley herbage at initiation of the experiment was similar to that at tree planting: a complex mixture of cool- and warm-season grasses and forbs (Burner and Brauer, 2003). Before the experiment, alleys had been cultivated only once, in November 2000, when they received an amendment of 4.5 Mg lime ha^–1, 30 kg P ha^–1, and 30 kg K ha^–1 which was incorporated to the 15-cm depth.

View larger version (57K): [in this window] [in a new window]   Fig. 1. Arrangement of tillage treatments and subplots in one replication. Trees are indicated by symbols (daisy wheel). There were three tree rows sets per replication: three tree rows (1.8 m within row by 2.4 m between rows) and 9.7 m alley. The cultivated plot was 6.7 m wide and separated from northern- and southern-most tree rows by an uncultivated area 1.5 m wide. Horizontal lines indicate subplots (2.2 by 55 m, oriented east-west) designated as north, middle, and south. Symbols (empty circle/cross) indicate approximate location of photosynthetically active radiation (PAR) and topsoil temperature sensors.

Soil type and depth to friable shale (maximum tool length was 1.2 m) were assessed using 0.8-cm auger holes drilled at 6-m intervals along an east–west transect in the middle of each alley in September 2002. About 65% of the alley area in two replicates was Leadvale silt loam (fine-silty, siliceous, semiactive, thermic Typic Fragiudult), and about 35% was Enders silt loam (fine, mixed, active, thermic Typic Hapludult). The third replicate was entirely Leadvale silt loam. Alleys had negligible slope, ranging from 1.3 to 2.5%. Depth to friable shale was typically 1.2 m for the Leadvale soil (range 0.7 to 1.2 m), and 1 m (range 0.6 to 1.2 m) for the Enders soil.

Topsoil from September 2002 was analyzed for pH, and available P and K by the University of Arkansas Soil Test Laboratory. Initial topsoil conditions were pH 6.2, 5.7 µg available P g^–1, and 63 µg available K g^–1.

Tillage Treatments and Herbage Establishment
The arable region of the alley was 6.7 m wide by 55 m long, which excluded a 1.5 m wide region from the north and south border tree rows to the plot edge which remained undisturbed during the experiment. The 1.5 m exclusion zone was located at about the tree foliage drip line. Vegetation in the tree understory was predominantly tall fescue [Lolium arundinaceum (Schreb.) Darbysh.] and was undisturbed during the experiment.

Alleys received one of three tillage (main plot) treatments in January 2003: control (surface tillage with rototiller to the 15-cm depth sufficient to prepare a seed bed), rip (60 cm deep at 1.5-m intervals across the alley followed by surface tillage), and trench (13 cm wide by 1 m deep, fitted with a 6 mil polyethylene root barrier and backfilled) along north and south border tree rows followed by surface tillage. Tillage treatments were imposed no closer than the 1.5-m drip line from the tree stem. The root barrier extended above the soil surface and was undisturbed during the experiment. Plots were trenched only once. The rip treatment was reapplied in October 2004. Intensity of root pruning was slightly less severe than that of Miller and Pallardy (2001) for silver maple (Acer saccharinum L.), but more severe than that of Sudmeyer et al. (2002) for maritime pine.

Plots were surface tilled twice annually to prepare a seed bed for each crop. Subplots (55 m long) were assigned to each alley (main plot) based on location: north = northern strip of alley extending 0 to 2.2 m southwards from the northern plot edge, middle = central region of alley 2.2 to 4.5 m southwards from the northern plot edge, and south = southern strip of alley extending 4.5 to 6.7 m southwards from the northern plot edge.

Alleys were sown in May 2003, May 2004, and June 2005 with ‘Tifleaf 3’ pearl millet at about 40 kg seed ha^–1, and harvested either once (September 2003) or twice (August and October 2004 and 2005). The August harvest occurred when about 50% of plants were at anthesis; second harvests were at vegetative growth stage. Yield of pearl millet in 2004 and 2005 was the sum of August and October harvests. Alleys were sown in September 2003 and November 2004 with ‘Marshall’ annual ryegrass at about 30 kg seed ha^–1, and harvested once (at anthesis) from each planting (May 2004 and 2005). On any given planting date, seed was either broadcast or drilled depending on equipment availability. After planting, each grass crop was topdressed with a blended fertilizer mixture which supplied 56, 10, and 20 kg ha^–1 of N, P, and K, respectively.

Environmental Monitoring
Topsoil temperature, measured 15-cm below soil surface, and photosynthetically active radiation (PAR), measured 1.4 m above soil surface, were continuously recorded at 0.5 h intervals from 1 May through 30 Sept. 2004 and 2005 in the middle of three subplots of one tillage treatment (no spatial replication) to characterize the subplot microenvironment during the pearl millet crop. Topsoil temperature was measured with a Model 3667 external temperature probe, and PAR (400–700 nm) was measured with a Model 3668 quantum light sensor (Spectrum Technologies, Plainfield, IL). Sensors were moved only during planting and harvesting.

Air temperature, topsoil temperature 15 cm beneath grass sod, PAR, and rainfall were similarly recorded at a weather station (unshaded conditions) located 1.3 km northwest of the experimental site. Long-term (1971–2000) air temperature and rainfall data were from an official weather station (National Oceanic and Atmospheric Administration, 2002) located 7.0 km east of the experimental site. Long-term topsoil temperature was not locally available. Topsoil was sampled from each subplot at monthly intervals from August 2003 to September 2005 for measuring the concentration of gravimetric water (48 h at 95°C).

Diurnal shade dynamics were modeled using the procedure of Hooge (2005) to predict the alley area covered by shade. Input variables were latitude, longitude, month, day, time (Central Standard), and barrier (tree) height. The model was run for the 21st of each month to capture summer (July) and winter (December) solstices, but only data for May, July, and September were presented. Tree height was either 7.3 or 10.7 m (mean height of border trees in 2002 and 2005, respectively). The proportion of the alley in shade was constrained to a maximum 10 m wide alley width. A deficiency of the model was that it assumed an opaque barrier, while trees provided varying levels of both opaque and translucent shade, as well as sun flecks.

Plant Growth and Nutritive Value Measurements
Twenty codominant trees in each north and south border row (about every second tree) were permanently tagged and measured for stem DBH with a diameter tape. Height of every second tagged tree was measured with a clinometer. Tree DBH and height were measured in January 2003 and 2004, and December 2004 and 2005, representing the four consecutive tree growing seasons from 2002 to 2005.

Fresh herbage mass was measured by clipping a strip in the middle of each subplot (15 cm stubble height in a swath 0.9 by 55 m) with a flail harvester. Dry mass yield (kg ha^–1) was calculated from a sample dried 48 h at 60°C. Sampled herbage was ground to pass a 1 mm screen and stored at –20°C before nutritive value analysis. Only the first annual harvests of pearl millet herbage were analyzed for nutritive value. Herbage N was determined using a Carlo Erba EA1112 combustion analyzer (Thermo Electron Corp., Waltham, MA). Herbage crude protein was calculated from herbage N x 6.25. Herbage in vitro dry matter digestibility (IVDMD) was determined using the procedure of Goering and Van Soest (1970), modified for the Ankom Daisy II fiber analyzer no. F200 (ANKOM Technology Corp., Fairport, NY). Rumen fluid for the IVDMD procedure was collected from a ruminally fistulated steer (Bos taurus L.) fed a diet of bermudagrass hay ad libitum and 0.5 kg d^–1 soybean (Glycine max L.) meal. Total nonstructural carbohydrate (TNC) was determined by the method of Smith (1981) as modified by Denison et al. (1990). Nutritive value was not determined for pearl millet harvested in September 2003.

Statistical Analysis
Air temperature and rainfall data were not spatially or temporally replicated at the experimental site, so data were presented as monthly means. Diurnal means and standard errors for subplot topsoil temperature and PAR were calculated across days and years. For PAR, only daylight hours were included in the analysis.

Analyses of variance of spatially and temporally replicated data (topsoil water, tree DBH and height, and herbage yield and nutritive value) were conducted using a mixed linear model, Proc Mixed (Littell et al., 1996; SAS Institute, 2002). Random effects were replication and its interactions with fixed effects. Sample date, or year, was analyzed as a repeated measure with an autoregressive order 1 covariance structure (Littell et al., 1996). Fixed effects for topsoil water were year (2 df), tillage (2 df), subplot within tillage (6 df), and the interactions year x tillage (4 df) and year x subplot within tillage (12 df). For tree DBH and height, fixed effects were year (3 df), tillage (2 df), and year x tillage (6 df). For annual ryegrass yield and nutritive value, fixed effects were year (1 df), tillage (2 df), subplot within tillage (6 df), year x tillage (2 df), and year x subplot within tillage (6 df). For pearl millet yield, fixed effects were year (2 df), tillage (2 df), subplot within tillage (6 df), year x tillage (4 df), and year x subplot within tillage (12 df). For pearl millet nutritive value, fixed effects were year (1 df), tillage (2 df), subplot within tillage (6 df), year x tillage (2 df), and year x subplot within tillage (6 df). Denominator df were calculated by a general Satterthwaite approximation method (Littell et al., 1996). Means were separated using the Tukey HSD test at P 0.05 using (SAS Institute, 2002).

	RESULTS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Mean monthly air temperatures in 2005 were roughly comparable to the 30-yr (long-term) mean, while 2004 tended to be slightly cooler (Fig. 2A ). Topsoil temperatures tended to be warmer from May through July 2004, but cooler in August, than in 2005 (Fig. 2B). Rainfall was variable in 2004 and 2005 and did not closely approximate the long-term mean either year except during August (Fig. 2C), a normally dry month. Rainfall tended to be greater than the long-term mean during June and July 2004 and September 2005, while other months within the year tended to be dryer than the long-term mean. May through August 2005 was unseasonably dry.

View larger version (14K): [in this window] [in a new window]   Fig. 2. Mean monthly climatic conditions at a weather station located 1.3 km from the experimental site in 2004 (empty circle) and 2005 (empty square). The long-term mean (filled circle) for air temperature (A) and rainfall (C) was for the period 1971 to 2000 mean from a station located 7.0 km east at Booneville, AR (National Oceanic and Atmospheric Administration, 2002).

View larger version (34K): [in this window] [in a new window]   Fig. 3. Mean diurnal soil temperature at 15 cm depth in north (empty triangle), middle (empty square), and south (filled triangle) alley subplots and at an unshaded weather station located 1.3 km from the experimental site (empty circle) during May through September 2004 and 2005.

View larger version (22K): [in this window] [in a new window]   Fig. 4. Mean diurnal photosynthetically active radiation (PAR) in north (empty triangle), middle (empty square), and south (filled triangle) alley subplots and at an unshaded weather station located 1.3 km from the experimental site (empty circle) during May through September 2004 and 2005.

View larger version (14K): [in this window] [in a new window]   Fig. 5. Diurnal shading of a 10 m-wide alley by an opaque barrier 7.3 or 10.7 m tall (mean height of bordering trees in 2002 and 2005, respectively) on 21 May, 21 July, and 21 Sept. 2002 (empty symbols) and 2005 (filled symbols) at Booneville, AR according to the model of Hooge (2005).

View larger version (16K): [in this window] [in a new window]   Fig. 6. Effect of subplot (north, middle, and south) within tillage treatment on mean topsoil water content of loblolly pine alleys. Vertical bars are the standard error of the mean (n = 78). Bars within a tillage treatment having a common letter do not differ (P > 0.05).

View larger version (15K): [in this window] [in a new window]   Fig. 7. (A) Effect of year, and (B) effect of tillage on diameter breast height (DBH) and height of loblolly pine. Vertical bars are the standard error of the mean (A: n = 360 and 180, and B: n = 480 and 240 for DBH and height, respectively). Bars within a figure followed by a common capital (DBH) or lowercase letter (height) do not differ (P > 0.05).

The middle subplot yielded more annual ryegrass herbage than the north or south subplots in the control in 2004 (Fig. 8 ), suggesting an edge effect of trees on herbage yield. Yield in 2005 ranged from 519 to 708 kg ha^–1. There were no other subplot within tillage differences (P 0.30) in 2004 or 2005.

View larger version (14K): [in this window] [in a new window]   Fig. 8. Effect of year x subplot within tillage treatment on herbage yield of annual ryegrass. Bars within a year and tillage treatment followed by a common letter do not differ (P > 0.05).

View larger version (15K): [in this window] [in a new window]   Fig. 9. Effect of year x subplot within tillage treatment on herbage yield of pearl millet. Bars within a year and tillage treatment followed by a common letter do not differ (P > 0.05).

Pearl millet herbage had less crude protein with trenching (104 g kg^–1) than the control (158 g kg^–1) and rip (139 g kg^–1) treatments in 2005, but crude protein was unaffected (P 0.99) by tillage in 2004 (72–78 g kg^–1). Herbage also had less IVDMD with trenching (573 g kg^–1) than control (648 g kg^–1) and rip (639 g kg^–1) treatments in 2005, but IVDMD was unaffected (P 0.99) by tillage in 2004 (579–582 g kg^–1). There was a year x subplot within tillage effect for herbage crude protein caused by large annual differences (75 and 134 g kg^–1 for 2004 and 2005, respectively). The subplot within tillage effect showed that the middle subplot had less crude protein (90 g kg^–1) than the south subplot (132 g kg^–1) in the control tillage, and less than that of the north subplot with ripping (77 and 132 g kg^–1, respectively). There were no differences (P 0.96) in pearl millet crude protein between subplots in the trench treatment (85 to 99 g kg^–1). Pearl millet TNC differed between year (74 and 64 g kg^–1 for 2004 and 2005, respectively). For any given tillage treatment, herbage had about 30 g kg^–1 less TNC in the south than the north subplot, but the difference was not significant (P 0.15).

	DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Trees exert a competitive root zone (edge effect) which, under some situations, can severely limit alley crop production (Sudmeyer et al., 2002). Factors such as tree species, age, height, alley crop species, soil, and climate likely affect the degree of adverse impact. We studied effects of root pruning on herbage and tree growth in a loblolly pine plantation to determine if disruption of tree roots at the crop–tree interface could improve rotational herbage productivity more than it reduced tree growth.

There was obvious root disruption on initial application of the deep tillage treatments. Numerous tree roots were severed and brought to the surface during ripping, especially by the ripping tine closest (1.5 m) to the border tree rows. There also were numerous roots severed during trenching. Taproots of 27-yr-old loblolly pine can grow to a depth of 1 to 2 m, and lateral roots of 7-yr-old trees are concentrated within a zone 15 to 20 cm below the soil surface and about 1.2 m from the stem (Schultz, 1997). Lateral feeder roots can emerge from the taproot at depths up to 1 m and extend >4.9 m from the stem. Few roots appeared to be damaged during surface tillage of the alleys, and few tree roots were observed in the alley middle during ripping or surface tillage.

We found that pruning the tree roots 1.5 m along one side of the loblolly pine stem (at the tree drip line) significantly reduced DBH and height compared to the control (Fig. 7). Either root pruning method caused tree DBH to be less than that of the control. Trenching had a greater impact on tree height than ripping, perhaps because the ripping tines were shallower (0.6 m) than the trench (1 m). Also, roots could recolonize the alley after ripping, allowing some recovery from damage, while the trench root barrier prevented root recolonization of the alley. Root pruning reduced DBH growth (height was not measured) of silver maple (Miller and Pallardy, 2001), but not height or DBH growth of maritime pine (Sudmeyer et al., 2002), perhaps because roots were pruned at about 1 and 5 m from the tree stem, respectively.

This soil tends to have low concentrations of available soil N and P, and a fragipan at the 40- to 60-cm depth seems relatively impervious to root penetration (Burner and MacKown, 2005). The fragipan could intensify the deleterious effect of root pruning, especially during periods of low rainfall (Fig. 2), if it caused the tree roots to have less depth penetration and more lateral spread. These factors could inhibit tree growth, although loblolly pine generally grows well on suboptimal upland sites (Schultz, 1997). Trees in this experiment were estimated to have an age-height relationship (site index) of about 17 m at age 25 yr (Carmean et al., 1989), indicating that loblolly pine was relatively unproductive on the site. In an economy of rising costs of production and unstable commodity values, production of even suboptimal tree growth in a managed silvopastoral practice on marginal land can be profitable due to increasing value of wood fiber (Dangerfield and Harwell, 1990). Additional value can be obtained by linking fee hunting, pine straw, and floriculture with silvopasture (Grado et al., 2001; Husak and Grado, 2002).

At northern latitudes, hedgerows cast a northern shadow of varying, predictable length depending on sun angle, which is a function of location, date, and time of day, and tree age, height, and spacing (Brandle et al., 2000). Four-meter tall apple (Malus P. Mill.) hedgerows oriented north–south at 30° N lat. are predicted to intercept 10% more solar radiation on 21 June than those oriented east–west (Palmer, 1989). It is likely that ordinal direction of loblolly pine hedgerows rows also affects the quantity of light received at the alley surface. A model of the spatiotemporal shade dynamics (Hooge, 2005) tended to support PAR subplot responses (Fig. 4) that at least some portion of the alley was shaded during much of the day from May through September (Fig. 5). Most of the alley was unshaded from May through July due to a low solar angle (12° from vertical). However, south subplots received increasingly brief periods of direct irradiance during May to July across years. Neither grass species was able to fully use direct irradiance during May and June because of growth stage and harvest cycles. The vertical solar angle increased with time to the point that alleys were in almost continuous shade for several months of the year. The variation in solar irradiance (shading) caused subplot differences in mean topsoil temperature (Fig. 3), and probably contributed to differences in topsoil water in control and rip treatments (Fig. 6). Trenching tended to conserve topsoil water in north and south subplots, unlike that of control or rip treatments. The inability of tree roots to recolonize the alleys in the trench treatment, unlike the control or rip treatments, probably contributed to this apparent redistribution of topsoil water.

Yield of annual ryegrass herbage was rarely influenced by subplot or tillage effects (Fig. 8), probably because shade constrained productivity of this grass species during most of its growth cycle. The absence of tillage and subplot effects on annual ryegrass herbage yield, crude protein, and IVDMD was consistent with growth in shade. However, herbage TNC concentration was greater in middle than south subplots in control and rip treatments, perhaps because of differences in topsoil water. Effect of establishment method, broadcast vs. drilling, was not measured, but could have affected annual herbage yield.

Subplot and tillage influenced yield (Fig. 9) and nutritive value of pearl millet herbage. Pearl millet herbage yield did not differ among subplots (P = 0.99) except in 2004. The middle subplot yielded more than the north or south subplots in the control and rip treatments in 2 of the 3 yr of the experiment. Pearl millet was perhaps better fit than annual ryegrass for this cropping practice because alleys received more solar irradiance during its growth period, and its taller growth habit (about 1.2 m) improved competition for light. Further, the edge effect border trees had on pearl millet yield was not as strongly ameliorated by root pruning as it was in annual ryegrass. The control yielded less pearl millet herbage (4990 kg ha^–1) than the rip treatment (6760 kg ha^–1) in 2003, and less in the control (1260 kg ha^–1) than the trench treatment (3300 kg ha^–1) in 2005. However, tillage treatments did not differ in 2004 (P 0.13). Edge effect responses of the two grass species could have been caused by inherent differences in shade tolerance. General guidelines indicate that pearl millet and annual ryegrass are intolerant and intermediate, respectively, to growth in shade (Plants Database, USDA, NRCS, http://plants.usda.gov/java/charProfile?symbol=TRDA3 and http://plants.usda.gov/java/charProfile?symbol=LOPEM2, for pearl millet and annual ryegrass respectively, cited 10 Sept. 2007; verified 10 Nov. 2008).

The IVDMD and TNC of pearl millet herbage were influenced by subplot and tillage effects, probably because of microsite (PAR, and topsoil temperature and water), but interactions with year made broad inferences difficult. The response of IVDMD to shade is inconsistent in alley cropping studies (Burner, 2003; Burner and Belesky, 2004; Burner and Brauer, 2003; Burner and MacKown, 2006). Growth in the shade tends to increase the apparent concentration of herbage crude protein (Burner and MacKown, 2006), and this was generally observed for subplots in control and rip treatments. Subplots did not differ in crude protein in the trench treatment, suggesting that factors other than shade, that is, topsoil temperature or water, could affect the herbage N response. Growth at low irradiance also can foster NO₃^––N accumulation in herbage (Burner and MacKown, 2006), which was not examined in this experiment. Tillage caused pearl millet herbage crude protein to increase, and IVDMD to decrease, with water stress (2005). In 2004, with reduced water stress, there were no significant tillage effects on yield, crude protein, or IVDMD.

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Contrary to our hypothesis, either method of root pruning significantly reduced tree growth. Trenching had a greater impact on tree height than ripping, perhaps because the ripping tines were shallower (0.6 m) than the trench (1 m), and roots could recolonize the alley following ripping. Pruning of loblolly pine roots as close as 1.5 m from the tree stem is not recommended unless productivity of the alley crop can compensate for pruning costs and reduction in tree growth. These 9.7-m wide alleys bounded by 7- to 10-m tall trees had only marginal potential for herbage production even with root pruning. Annual and subplot variation in alley crop yield were large, and variation probably was exacerbated by crop–tree competition for light and topsoil water. There was no benefit of ripping and trenching on annual ryegrass herbage yield, but either practice usually improved pearl millet herbage yield. However, the yield increases did not appear to be sufficiently great to justify the cost, especially if these practices reduce tree growth. Silvopasture using a perennial shade tolerant herbage might be a better practice than annual cropping for similarly configured loblolly pine stands.

	ACKNOWLEDGMENTS

 
The authors appreciate the efforts of Dr. Henry Pearson (USDA-ARS, retired) who established the loblolly pine plantation. Randy King (USDA-NRCS, Booneville) provided the trencher and forage plot harvester. Technical assistance was provided by J. Ruckle (USDA-ARS, Beaver, WV), and K. Chapman, J. Whiley, and S. Haller (USDA-ARS, Booneville, AR). Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
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.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Brandle, J.R., L. Hodges, and B. Wight. 2000. Windbreak practices. p. 79–118. In H.E. Garrett et al. (ed.) North American agroforestry. ASA, Madison, WI.
• Brauer, D., D. Burner, and M. Looper. 2004. Effects of tree configuration on the understory productivity of a loblolly pine-forage agroforestry practice. Am. Forage Grassl. Counc. Proc. 13:412–416.
• Burner, D.M. 2003. Influence of alley crop environment on orchardgrass and tall fescue herbage. Agron. J. 95:1163–1171.[Abstract/Free Full Text]
• Burner, D.M., and D.P. Belesky. 2004. Diurnal effects on nutritive value of alley-cropped orchardgrass herbage. Crop Sci. 44:1776–1780.[Abstract/Free Full Text]
• Burner, D.M., and D.P. Belesky. 2008. Relative effects of irrigation and intense shade on productivity of alley-cropped tall fescue herbage. Agrofor. Syst. 73:127–139.
• Burner, D.M., and D.K. Brauer. 2003. Herbage response to spacing of loblolly pine trees in a minimal management silvopasture in southeastern USA. Agrofor. Syst. 57:69–77.[Medline]
• Burner, D.M., and C.T. MacKown. 2005. Herbage nitrogen recovery in a meadow and loblolly pine alley. Crop Sci. 45:1817–1825.[Abstract/Free Full Text]
• Burner, D.M., and C.T. MacKown. 2006. Nitrogen effects on herbage nitrogen use and nutritive value in a meadow and loblolly pine alley. Crop Sci. 46:1149–1155.[Abstract/Free Full Text]
• Cameron, D.M., S.J. Rance, R.M. Jones, D.A. Charles-Edwards, and A. Barnes. 1989. Project STAG: An experimental study in agroforestry. Aust. J. Agric. Res. 40:699–714.
• Carmean, W.H., J.T. Hahn, and R.D. Jacobs. 1989. Site index curves for forest tree species in the eastern United States. Gen. Tech. Rep. NC-128. USDA, For. Serv., North Central For. Exp. Stn., St. Paul, MN.
• Dadhwal, K.S., and V.P.S. Tomar. 1999. Enhancing production by managing tree roots in an agroforestry land use system. In M.D. Razzaque et al. (ed.) South Asian Assoc. for Region. Coop., Agric. Info. Cent. Newsl. 9(2): 5–6. Available at http://www.saic-dhaka.org/newsletter/Apri-june1999/#Agroforestry (accessed 10 July 2007; verified 10 Nov. 2008).
• Dangerfield, C.W., Jr., and R.L. Harwell. 1990. An analysis of a silvopastoral system for the marginal land in the southeast United States. Agrofor. Syst. 10:187–197.
• Denison, R.F., J.M. Fedders, and C.B.S. Tong. 1990. Amyloglucosidase hydrolysis can overestimate starch concentration of plants. Agron. J. 82:361–364.[Abstract/Free Full Text]
• Gillespie, A.R., S. Jose, D.B. Mengel, W.L. Hoover, P.E. Pope, J.R. Seifert, D.J. Biehle, T. Stall, and T.J. Benjamin. 2000. Defining competition vectors in a temperate alley cropping system in the Midwestern USA. 1. Production physiology. Agrofor. Syst. 48:25–40.[CrossRef]
• Goering, H.K., and P.J. Van Soest. 1970. Forage fiber analyses (apparatus, reagents, procedures, and some applications). Agric. Handb. 379. USDA Agric. Res. Serv., Washington, DC.
• Grado, S.C., C.H. Hovermale, and D.G. St. Louis. 2001. A financial analysis of a silvopasture system in southern Mississippi. Agrofor. Syst. 53:313–322.[CrossRef]
• Hooge, O. 2005. Zenith angle, solar declination, solar energy, shadow length, and sunrise/sunset using MS Works spreadsheets. Available at http://members.shaw.ca/omjhooge/MS%20Works%20Spreadsheet.htm (accessed 10 July 2007; verified 10 Nov. 2008).
• Husak, A.L., and S.C. Grado. 2002. Monetary benefits in a southern silvopastoral system. South. J. Appl. For. 26:159–164.
• Jose, S., A.R. Gillespie, J.R. Seifert, and D.J. Biehle. 2000. Defining competition vectors in a temperate alley cropping system in the Midwestern USA. 2. Competition for water. Agrofor. Syst. 48:41–59.[CrossRef]
• Lin, C.H., R.L. McGraw, M.F. George, and H.E. Garrett. 1999. Shade effects on forage crops with potential in temperate agroforestry practices. Agrofor. Syst. 44:109–119.
• Littell, R.C., G.A. Milliken, W.W. Stroup, and R.D. Wolfinger. 1996. SAS system for mixed models. SAS Inst., Cary, NC.
• Miller, A.W., and S.G. Pallardy. 2001. Resource competition across the crop-tree interface in a maize-silver maple temperate alley cropping stand in Missouri. Agrofor. Syst. 53:247–259.
• National Oceanic and Atmospheric Administration. 2002. Climatography of the United States No. 81, Arkansas. Natl. Climatic Data Ctr., Asheville, NC.
• Palmer, J.W. 1989. The effects of row orientation, tree height, time of year and latitude on light interception and distribution in model apple hedgerow canopies. J. Hortic. Sci. 64:137–145.
• Payne, W.A. 2000. Optimizing crop water use in sparse stands of pearl millet. Agron. J. 92:808–814.
• Pearson, H.A., R.L. Knowles, P.G. Middlemiss, V.C. Baldwin, and R.L. Busby. 1995. United States agroforestry estate model. Compiler 13:27–37.
• SAS Institute. 2002. SAS/STAT user's guide. Release 9.1. Windows version 5.1.2600. SAS Inst., Cary, NC.
• Schultz, R.P. 1997. Loblolly pine. Agric. Handb. 713. USDA., For. Serv., U.S. Gov. Print. Office, Washington, DC.
• Smith, D. 1981. Removing and analyzing total nonstructural carbohydrates from plant tissue. Bull. 2107. Wisconsin Agric. Exp. Stn., Madison.
• Ssekabembe, C.K., P.R. Henderlong, and M. Larson. 1997. Effect of hedgerow orientation on maize light interception and yield in black locust alleys. Int. Tree Crops J. 9:109–118.
• Sudmeyer, R.A., D.J.M. Hall, J. Eastham, and M.A. Adams. 2002. The tree-crop interface: The effects of root pruning in south-western Australia. Aust. J. Exp. Agric. 42:763–772.
• Tyler, D.D., and T.C. McCutchen. 1980. The effect of three tillage methods on soybeans grown on silt loam soils with fragipans. Tenn. Farm Home Sci. 114:23–26.
• USDA. Natural Resources Conservation Service. 2002. Environmental Quality Incentives Program (EQIP). Statewide natural resource concerns. 2002 program policies and cost share opportunities. Available at http://agebb.missouri.edu/modbu/archives/v2n8/EQIPProgramPolicies.doc (accessed 10 July 2007; verified 10 Nov. 2008). USDA-NRCS. Washington, DC.

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Burner, D. M. Articles by Belesky, D. P. Search for Related Content PubMed Articles by Burner, D. M. Articles by Belesky, D. P. Agricola Articles by Burner, D. M. Articles by Belesky, D. P. Related Collections Other Forage Crops Agroforestry Forage quality Crop Systems Plant and Environment Interactions Tillage