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 Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by McBroom, M. W. Articles by Ice, G. G. Search for Related Content PubMed PubMed Citation Articles by McBroom, M. W. Articles by Ice, G. G. Agricola Articles by McBroom, M. W. Articles by Ice, G. G. Related Collections Surface Water Quality Best Management Practices Nutrients

TECHNICAL REPORTS

Surface Water Quality

Water Quality Effects of Clearcut Harvesting and Forest Fertilization with Best Management Practices

^a Arthur Temple College of Forestry and Agriculture, Stephen F. Austin State Univ., Box 6109 SFA Station, Nacogdoches, TX 75962
^b National Council for Air and Stream Improvement, PO Box 458, Corvallis, OR 97339

^* Corresponding author (mcbroommatth{at}sfasu.edu).

Received for publication December 20, 2006.

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and Discussion Conclusions REFERENCES

 
Nine small (2.5 ha) and four large (70–135 ha) watersheds were instrumented in 1999 to evaluate the effects of silvicultural practices with application of best management practices (BMPs) on stream water quality in East Texas, USA. Two management regimes were implemented in 2002: (i) conventional, with clearcutting, herbicide site preparation, and BMPs and (ii) intensive, which added subsoiling, aerial broadcast fertilization, and an additional herbicide application. Watershed effects were compared with results from a study on the same small watersheds in 1981, in which two combinations of harvesting and mechanical site preparation without BMPs or fertilization were evaluated. Clearcutting with conventional site preparation resulted in increased nitrogen losses on the small watersheds by about 1 additional kg ha^–1 each of total Kjeldahl nitrogen (TKN) and nitrate-nitrogen (NO₃–N) in 2003. First-year losses were not significantly increased on the large watershed with a conventional site preparation with BMPs. Fertilization resulted in increased runoff losses in 2003 on the intensive small watersheds by an additional 0.77, 2.33, and 0.36 kg ha^–1 for NO₃–N, TKN, and total phosphorus, respectively. Total loss rates of ammonia nitrogen (NH₄–N) and NO₃–N were low overall and accounted for only 7% of the applied N. Mean loss rates from treated watersheds were much lower than rainfall inputs of about 5 kg ha^–1 TKN and NO₃–N in 2003. Aerial fertilization of the 5-yr-old stand on another large watershed did not increase nutrient losses. Intensive silvicultural practices with BMPs did not significantly impair surface water quality with N and P.

Abbreviations: BMPs, best management practices • DAP, diammonium phosphate • LW, large watershed • SMZ, streamside management zone • SW, small watershed • TKN, total Kjeldahl nitrogen • TN, total nitrogen • TP, total phosphorus

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and Discussion Conclusions REFERENCES

 
THE southeastern USA produces 258 million cubic meters of wood products per year, contributing over 60% of the timber products for the nation (Wear and Greis, 2002). The southeastern USA produces a greater volume of timber products than any other country in the world, and almost all of it comes from private lands. To sustainably produce this volume of wood on a shrinking land base, forest management in the southeastern USA has become increasingly more chemically intense over the past couple of decades. Herbicide use has become more prevalent as the preferred means of controlling competing vegetation, with about 0.8 million ha yr^–1 in the Southeast receiving herbicide applications (Wear and Greis, 2002). In addition, increased fertilization is partly responsible for nearly doubling stand productivity during this time (Wear and Greis, 2002). Forest fertilization increased nearly 800% between 1990 and 1999, with almost 4 million ha fertilized in the Southeast since 1969 (Wear and Greis, 2002). This area exceeds the amount of forest area fertilized in the rest of the world (Wear and Greis, 2002). Forest fertilization is predicted to continue to increase in the future. This has generated concern about the potential water quality effects of forest fertilization.

Nutrient concentrations are generally lower in streams from forested watersheds than adjacent agricultural areas (McBroom et al., 1999; Chang et al., 1980), and the effects of forest practices such as clearcutting on nutrient losses are generally small (Van Lear et al., 1985; Lawson and Hileman, 1982; McClurkin et al., 1985). Atmospheric inputs of N and P can be greater than runoff losses (McClurkin et al., 1985). Large spatial variation in streamwater chemistry of headwater systems is often observed, even in relatively pristine areas (Temnerud and Bishop, 2005). This variation can be influenced by plant communities, geology, and other factors (Holloway et al., 1998).

Researchers have found either no increases in nutrient losses after harvest (McClurkin et al., 1985; Lawson and Hileman, 1982; Lawson et al., 1984) or only slight first-year increases (Aubertin and Patric, 1974). Site-preparation practices, such as prescribed burning, have the potential to result in increased nutrient export through increased leaching loss and erosion losses (Binkley et al., 1999; Fisher and Binkley, 2000; Van Lear et al., 1985). Herbicide applications can also result in higher N losses through increased leaching and denitrification (Vitousek and Matson, 1985; Likens et al., 1970).

At the Alto Watersheds in East Texas, Blackburn and Wood (1990) found significant increases in first-year N, P, and K losses after clearcutting with shear and windrow site preparation. Very high rates of sediment loss were also observed due to the steep slopes in the area. Prescribed burning conducted on the sites released nutrients stored in plant biomass. Although yields increased over the control watersheds at Alto, overall nutrient concentrations were low and were not high enough to result in eutrophication of these streams (Blackburn and Wood, 1990).

Forest fertilization has the potential to increase nutrient losses (Riekerk, 1989; Bisson et al., 1992). The potential impact on streams depends on the quantity applied and whether nutrients are applied directly to drainage systems or unfertilized buffer zones are maintained along streams. Conditions at the time of and immediately after application can affect runoff, urea hydrolysis, denitrification, and other processes. Applications during summer months potentially result in lower runoff concentrations because tree uptake rates are higher (Blazier et al., 2006), volatilization and oxidation rates are greater (Fisher and Binkley, 2000; Liechty et al., 1999; Bisson et al., 1992), and runoff rates are usually lower (Blackburn et al., 1986).

Studies in which nutrients are not directly applied to streams often result in lower concentrations of NO₃–N and phosphorus as ortho-phosphate (PO₄–P) (Binkley et al., 1999; Bisson et al., 1992). For example, direct application to streams generally results in short-term increases of nutrient concentrations, followed by rapid dilution occurring over time and downstream. Even narrow buffers tend to reduce nutrient concentrations, but treatment effects can be lower if wider buffers are used (Binkley et al., 1999). In general, width should be proportional to slopes, management practices, and the nature of the drainage (Comerford et al., 1992).

Little published literature exists that is specific to the effects of fertilization on southern forested streams. Liechty et al. (1999) reported the effects of an aerial application of 437 kg ha^–1 of urea and 140 kg ha^–1 of diammonium phosphate (DAP) to a 150-ha watershed in the Ouachita Mountains of Arkansas. Fertilizers were not applied to streamside management zones (SMZs) on perennial streams, although applications did occur on ephemeral and intermittent streams. After application, NO₃–N and NH₄–N increased from 0.30 and <0.01 mg L^–1 during pretreatment, respectively, to peak concentrations of 3.58 mg L^–1 for NO₃–N and 4.91 mg L^–1 for NH₄–N after the urea application. This peak concentration occurred 24 h after application during a large storm event. Total organic N peaked at 44.5 mg L^–1 after urea application. Diammonium phosphate application resulted in less dramatic effects on stream N concentrations, with a maximum NO₃–N concentration of 1.37 mg L^–1 6 h after application. Concentrations were below detection limits 30 h after application. Diammonium phosphate application resulted in lower N concentrations because less N was applied in the DAP application and there was not a significant storm event soon after application.

Binkley et al. (1999) reported on three studies published internally by Weyerhaeuser Corporation. In one study, the applications of 170 kg N ha^–1 and 28 kg P ha^–1 applied as urea and DAP to a 15-yr-old loblolly pine stand in the Coastal Plain of North Carolina using ground-based equipment resulted in a short-term spike in NO₃–N, from a mean of 0.6 mg L^–1 pretreatment to a maximum of 1.2 mg L^–1 post-treatment. In another Weyerhaeuser study reported in Binkley et al. (1999), 145 kg N ha^–1 and 40 kg P ha^–1 were applied as urea and DAP in North Carolina to a 15-yr-old loblolly pine plantation. Streams were avoided. Average annual NO₃ decreased from 0.80 to 0.14 mg L^–1, and PO₄–P decreased from 0.07 to 0.03 mg L^–1 after treatment.

Although silvicultural practices have become more intense in recent decades, BMPs have been implemented during this time to mitigate potential water quality effects. The retention of SMZs along perennial, intermittent, and many ephemeral streams in particular has become a standard component of the BMP program in many states. Best management practices are effective in reducing nonpoint-source pollution from silvicultural operations (Lynch and Corbett, 1990; Arthur et al., 1988; Comerford et al., 1992; Kochenderfer et al., 1997; Stuart and Edwards, 2006). Best management practices have been shown not to be the "weak sister" to point source controls but in fact have been shown to be the optimum means for controlling nonpoint-source pollution in extremely complex and variable watershed conditions (Ice, 2004). Some states, such as California and Oregon, have made BMPs mandatory for silvicultural operations. In most of the southeastern states, BMPs are voluntary, and there are no legal requirements for BMP implementation on lands on which normal silvicultural operations are conducted. Voluntary BMPs in Texas are implemented on 91.5% of silvicultural operations (Carraway et al., 2002).

The Texas Intensive Silviculture Study was initiated to evaluate the water quality and quantity effects of clearcut harvesting, contemporary site preparation techniques, and forest fertilization, all with BMPs. The study was conducted on nine small (2.5 ha) and four large (70–135 ha) watersheds just west of the town of Alto near the Neches River. The nine small watersheds were used in a study in the late 1970s and early 1980s by Texas A&M University to evaluate water quality effects from silvicultural practices of that era. Among the watersheds studied in the western Gulf Coast in 1980s, these watersheds had the highest losses of sediment (2937 kg ha^–1 on sheared watershed the first year after treatment), indicating that they are on the upper end of the range of sensitivity to treatments (Blackburn et al., 1986). Assessing the effects of contemporary silvicultural practices on water quality and quantity in an area that is highly sensitive to treatments is important for nonpoint-source pollution management. Four large watersheds were added to the original study to evaluate water quality effects on stand-size treatment areas. The purpose of this study was to examine the effects of contemporary silvicultural activities on stream nutrient concentrations and watershed losses and to compare these results with results from the 1980 study conducted before silvicultural BMP adoption.

	Materials and Methods

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and Discussion Conclusions REFERENCES

 
Study Watersheds
Nine small (2.5 ha) and four large (80–120 ha) watersheds (31°36' 07'' N, 95°14'12'' W), located in the Neches River Watershed about 16 km west of Alto, Texas were used in this study (Fig. 1 ). This area has a humid, subtropical climate with hot summers and cool winters. At Nacogdoches, Texas, the mean (1901–1994) annual rainfall and rain days were 117 cm and 89 d, respectively, with April and May being the wettest months. The mean (1901–1994) annual temperature is 18.7°C, with an average summer temperature of 27.2°C and an average winter temperature of 9.5°C (Chang et al., 1996).

View larger version (57K): [in this window] [in a new window]   Fig. 1. Texas Intensive Silviculture Study Alto experimental watersheds in East Texas, USA, by treatment type and acreage.

Study Design and Treatments
A paired watershed approach was used for the small watersheds (SW) and large watersheds (LW). There was one control (LW1), and there were three treatments for large watersheds. For the nine small watersheds, there were three replicates of the two treatment regimes and controls (SW3, SW5, and SW8). Watersheds were calibrated from January 1999 through April 2002, and treatments were conducted from April through December 2002. A more detailed discussion of the study design and treatments is given by McBroom (2005).

Watershed treatments (except LW3) consisted of clearcut harvesting of loblolly pine plantations beginning in April and finishing by June 2002. Trees were harvested with rubber tired feller bunchers equipped with a rotary shear head. Logs were skidded tree-length to landings, where they were merchandized with a cut-to-length processor that provided for optimal log value. Tops and nonmerchantable stem portions were removed from landings and placed on skid trails to minimize soil compaction, or they were distributed throughout the watershed, especially in erosion-sensitive areas.

After harvest, watersheds received intensive or conventional site preparation. Conventional site preparation (conducted in LW2, SW2, SW4, and SW9) was considered to be the minimum necessary treatments to re-establish loblolly pine plantations. This involved an initial aerial broadcast application of 1.17 L ha^–1 imazapyr {(±)-2-[4.5-dihydro-4-methyl-4-(1-methylethyl)-5-oxo-1H-imidazol-2-yl]-3-pyridinecarboxylic acid}and 4.68 L/ha (2qt/ac) glyphosate (N-(phosphonomethyl) glycine) with 1.17 L ha^–1 of Rebound surfactant in September 2002 to suppress woody and herbaceous vegetation. Watersheds were machine planted in December 2002 with around 1200 loblolly pine stems per hectare. In March 2003, a hand-applied, banded herbaceous weed control consisting of 0.81 L ha^–1 Oustar (a mixture of hexazinone (3-cyclohexyl-5-(dimethylamino)-1-methyl-1,3,5-triazine 2,4(1H,3H)-dione) and sulfometuron methyl [methyl 2-[[[[(4,6-dimethyl-2-pyrimidinyl)amino]carbonyl]amino]sulfonyl]benzoate]) was implemented.

Intensive site preparation (conducted in LW4, SW1, SW6, and SW7) was considered to be a treatment that would allow for more rapid stand establishment and higher productivity. This treatment included everything in the conventional method and added a subsoiling operation in November 2002. Subsoiling involved the use of large bulldozers (Caterpillar D8) pulling savannah plows through the soil with the coulters chained up. This operation fractured the soil to a depth of around 60 cm, allowing for better seedling root penetration. Coulters were raised because on erodible sites with steeper topography, bedding and ripping is not recommended due to the potential for increased sediment losses (Beasley, 1979). Subsoiled rows were hand planted with around 1200 stems ha^–1 in mid-December 2002. Also in mid-December, 280.2 kg ha^–1 of granular DAP was applied by aerial broadcast to the intensive watersheds. In addition to the banded herbaceous weed control in the first growing season as done with the conventional site preparation regime, the intensive watersheds received an additional aerial broadcast herbaceous weed control in the spring of the second growing season with the same herbicide at the same concentration as the banded application (results not reported here). The treatment on LW3, a 5-yr-old plantation at time of treatment, involved fertilization of 280.2 kg ha^–1 DAP in August 2002 followed by a herbicide release of 0.73 L ha^–1 of imazapyr in the following month.

Data obtained by GPS of the flight lines were provided by the contractor, and no direct application of fertilizer into the SMZ was documented in any of the aerial applications. Targets were set up along in the SMZ for the aerial DAP application, and although some pellets drifted into the SMZ, there was no evidence of any significant application directly to the stream. Post-treatment BMP inspections by the Texas Forest Service verified that treatments complied with Texas BMPs.

Sample Collection and Analysis
Streamflow on small watersheds was monitored with 0.91-m H-flumes. Runoff from the watershed first flows into a sediment trap then into an approach section (4.3 m long, 1.2 m wide, 0.9 m high) before passing through the 0.91-m H-flume. The approach section promotes laminar flows for accurate depth-discharge determinations. A three-turn potentiometric float and pulley level recorder (Intermountain Environmental, Logan, UT) installed in the stilling well at the sidewall of the flume measured stage. Discharge was calculated from stage recordings stored at 5-min intervals in the datalogger (CR500/510; Campbell Scientific, Logan, UT). Stage data were retrieved from the datalogger every 2 wk and after each storm-runoff event using a laptop computer. An automatic pumping sampler (Model 3700; Isco, Lincoln, NE) allowed for collection of discrete water samples along the runoff hydrograph, typically in 30-min intervals.

The large watershed study used concrete control structures for flow measurements on LW1, LW2, and LWW3 (1.5 m high x 2.5 m wide). A 1.8-m corrugated iron culvert was used as a control structure on LW4. Stage discharge rating curves were established by measuring stream discharge for each structure (USGS, 1980). A portable Doppler flow meter (Flow Mate Model 2000; Marsh-McBurney, Fredrick, MD) with a top-setting wading rod was used to measure velocity. The Manning equation was used to estimate discharge on LW4 (Chang, 2006).

Water samples were retrieved within a few hours after a storm-runoff event. Samples were iced and transported to the forest hydrology laboratory at Stephen F. Austin State University for compositing and preservation. Stage data were processed. Hydrographs were developed for each watershed, and sample collection times on the hydrograph were determined. For most runoff events, samples were equal volume weight composited so that one set represented the rising limb, one set the peak, and one set the recession limb of the storm hydrograph. Additional composites were generated in the same manner for larger storm events or for complex hydrographs that had multiple peaks. Individual samples were sent when few samples were collected or during more extreme runoff events.

Aliquots from each composite were poured into 1-L unpreserved bottles for analysis of NO₃ and PO₄–P. Aliquots were poured into 500-mL bottles preserved with sulfuric acid to pH <2 for analysis of NH₄–N and total Kjeldahl nitrogen (TKN).

Samples were packed on ice and delivered to a contract laboratory (Ana-Lab, Kilgore, TX) for analysis of the parameters listed previously. Analysis methods conformed to established APHA and EPA methodology (Table 1 ) (USEPA, 2003). Soil samples were collected pre- and post-treatment on intensive watersheds on a 100-m grid from each of three slope positions: lower slope, near the stream; mid-slope; and upper slope, near the watershed divide. Samples were collected from the top 30 cm of the soil profile. Soil samples were analyzed for nutrients (NO₃, TP, and K) by the Stephen F. Austin State University Soil Testing Laboratory using a modified Morgan extractant (Texas Agricultural Extension Service, 1980).

Statistical Analysis
Treatments effects for watersheds were determined using the paired watershed approach (Chang, 2006). Three years of pretreatment data served as a basis for developing pretreatment calibration regression equations between treatment and control watersheds.

The calibration equation for streamflow took the simple linear form:

[1]

where y_t = estimated runoff for the treatment watershed, y_c = intercept of the regression line in the treatment watershed, b = regression coefficient (slope), and x_c = runoff for the control watershed. The difference between the observed (y) and predicted (y_t) storm runoff in the treatment watershed is the treatment effect (q), where

[2]

For typical paired watershed studies, the ANCOVA is an effective method for determining treatment differences. The response variable from the control watershed is treated as a covariate, so that the change in the treated watershed response during the treatment period can be determined. However, the parametric ANCOVA has three main assumptions: (i) the dependent variable is unaffected by the group variable, (ii) the standard analysis assumes the relationship between the independent and dependent variables is significant and identical among the groups, and (iii) the residuals are independently and identically distributed.

For the Alto Watershed data, parameter frequency distributions were typically skewed, with many smaller observations and a few larger ones. These types of situations in hydrologic data, although anticipated, result in skewed distributions that violate parametric ANCOVA assumptions. Transformations such as a log₁₀+1 can be used to correct this problem. However, ordinary least squares regression on which ANCOVA is based may not provide an unbiased estimate for nonlinear transformations (Cade and Richards, 1996). Least absolute deviation regression has been shown to have greater power than ordinary least squares regression for asymmetric error distributions and heavy-tailed, symmetric error distributions and has greater resistance to the influence of a few outlying values of the dependent variable (Cade and Richards, 1996).

Least absolute deviation regression was used with a rank-score hypothesis test to conduct an ANCOVA-like analysis, where an interaction term consisting of the dependent variable multiplied by the group variable was included in the model (Cade and Richards, 2001). Then a reduced form of the model, lacking the interaction variable, was evaluated, and the null hypothesis that the estimated interaction term is equal to zero or that the differences in slopes are equal to zero was evaluated (Cade and Richards, 2001). A rank-score procedure for hypothesis testing and confidence interval construction was used for regression quantile estimates that do not assume independently and identically distributed errors. The BLOSSOM software package, developed by the United States Geological Survey, was used to conduct these tests.

Pretreatment and first year post-treatment results are reported here. Greatest treatment effects are typically measured during the first year after harvest in the Southeast, with subsequent years showing less impact (Blackburn and Wood, 1990). Therefore, first-year results are often the most valuable for determining water quality impacts of silvicultural practices.

	Results and Discussion

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and Discussion Conclusions REFERENCES

 
Baseflow Concentrations
On overage, 17 and 3 baseflow samples were collected in the pre- and post-treatment periods, respectively. Before harvest and fertilization, baseflow concentrations of nutrients were relatively low and consistent with expected levels in streams draining forested watersheds (Table 2 ). These streams tend to be ephemeral or intermittent in nature, so baseflow samples were only collected when streams were flowing. After harvest and fertilization, no change in baseflow nutrient concentrations was observed. Fertilizers were not applied directly to streams, and forested buffers were maintained along streams. No spike was observed in baseflow concentrations during the first sampling event after fertilization. For LW3, where nutrients were applied to the 5-yr-old loblolly pine plantation, no change in baseflow nutrient concentrations was observed.

Stormflow Concentrations
A total of 38 and 20 storm events were analyzed in the pretreatment and post-treatment periods, respectively. The largest stormflow concentrations of N were observed after fertilization on the intensive watersheds. Watersheds were fertilized with DAP on 15 Dec. 2002, and on the 30 Dec. 2002 runoff event the peak concentration of 2.4 mg L^–1 NH₄–N was observed on LW4 (Fig. 2 ). Subsequent events had lower peak concentrations. For NO₃–N, peak concentrations occurred later, during the late February events. For TKN, the peak was still later, during the June events. This trend may be attributed to the form of N applied. Because NH₄–N was applied as DAP, NH₄–N concentrations peaked immediately and then declined. As nitrification occurred, NH₄–N was converted to NO₃–N, and NO₃–N concentrations peaked a couple of months later. This lag in nitrification could be due to low temperatures at the time of application, which are often associated with lower nitrification rates (Bisson et al., 1992). Because N was taken up into organic forms, TKN concentrations peaked in summer. A similar trend was observed on intensive small watersheds as well. Stormflow concentrations returned to near prefertilization levels by the end of the year. This trend in peak N concentrations through winter and spring was not observed on the conventional large or small watersheds, indicating that this trend may be attributed to the DAP application.

View larger version (15K): [in this window] [in a new window]   Fig. 2. Maximum post-treatment stormflow LW4 (intensive large watershed) nitrogen concentrations for the Alto Watersheds in East Texas (fertilization on 15 Dec. 2002).

View larger version (12K): [in this window] [in a new window]   Fig. 3. Maximum post-treatment stormflow LW4 (intensive large watershed) phosphorus concentrations from the Alto Watersheds in East Texas (fertilized on 15 Dec. 2005).

Nutrient Concentrations during Storm Events
On larger storm events, greatest TKN and TP concentrations were typically observed just before peak discharge, near the top of the rising phase of the hydrograph. Concentrations fell off after peak discharge (Fig. 4 ). For storm events with multiple peaks, peak nutrient concentrations occurred with the first main peak, with the secondary flow peak having lower nutrient concentrations. This general trend was observed on all watersheds. This indicates that most of the nutrients available for transport move with the first storm flush, with fewer nutrients available in subsequent flushes. These peak concentrations typically occur for 15 to 20 min during a storm event before dropping.

View larger version (10K): [in this window] [in a new window]   Fig. 4. Stream discharge and total Kjeldahl nitrogen (TKN) concentration from SW6 (intensive small watershed) for a post-treatment storm event on 4 Nov. 2002 at the Alto Watersheds in East Texas.

On the clearcut large watersheds, significantly greater growing season storm runoff was measured from treated watersheds than from the control, but differences in watershed response between the control and treated watersheds were insignificant during the nongrowing season. This resulted overall in a change in volume of storm runoff that was not statistically significant for the year. Differences in large versus small watershed flow response were attributed to the small, headwater watersheds generally having a steeper channel, less storativity, and more circular basin shapes (McBroom et al., 2007).

Stormflow Nitrogen Mass Losses
For the large watersheds, there was no significant increase in loss rates for TKN for either of the clearcut watersheds after harvest and site preparation (Table 3 ). However, NH₄–N increased significantly on LW4 after fertilization, with an estimated 0.097 kg ha^–1 increase in NH₄–N losses (Table 4 ). This increase can be attributed to the aerial broadcast application of 113.4 kg ha^–1 of DAP on 15 Dec. 2002 on LW4 but not LW2. This application rate equates to about 40.2 kg ha^–1 of elemental N applied to the watershed as NH₄–N. This means that only about 3% of the applied NH₄–N was observed in the streamflow. Nitrate-N also increased significantly, at an estimated rate of 2.618 kg ha^–1 over what would have been observed without treatment (Table 4). There were no increases in any of the N parameters on LW3, indicating that the release fertilization had no measurable impact on stream nutrient status. Stormflow did not increase significantly on any of the large watersheds after harvest (Table 4).

Concentrations of N forms in rainfall were also measured during the study. Accordingly, the amount of N input from precipitation was estimated based on these concentrations and annual precipitation depth (Fig. 5 ). Annual N inputs were greater than losses measured in streamflow for each year and treatment. Similarly, McClurkin et al. (1985) found that in Tennessee forested watersheds the rainfall input of TKN (8.0 kg ha^–1 yr^–1) was nine times greater than TKN stormflow export. Schreiber et al. (1980) concluded that N deposited in rainfall can account for a significant portion of the overall annual N budget. For the fertilized watersheds, 40.2 kg ha^–1 of N was applied as DAP. After fertilization, the predicted increase in NH₄–N for 2003 for SW6 was 0.6 kg ha^–1 and around 0.1 kg ha^–1 for the other three fertilized watersheds (Table 4). This is still well below the 1.36 kg ha^–1 of NH₄–N that was contributed by rainfall for that year. This implies that even after fertilization at a rate almost 30 times greater than rainfall input, these watersheds are still serving as sinks. Soil NO₃–N modestly increased from 2 to 11 ppm on LW4 after fertilization, supporting this hypothesis. The amount of N leaving these watersheds in stormflow is a small fraction of the amount that is used by vegetation and stored in organic matter and soils.

View larger version (17K): [in this window] [in a new window]   Fig. 5. Annual estimated nitrogen contributions from rainfall and losses in stormflow on the Alto Watersheds in East Texas.

These watersheds serve as P sinks. Soil test P did increase after fertilization from 4 to 8 ppm on LW4. However, these P levels are considered to be low when compared with East Texas agricultural soils, indicating that these soils likely have additional P absorption potential (Texas Agricultural Extension Service, 1980).

Comparison with 1980 Silviculture
Blackburn and Wood (1990) found significantly greater losses of NO₃–N, PO₄–P, and TP from the sheared watersheds than the control or roller-chopped watersheds. They reported no significant change in NH₄–N. Losses in the current study were generally higher than those from 1981, especially for NO₃–N, for the fertilized and unfertilized watersheds (Table 6 ). Blackburn and Wood (1990) indicated that the site preparation burn in 1981 may have made more nutrients available for transport. In the current study, additional nutrients available for transport would have resulted from fertilizer application on intensive watersheds, herbicide treatments, and the additional streamflow moving more organic matter and nutrients from the conventional watersheds. Generally, fertilized watersheds in the current study had higher loss rates for NO₃–N, NH₄–N, and PO₄–P, whereas TP loss rates were slightly lower than in 1981. Losses of NO₃–N and NH₄–N were also slightly higher in 2003 than 1981 for control and conventional watersheds.

	Conclusions

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and Discussion Conclusions REFERENCES

 
Clearcutting with intensive stand reestablishment practices using BMPs did not dramatically affect runoff concentrations and losses of nutrients from these watersheds. Nutrient export increased slightly after fertilization on the intensive watersheds. However, the total annual loss was a small fraction of the applied nutrient. Elevated nutrient loss rates were observed only for the first few storms after the December DAP application. Although nutrient loss rates were statistically significant, they are not likely to have any negative impact on stream biota. Total nutrient loss rates from treated watersheds were much lower than rainfall inputs. Fertilization of the 5-yr-old stand resulted in no measurable increase in nutrient losses.

Silvicultural activities generally have a small, short-lived impact on water quality, especially when compared with other land uses, such as agriculture or urban development. Fertilization of pine plantations typically only occurs once during a stand rotation, and most of the applied nutrients are rapidly taken up by soils and vegetation (Fisher and Binkley, 2000). Unlike agricultural watersheds where nutrients are applied annually, forested watersheds are not likely to experience saturation due to application rates exceeding receiving crop demands. In addition, the use of silvicultural BMPs helps mitigate possible water quality effects. Streamside buffers in particular help stabilize stream channels and prevent direct application of fertilizers to streams, thus reducing potential water quality impacts.

	ACKNOWLEDGMENTS

 
Funding was provided by the Arthur Temple College of Forestry and Agriculture at Stephen F. Austin State University, the National Council for Air and Stream Improvement (NCASI), and Temple-Inland Forest Products Corporation. Assistance from these organizations and the Texas Institute for Applied Environmental Research (TIAER) is gratefully acknowledged. Temple-Inland provided the research sites and numerous in-kind contributions. The assistance of Dr. Don Turton with Oklahoma State University in project setup is appreciated. Dr. Brian Cade with the United States Geological Survey provided valuable assistance with statistical analyses.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION Materials and Methods Results and 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 and Discussion Conclusions REFERENCES

 

• Arthur, M.A., G.B. Coltharp, and D.L. Brown. 1988. Effects of best management practices on forest streamwater quality in eastern Kentucky. J. Am. Water Resour. Assoc. 34:481–495.[CrossRef]
• Aubertin, G.M., and J.H. Patric. 1974. Water quality after clearcutting a small watershed in West Virginia. J. Environ. Qual. 3:243–249.[Abstract/Free Full Text]
• Beasley, R.S. 1979. Intensive site preparation and sediment losses on steep watersheds in the Gulf Coastal Plain. Soil Sci. Soc. Am. J. 43:412–417.[Abstract/Free Full Text]
• Binkley, D., H.L. Allen, and H. Burnham. 1999. Water quality effects of forest fertilization. NCASI Tech. Bull. No. 782. National Council of the Paper Industry for Air and Stream Improvement, Inc., Research Triangle Park, NC.
• Bisson, P.A., G.G. Ice, C.J. Perrin, and R.E. Bilby. 1992. Effects of forest fertilization on water quality and aquatic resources in the Douglas-fir region. In H.N. Chappel et al. (ed.) Forest fertilization: Sustaining and improving nutrition and growth of western forests. Inst. For. Resour. Contribution 73. Univ. of Washington, Seattle, WA.
• Blackburn, W.H., and J.C. Wood. 1990. Nutrient export in stormflow following forest harvesting and site preparation in East Texas. J. Environ. Qual. 19:402–408.[Abstract/Free Full Text]
• Blackburn, W.H., J.C. Wood, and M.G. DeHaven. 1986. Storm flow and sediment loss from site-prepared forestland in East Texas. Water Resour. Res. 22:776–784.[CrossRef]
• Blazier, M.A., T.C. Hennessey, P. Dougherty, and R. Campbell. 2006. Nitrogen accumulation and use by a young loblolly pine plantation in southeast Oklahoma: Effects of fertilizer formulation and date of application. South. J. Appl. For. 30:66–78.
• Cade, B.S., and J.D. Richards. 1996. Permutation tests for least absolute deviation regression. Biometrics 52:886–902.[CrossRef][Web of Science]
• Cade, B.S., and J.D. Richards. 2001. User manual for Blossom statistical software. Midcontinent Ecological Center, USGS, Fort Collins, CO.
• Carraway, B., J. Donnelan, H. Simpson, and D. Work. 2002. Voluntary implementation of forestry best management practices in East Texas. Texas Forest Service Forest Resources Development Project Best Management Practices Project.
• Chang, M. 2006. Forest hydrology: An introduction to water and forests, 2nd ed. CRC Press, Taylor and Francis, Boca Raton, FL.
• Chang, M., J.D. McCullough, and F.A. Roth, Jr. 1980. Nitrogen yields and land use in southern streams. p. 951–960. In Proc. Symp. Watershed Management. ASAE, Boise, ID.
• Chang, M., L.D. Clendenon, and H.C. Reeves. 1996. Characteristics of a humid climate, Nacogdoches, Texas. College of Forestry, Stephen F. Austin State Univ., Nacogdoches, TX.
• Comerford, N.B., D.G. Neary, and R.S. Mansell. 1992. The effectiveness of buffer strips for ameliorating offsite transport of sediment, nutrients, and pesticides from silvicultural operations. NCASI Tech. Bull. No. 631. National Council of the Paper Industry for Air and Stream Improvement, Inc., Research Triangle Park, NC.
• Compton, J.E., M.R. Church, S.T. Larned, and W.E. Hogsett. 2003. Nitrogen export from forested watersheds in the Oregon Coast Range: The role of N₂–fixing red alder. Ecosystems 6:773–785.[CrossRef]
• Fisher, R.F., and D. Binkley. 2000. Ecology and management of forest soils, 3rd ed. John Wiley & Sons, New York.
• Holloway, J.M., R.A. Dahlgren, B. Hansen, and W.H. Casey. 1998. Contribution of bedrock nitrogen to high nitrate concentrations in stream water. Nature 395:785–788.[CrossRef]
• Ice, G. 2004. History of innovative best management practice development and its role in addressing water quality limited waterbodies. J. Environ. Eng. 130:684–689.[CrossRef]
• Ice, G., and D. Binkley. 2003. Forest streamwater concentrations of nitrogen and phosphorus: A comparison with EPA's proposed water quality criteria. J. For. 101:21–28.
• Kochenderfer, J.N., P.J. Edwards, and F. Wood. 1997. Hydrologic impacts of logging an Appalachian watershed using West Virginia's best management practices. North. J. Appl. For. 14:207–218.
• Lawson, E.R., and L.H. Hileman. 1982. Nutrient distributions in runoff from Ouachita Mountain watersheds. p. 477–482. In Proc. 2nd Biennial Southern Silvicultural Res. Conf. USDA Forest Service Southern Forest and Range Exp. Stn., Ashville, NC.
• Lawson, E.R., T.L. Rogerson, and L.H. Hileman. 1984. Nutrients in storm runoff from hardwood-forested watersheds in the Ozark highlands. p. 558–563. In Proc. 3rd Biennial Southern Silvicultural Research Conf., USDA Forest Service Southern Forest and Range Exp. Stn., Ashville, NC. SO-54.
• Liechty, H.O., J. Nettles, D.A. Marion, and D.J. Turton. 1999. Stream chemistry after operational fertilizer application in the Ouachita Mountains. p. 265–270. In Proc. 10th Biennial Southern Silvicultural Research Conf., Shreveport, LA. USDA Forest Service Southern Research Stn. General Tech. Rep. SRS-30.
• Likens, G.E., F.H. Bormann, N.M. Johnson, D.W. Fisher, and R.S. Pierce. 1970. Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecol. Monogr. 40:23–47.[CrossRef][Web of Science]
• Lynch, J.A., and E.S. Corbett. 1990. Evaluation of best management practices for controlling nonpoint pollution from silvicultural operations. Water Resour. Bull. 26:41–52.
• McBroom, M.W. 2005. The effects of clearcutting, intensive site preparation, herbicide release, and fertilization on surface water quantity and quality at the Alto Experimental Watersheds in East Texas. Ph.D. diss. Stephen F. Austin State Univ., Nacogdoches, TX.
• McBroom, M.W., R.S. Beasley, M. Chang, and G.G. Ice. 2007. Storm runoff and sediment losses from forest clearcutting and stand reestablishment with best management practices in the Southeastern United States. Hydrol. Processes (in press) DOI: 10.1002/hyp.6703.
• McBroom, M.W., R.S. Beasley, M. Chang, B. Gowin, and G.G. Ice. 2003. Runoff and sediment losses from annual and unusual storm events from the Alto Experimental Watersheds, Texas: 23 years after silvicultural treatments. p. 607–613. In K.G. Renard et al. (ed.) First Interagency Conf. on Research in the Watersheds, 27–30 Oct. 2003. USDA-ARS, Washington, DC.
• McBroom, M.W., M. Chang, and M.C. Cochran. 1999. Water quality conditions of streams receiving runoff from land-applied poultry litter in East Texas. p. 549–552. In R. Sakrison and P. Sturtevant (ed.) Watershed management to protect declining species, Proc. Am. Water Resour. Assoc. 1999 Ann. Water Resour. Conf., AWRA Tech. Pub. TPS-99-4.
• McClurkin, D.C., P.D. Duffy, S.J. Ursic, and N.S. Nelson. 1985. Water quality effects of clearcutting upper coastal plain loblolly pine plantations. J. Environ. Qual. 14:329–332.[Abstract/Free Full Text]
• Mowery, I.C. 1959. Soil survey for Cherokee County, Texas. USDA-SCS, U.S. Gov. Print. Office, Washington, DC.
• Riekerk, H. 1989. Forest fertilizer and runoff-water quality. Proc. Soil Crop Sci. Soc. Fla. 48:99–102.
• Schreiber, J.D., P.D. Duffy, and D.C. McClurkin. 1980. Aqueous- and sediment-phase nitrogen yields from five southern pine watersheds. Soil Sci. Soc. Am. J. 44:401–407.[Abstract/Free Full Text]
• Stuart, G.W., and P.J. Edwards. 2006. Concepts about forests and water. North. J. Appl. For. 23:11–19.
• Temnerud, J., and K. Bishop. 2005. Spatial variation of streamwater chemistry in two Swedish boreal catchments: Implications for environmental assessment. Environ. Sci. Technol. 39:1463–1469.[Medline]
• Texas Agricultural Extension Service. 1980. Soil testing procedures. Texas A&M Univ. System, College Station, TX.
• University of Texas at Austin Bureau of Economic Geology. 1968. Geologic atlas of Texas, Palestine sheet. Univ. Texas at Austin Bureau of Economic Geology.
• USEPA. 2003. Index to EPA Test Methods. EPA 901/3-88-001. Office of Water. U.S. Gov. Print. Office, Washington, DC.
• USGS. 1980. National handbook of recommended methods for water-data acquisition. U.S. Dep. of Interior, Reston, VA.
• Van Lear, D.H., J.E. Douglas, S.K. Cox, and M.K. Augspurger. 1985. Sediment and nutrient export in runoff from burned and harvested pine watersheds in the South Carolina Piedmont. J. Environ. Qual. 14:169–174.[Abstract/Free Full Text]
• Vitousek, P.M., and P.A. Matson. 1985. Disturbance, nitrogen availability, and nitrogen losses in an intensively managed loblolly pine plantation. Ecology 66:1360–1376.[CrossRef][Web of Science]
• Wear, D.N., and J.G. Greis (ed.) 2002. Southern forest resource assessment. USDA Forest Service Southern Research Stn., Ashville, NC. General Tech. Rep. SRS-53.

This article has been cited by other articles:

				W. A. Pratt and T. R. Fox Streamside Management Zones Effectiveness for Protecting Water Quality after Forestland Application of Biosolids J. Environ. Qual., August 24, 2009; 38(5): 2106 - 2120. [Abstract] [Full Text] [PDF]

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 Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by McBroom, M. W. Articles by Ice, G. G. Search for Related Content PubMed PubMed Citation Articles by McBroom, M. W. Articles by Ice, G. G. Agricola Articles by McBroom, M. W. Articles by Ice, G. G. Related Collections Surface Water Quality Best Management Practices Nutrients