| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Horticulture and Landscape Architecture
b Dep. of Biosystems and Agricultural Engineering
c Dep. of Statistics
d Dep. of Plant and Soil Sciences, Oklahoma State Univ., Stillwater, OK 74078
* Corresponding author (Greg.Bell{at}okstate.edu)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The amount of N and P lost in surface runoff depends on such factors as soil water content, amount and timing of precipitation or irrigation, method and form of fertilizer application, slope, and various soil properties (Walker and Branham, 1992). Nitrogen (primarily NO3– and NH4+) lost in water runoff may lead to eutrophication in bodies of water at concentrations as low as 1 mg L–1 (Walker and Branham, 1992). Excessive nutrient levels cause an abundance of algae and aquatic plant blooms that deplete oxygen in the water and may cause the death of plants and fish. Dissolved reactive phosphorus (DRP), a term that refers to HPO42– and H2PO4–, is believed to contribute to the eutrophication of bodies of water at concentrations as low as 25 µg L–1 (Walker and Branham, 1992).
The importance of environmentally prudent golf course nutrient management was identified by Kunimatsu et al. (1999), who studied nutrients discharged from a golf course in Japan. Samples were taken from a stream that ran from a forested basin through the golf course. They determined that an increase in nutrient discharge from the stream was due primarily to the N, P, and K applied to the golf course as fertilizer.
A properly established vegetative buffer may help to reduce or eliminate the movement of sediments (Barfield et al., 1979; Hayes et al., 1979), pesticides (Baker et al., 2000; Baird et al., 2000) and nutrients (Gross et al., 1990). Grass vegetation is the primary material recommended for vegetative buffers, also referred to as vegetative filter strips (USDA-NRCS, 1997). Vegetative buffers have been widely studied in agricultural settings and have reduced nutrient losses from agricultural lands (Baker et al., 2000), but limited research has been conducted in the area of golf course management and grass buffer performance.
While previous studies have indicated that turf is helpful for reducing or preventing nutrient runoff from well-maintained turf areas, little research has been completed concerning grass buffer performance for reducing runoff from turf. Researchers have suggested that nutrient runoff from highly maintained turfgrass areas is low, but N and P concentrations in runoff water were high enough to cause eutrophication in many previous studies (Gross et al., 1990; Gross et al., 1991; Harrison et al., 1993; Linde et al.,1994; Krenitsky et al., 1998). Public concern of golf course management and its impacts on the environment demand that strategies be developed and implemented to mitigate the potentially harmful effects of nutrient runoff from golf courses.
Cole et al. (1997) studied runoff from bermudagrass plots simulating golf course fairways that were bordered by grass buffers of differing length and mowing height. They concluded that neither length nor mowing height significantly affected nutrient or pesticide runoff. These results suggest that the initial barrier provided by the grass buffers at the interface of low-cut to high-cut turf was more important for nutrient runoff reduction than the buffer width or height. Consequently, we hypothesized that a buffer mowed at increasingly higher heights may further inhibit nutrient runoff by presenting a series of low-cut to high-cut obstacles.
The objective of this study was to determine if bermudagrass buffers mowed at increasingly higher heights could reduce nutrient runoff from golf course fairways better than bermudagrass buffers mowed at a single height by providing multiple physical barriers to runoff.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Covered troughs collected runoff water from each experimental unit and channeled it through calibrated Parshall flumes by gravity flow. The collection troughs were made of polyvinylchloride pipe (15 cm diam) cut in half lengthwise, and were mounted on wooden support posts. The posts (10 x 10 x 60 cm long) were buried in concrete below the soil frost line to stabilize the troughs. A galvanized shingle-type attachment was fixed to an aluminum angle along the bottom edge of each plot to channel runoff water into the corresponding collection trough (Cole et al., 1997). The shingle-type attachment was sealed to the soil with paraffin wax to prevent water from running beneath the trough. Stainless steel bolts supported a galvanized cover 7.6 cm above the shingle to allow runoff collection while eliminating the entry of unwanted irrigation or rainfall into the trough.
Isco 6700 portable samplers (Isco, Lincoln, NE) were secured to concrete platforms located between each experimental block. Ultrasonic Modules (Isco 710) mounted over each Parshall flume used ultrasonic reflection to measure water level. The sampler was programmed to determine water flow rate from these water level measurements based on a predetermined calibration of each Parshall flume. A pump in each sampler provided runoff sample collection through vinyl suction line tubing (0.95 cm) fitted with a screen strainer and secured to the Parshall flume. A Rapid Transfer Device (Isco 581) enabled information transfer from samplers to a computer.
Time domain reflectometry probes were permanently buried along the slope in each experimental unit to assess soil water content and to help maintain antecedent soil water at uniform conditions before natural rainfall. These probes were centered within each experimental unit at 2.1, 10.2, and 18.3 m from the top of the fall line and buried 15.2 cm deep. The experimental units were irrigated three times per week to field capacity determined by volumetric water content 24 h after saturation. The runoff site was at or very near field capacity (0.22 m3 water m–3 soil) immediately before all irrigation and natural rainfall events.
Runoff Sampling Methodology
To test nutrient runoff, urea and super triplephosphate fertilizer were applied at 49 kg ha–1 N and 24 kg ha–1 P 4 h before irrigating and again following irrigation events to await natural rainfall. The fertilizers were applied as granules and were not "watered in" before irrigation events so that the irrigation experiments represented worst-case conditions. In contrast, the fertilizers were "watered in" following irrigation events to await natural rainfall. The "watering in" process was performed by applying irrigation at 51 mm h–1 for 7 min immediately after fertilizer was applied. Fertilizers were applied to the simulated golf course fairway area six times in 2001 (3 before irrigation and 3 before natural rainfall) and six times in 2002. Fertilizer was not applied to the buffers.
Irrigation was applied at 51 mm h–1 using the in-ground sprinkler-type turf irrigation system three times in 2001. Time from rainfall initiation to runoff was recorded for each rainfall event and the runoff flow rate was measured every 5 min for 1 h after runoff began. Water samples were also collected from the runoff at 5-min intervals for 1 h. Following an hour of runoff, irrigation and sample collection were terminated. The water samples were tested for concentrations of NO3–N, NH4–N, and DRP. Flow rate and sample nutrient concentrations were used to calculate the total amount of nutrients lost for each event. Both irrigation and natural rainfall events were recorded during two growing seasons.
Irrigation runoff was collected on 14 Aug., 30 Aug., and 12 Sept. 2001 and on 18 June, 9 July, and 23 July 2002. Natural rainfall events were recorded on 5 Sept. 2001 (33 mm), and on 17 May (61 mm), 4 June (22 mm), and 12 June (30 mm) in 2002. The time from the beginning of precipitation to the initiation of runoff and the flow rates for each plot were recorded for all events.
Analytical Procedures
Water samples were analyzed for NO3–N and NH4–N using colorimetric methods by automated flow injection analysis and DRP using the phosphomolybdate colorimetric procedure employed by Murphy and Riley (1962). The detection limit was 0.01 mg L–1 for each nutrient in the runoff water samples. The average background levels of nutrients in the irrigation water samples were 0.26 mg L–1 for NO3–N, 0.10 mg L–1 for NH4–N, and 0.05 mg L–1 for DRP, and in natural rainfall samples 0.88 mg L–1 NO3–N, 0.04 mg L–1 NH4–N, and 0.05 mg L–1 DRP. The concentration of NO3–N, NH4–N, and DRP in the precipitation was measured during each event and subtracted from the measured concentrations in collected runoff before statistical analyses were performed.
Nutrient losses following runoff initiation were calculated by multiplying the nutrient concentrations during each sampling interval [(concentration at time 1 + concentration at time 2)/2] by the total amount of runoff that passed through the Parshall flume during each specific 5 min sample period in 60 min of runoff. The total nutrient loss for each treatment during the first 60 min following runoff initiation was computed by adding the total amount of NO3–N, NH4–N, or DRP (after subtracting the amounts measured in the precipitation) that were calculated for each 5 min sample period.
Statistical analyses were performed using SAS version 8.1. Analysis of variance procedures were used to determine nutrient runoff as a function of precipitation or irrigation duration for a randomized complete block design. Repeated measures analysis was performed using PROC MIXED with time as the repeated measure treatment. A model for intraplot variance was determined using an auto regressive variance model. There was no collection date x treatment interaction (P < 0.05), so data from irrigation events, natural rainfall events, and time from the beginning of precipitation to initial runoff are presented after averaging over all collection dates. Time from initial precipitation to runoff was also examined by analysis of variance for a randomized complete block design (P < 0.05).
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The delay from the beginning of precipitation to runoff of 4 (irrigation) or 2 (rainfall) min resulted in a minor reduction in nutrient losses compared with the reduction resulting from the lower runoff volumes. However, based on the rate of nutrient loss at the end of the irrigation event and after the first 60 min of natural rainfall (Fig. 2), the graduated buffer reduced N loss by 12.4 g ha–1 during irrigation [(11.4–8.3 g ha–1 min–1) x 4 min] and 0.8 g ha–1 during natural rainfall. The graduated buffer reduced P loss by 14.0 g ha–1 during irrigation and by 1.8 g ha–1 during natural rainfall by extending the time from precipitation to runoff (Fig. 3).
|
|
|
Nutrient Losses
Fertilizer losses in runoff were small compared with fertilizer applied. On average, across treatments, 1.5% of the N applied was lost to irrigation runoff and 0.5% to natural rainfall runoff in the first 60 min after runoff began. Irrigation runoff caused a 5.5% loss of P and natural rainfall runoff caused a 3.3% loss of P during 60 min of runoff. These results compare favorably with the results of other researchers, although studies differed in objectives and methods, and further support the contention that turf has a positive influence on the reduction of nutrient losses from runoff. Shuman (2002) collected no more than 0.6% of applied N as NO3–N and 10.9% of applied P as PO4–P in runoff from turf. Easton and Petrovic (2004) collected 6.2% of applied N as NO3–N (calculated from author's tables) during a year of natural rainfall runoff when a urea-based mixed fertilizer was applied.
The fertilizer application methods that were used for the irrigation experiments in this study were established to provide worst-case conditions. According to Titko et al. (1987), the N losses due to the volatilization of granular urea during the first 4 h after application are small regardless of environmental conditions. Titko et al. (1987) reported less than 0.3% volatilization losses 4 h after granular urea was applied to turf at temperature as high as 32.2°C, at relative humidity of 68%, and when no irrigation was applied to "water in" the fertilizer. Mancino et al. (1988) found that losses due to denitrification were low when granular urea was applied to turf on soil that was less than 80% of saturation. During that study, less than 0.4% of the applied urea-N was lost to denitrification in 10 d. Consequently, it was estimated that N losses during the 4-h interval between application and irrigation did not exceed 1% of the N applied regardless of environmental conditions.
Shuman (2004) demonstrated that light irrigation following fertilization reduced nutrient losses, and Walker and Branham (1992) stated that, as the period between the first runoff event and fertilizer application is extended, a greater proportion of nitrogen will be immobilized by plants or soil or leached past the active mixing zone reducing nitrogen runoff. Because of these and other recommendations, golf course superintendents generally do not apply fertilizer within 48 h before predicted rainfall and nearly always "water in" the fertilizer to minimize possible losses. The nutrient losses in the irrigation portion of this study are representative of a worst-case scenario and are likely to be more severe than what typically occurs.
The reduced runoff volume resulting from the use of the graduated buffer compared with the single buffer caused a significant reduction in the rate of N and P lost to both irrigation and natural rainfall runoff (Fig. 2 and 3). During irrigation the peak rate of N loss occurred at 35 min and the peak rate of P loss occurred at 30 min. The graduated buffer significantly reduced the peak N loss rate from 14.4 to 11.5 g ha–1 min–1 during irrigation and caused an 18% reduction in cumulative N loss (mean flow rate x 60 min) during 60 min of runoff from 603 to 493 g ha–1. The peak P loss rate was reduced from 29.2 to 24.7 g ha–1 min–1 helping to reduce the cumulative P loss by 14% from 1127 to 969 g ha–1.
The pooled results of natural rainfall runoff indicated that the peak nutrient losses occurred early in the runoff event at 10 min for both N and P (Fig. 2 and 3). The graduated buffer did not affect the peak N loss rate or the peak P loss rate when the peaks occurred only 10 min after runoff began. However, the graduated buffer caused a significant reduction in the nutrient loss rate during the 60-min runoff event resulting in a 17% reduction in cumulative N loss (218 vs. 181 g ha–1) and an 11% reduction in cumulative P loss (654 vs. 581 g ha–1).
There were no significant differences in NO3–N, NH4–N, or DRP concentrations between the graduated buffer and single buffer during either irrigation or natural rainfall over the course of the study (Fig. 5 and 6). During irrigation runoff, the average NO3–N concentration peaked at 0.3 mg L–1 following 30 min of runoff. The highest average NH4–N concentration was 3.4 mg L–1 at 30 min and the average peak DRP was 8.1 mg L–1 at 25 min of irrigation runoff. The average peak concentrations in natural rainfall runoff were 1.2 mg L–1 NO3–N (30 min), 0.9 mg L–1 NH4–N (30 min), and 8.0 mg L–1 DRP (25 min). Linde et al. (1994) observed similar results in their study of nutrient transport in runoff from creeping bentgrass and perennial ryegrass maintained as golf course fairways. They found that NO3–N concentrations in runoff were not affected by species, although time to runoff and runoff volume were significantly different.
|
|
This study found the same runoff activity, high nutrient runoff concentrations in the early stages of runoff followed by declining concentrations with time, suggested by Walker and Branham (1992). The N concentrations in both irrigation and natural rainfall accelerated rapidly from 5 to 25 min and were highest between approximately 25 to 35 min, causing rapid nutrient losses early in the runoff events (Fig. 5). The P concentrations also accelerated rapidly and were highest in both forms of precipitation at approximately 20 to 35 min (Fig. 6). The rapidly accelerating nutrient losses during those periods overcame the delay in time to runoff between treatments and effectively neutralized the beneficial effects of the graduated buffer during the initial stages of runoff (Fig. 2 and 3). After 20 to 25 min of runoff, nutrient losses were nearly equal among treatments in spite of the average 4- or 2-min delay in time to runoff caused by the graduated buffer and the greater volume of irrigation runoff from the single buffer treatment (Fig. 1, 2, and 3). Consequently, the graduated buffer did not affect nutrient runoff significantly for the first 30 to 35 min of runoff, but maintained an advantage following 35 min and until at least 60 min of runoff during both irrigation and natural rainfall. Assuming an average 37 min from the beginning of precipitation to runoff and sufficient precipitation to cause runoff, a rainfall event would have to last at least 72 min (37 min time to runoff + 35 min to significant runoff results) for the graduated buffer to make a significant difference in the amount of nutrient runoff that occurred.
Based on 55 yr of precipitation data collected at Stillwater, OK, an average of 81 rainfall events occurred each year (NCDC, 2005). Most of those events did not produce adequate precipitation to force runoff, but 7 events per year produced precipitation at an average rate greater than 13 mm h–1 (the surface infiltration rate at the research site) for a period longer than 72 min (the average time of precipitation required to produce significant differences in nutrient losses between buffer treatments). Consequently, the use of graduated buffers could make a meaningful difference in the amount of nutrient runoff during the 7 runoff-producing rainfall events that are likely to occur each year in Stillwater, OK, and additional reductions if excessive irrigation occurred. The average annual rainfall in Stillwater is 932 mm yr–1, a relatively dry climate compared with many regions of the world. The graduated buffer could make a greater difference in regions where rainfall runoff is more common.
The first 60 min of natural rainfall caused a slower loss of nutrients than 60 min of irrigation because the runoff volume from the natural rainfall was lower (Fig. 2 and 3). However, the P concentrations in the runoff from both forms of precipitation were nearly equal (Fig. 6). The N concentrations in natural rainfall runoff were lower than those in the irrigation runoff possibly because of volatilization, denitrification, and other forms of N degradation during the period between fertilization and natural rainfall (up to 3 wk). Torello et al. (1983), for instance, observed volatilization losses of 10.3% of the N in granular urea applied to Kentucky bluegrass (Poa pratensis L.) after 21 d at 24°C.
Despite Oklahoma's reputation for severe weather, a natural rainfall event that provides at least 51 mm h–1 precipitation for at least 1 h is considered a catastrophic event that occurs only once every 5 yr in Stillwater (Tortorelli et al., 1999). The DRP concentrations did not appear to differ between irrigation (51 mm h–1) and natural rainfall (mean = 16 mm h–1) despite large differences in precipitation rates and the resulting difference in runoff volume between the two forms of precipitation. Although the DRP losses were greater in irrigation due to greater runoff volume, the study suggested that DRP concentrations in runoff were not substantially affected by precipitation rate or runoff volume.
As expected, the graduated buffers caused significant delays in time to runoff and lower runoff volume regardless of whether the runoff occurred as a result of irrigation or natural rainfall. These results agree with our hypothesis that mowing at multiple heights results in multiple barriers that reduce runoff. A turfgrass stand is very dense, generally including 300 shoots or more per square meter. Because of this shoot density, multiple researchers have demonstrated and recommended grass buffers along crop production fields to reduce runoff. The dense shoot system in a grass buffer creates considerable resistance to water passage. A simple observation of turf following a severe rainstorm will indicate that runoff not only occurs through the shoots, but also occurs over the leaves. Areas of severe runoff are identified by the prostrate appearance of the turf. When runoff water from bare soil encounters a grass barrier, the runoff slows due to shoot resistance until sufficient volume accumulates to provide the energy necessary to bend the shoots and the lower leaves allowing the runoff to flow over or around the plants. We hypothesize that when the water encounters a second mowing height, a similar resistance occurs and sufficient volume must be accumulated to overcome this second barrier. During the study, a puddle of water formed each time the runoff encountered a buffer. The puddling was most noticeable at the interface of the fairway and initial buffer, but also occurred at the interface of each height increase in the multiple height buffers. Although turf density can be expected to increase with lower mowing height and have a negative effect on runoff (Linde et al., 1994; Easton and Petrovic, 2004), the work of Baird et al. (2000) indicated that when a buffer strategy is employed, the shoot height of the buffer vegetation had a greater effect on runoff than turf density. Baird et al. (2000) reported that a 76 mm buffer height was more effective for reducing water runoff than a 38 mm buffer in spite of the tendency for increasing turf density with decreasing mowing height. Multiple mowing heights result in multiple barriers that slow runoff and reduce runoff volume.
According to Baird et al. (2000), increasing the height of a vegetative buffer from 38 mm to 76 mm reduces runoff. Consequently, increasing the height of graduated buffers may cause higher reductions in runoff compared with those reported by this study. However, increasing the mowing height of bermudagrass golf course rough to 76 mm or more is not always practical. A survey of Oklahoma golf courses (n = 47) in 2004 indicated that the maximum mowing height of bermudagrass rough ranged from 19 to 100 mm with only six courses mowing bermudagrass rough at 76 mm or more (unpublished data). The 41 remaining courses maintained a mean maximum mowing height of 49 mm and a median mowing height of 51 mm in bermudagrass rough. Although high-cut bermudagrass rough could effectively reduce water runoff, golf courses must also maintain adequate playability. Dense bermudagrass rough mowed at more than 51 mm makes finding golf balls difficult and slows play considerably. The graduated rough described in this study could reduce nutrient runoff while maintaining playability.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Less than 2% of the N and 6% of the P applied as fertilizer were lost to irrigation runoff that occurred 4 h after fertilization. These losses, however small, were great enough to potentially cause unacceptable nutrient concentrations. The establishment of graduated buffers along golf course fairways and other turf areas could make a significant difference in the amount of N and P entering surface water in runoff during both excessive irrigation and natural rainfall.
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication February 3, 2005.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
