Print This Page

Fertilizer Recommendations Dramatic Changes for Austin


A new study commissioned by the City of Austin and conducted by Texas A&M, is expected to affect the entire state of Texas as awareness increases about our limited and vulnerable water resources. The study recommends new residential lawn fertilization practices, changing those that have been promoted state-wide for twenty years. In 2001, the City of Austin Watershed Protection and Development Review Department and Texas Cooperative Extension launched a pilot program, Stillhouse Spring Cleaning, to improve water quality in the Stillhouse Hollow area. Water quality data for the local spring, which is located in the environmentally-sensitive recharge zone of the Northern Edwards Aquifer, showed that nitrate levels were some of the highest in the city.

High nitrates can cause algae blooms which deplete the water of oxygen and can lead to fish kills and other risks to aquatic life. Since high nitrates in water are frequently associated with fertilizer use, the pilot program's goal was to educate the 300 Stillhouse neighborhood families on the earth-wise gardening practices promoted by the partnership's existing county-wide Grow Green program.

Soil tests of each of the pilot study yards included an assessment of nitrogen, phosphorous and potassium, the basic ingredients in fertilizer. The results were a surprise in that the majority of the yards had low to very low nitrates, but extremely high phosphorous and potassium levels. Instead of recommending high nitrogen fertilizers which could degrade the spring, the City decided to contract with Texas A&M to do a greenhouse study comparing nine different fertility treatments. The resulting lawn plots were judged for appearance as well as the amount of nitrogen, phosphorous and potassium that "leached" or filtered through the soil to groundwater. The findings have resulted in dramatic changes to both the quantity and types of fertilizer recommended.

Standing Recommendations:

Test soil to determine individual lawn needs. Know the square footage of your lawn and calculate the amount of fertilizer needed. Use only that amount. Don't Bag It! -- leave your clippings on the lawn. This practice results in at least 60% of the clipping's nitrogen and 100% of the phosphorous being available to the grass within the growing season. Additionally, iron and other important micronutrients are returned to your soil. New Earth-Wise Recommendations: Choose carefully -- certified organic and other labeled, natural fertilizers out-performed inorganics in the study. There was substantially less nitrogen leaching to groundwater in the organically- fertilized test plots. After a slightly slower start, the grass was denser and more attractive than those treated with inorganic fertilizer. The inorganic urea-coated, slow-release fertilizers evaluated in the study did not actually release nutrients in small increments as expected. (Other slow release coatings will be evaluated.) Other research supports the ability of organics to retain soil moisture and decrease runoff during heavy rainfall events. If using inorganic fertilizers, apply multiple, smaller applications (for a total of ½ lb. Nitrogen per 1000 square feet) per season. Larger amounts will run off or leach to the groundwater and your fertilizer will be wasted. Choose fertilizers with low phosphorous values because Austin soils are naturally high in phosphorous. (Fertilizer is based on a ratio of Nitrogen-Phosphorous-Potassium so phosphorous is the second number on the fertilizer bag..) 9-1-1, 8-2-4, and 6-1-1 are among the recommended options. Excessive phosphorous can block the availability of iron to your plants, causing yellowing. If your soil test shows: Use: Low to Very Low Nitrogen in Soil ½ lb. nitrogen/1,000 square feet 2 times/year Moderate Nitrogen ½ lb. nitrogen/1,000 square feet 1 time/year High to Very High Nitrogen DO NOT FERTILIZE

If you must fertilize without a soil test, never apply more than a moderate nitrogen rate or fertilize more than once a year. These new fertilizer recommendations are substantial changes from previous ones. First of all, while soil tests have long been recommended, few homeowners knew or used these services to determine existing nutrients in the soil. As we've seen in Stillhouse, there are considerable differences in soil nitrogen from yard to yard and fertility treatments should take that into account. Tony Provin, the A&M soil scientist leading the study, originally expected the synthetic fertilizers to look better than the organic but described a "180 degree change" in his opinion. The organic treatments fared better across the board on color and density. And now that the organic fertilizers are in a pelletized form, they are as easy to use as the synthetic fertilizers. While organics are often more expensive than synthetics, the fact that we are recommending substantially smaller application rates should more than balance out the cost.

Findings on the sulphur-coated urea, one form of slow release fertilizer, were also a surprise. It's been long held that it's environmentally beneficial to use these products, but due to the warmer central Texas heat, apparently the soil microbial activity breaks down the outer coating quickly, reducing its benefit. Also, a close look at a slow release fertilizer label reveals that only 5-10% of the product is actually coated for slow-release. The tests included a 41-0-0 product that is 100% slow release (not available to homeowners) and it showed very little benefit over the other synthetic options. Texas A&M University is planning additional nitrogen fertilizer source and rates studies. These studies will further help evaluate the slow-release benefits of other slow release coatings, polymers and formulations.

These recommendations include, but also go well beyond, the findings of the fertility study. They represent the collaboration of water quality and soil scientists and horticulturists who have compiled years of experience and study results to put together more inclusive, environmentally-responsible options for yard care.

Austin City Connection - The Official Web site of the City of Austin Contact Us: or 512-974-2446. Legal Notices | Privacy Statement © 1995 City of Austin, Texas. All Rights Reserved. P.O. Box 1088, Austin, TX 78767 (512) 974-2000

---------- Evaluating Potential Movement of Nitrogen and Phosphorus in City of Austin Soils Following Varying Fertility Regimes: Greenhouse Simulations Issue:

The Stillhouse watershed in the Austin, TX area has significant impairments due to nitrate-N and phosphorus levels. The source of these nutrients is postulated as urban/homeowner turfgrass fertilization.

Objective of study:

1) Assess the potential for off-site movement of nitrogen and phosphorus from various organic and inorganic nutrient sources. 2) Assist the City of Austin and Travis County Cooperative Extension in the development of environmentally-sound turfgrass fertilization recommendations.


Lysimeter boxes were constructed using pressure treated SPF 2"x6" lumber and ½" pressure treated plywood. Each lysimeter was glued and screwed together and all joints/seams were sealed with silicon sealant to prevent side and bottom seam leakage. The upper sides and back pieces (2"x6" lumber) of each lysimeter were cut at a 10 degree angle, sloped to allow water hitting the lumber to flow to the outside of the box. The front of each lysimeter was trimmed to 5", with a ¼" hardware cloth insert installed to the original 5.5" height. The effective soil capacity was 2073.5 cubic inches (W=13", L=29" D=5.5"). A covered sheet-metal weir was fabricated and glued/sealed with silicon onto the front of each lysimeter. This lumber angling and covered weir design allowed for a 13"x29" effective runoff area. Each lysimeter box was bench mounted at a 3-degree slope. This slope, in conjunction with a glass wool-lined recession located in the bottom-front of each lysimeter allowed for easy collection of leachate water.

Two collection boxes for each lysimeter were located beneath each lysimeter to capture leachate and runoff generated during each leaching event. Overhead sprinkler irrigation was installed to simulate a 1" per hour irrigation/rainfall event. This system, also used daily to provide needed irrigation water needs, was fed by a large capacity reverse osmosis water purification system. Routine tests of the reverse osmosis water were conducted to verify the quality of this treated water. All primarily ions of concern (Ca, Mg, K, P, S and NO3-N) were below detection limits. Sodium levels never exceeded 2 mg/L-1.

Two soils from the Austin area were collected, sieved in a moist state to pass a 2 mm opening, mixed and placed in greenhouse lysimeter boxes. All rocks and consolidated materials larger than 2 mm were discarded. The soils, described in Table 1, were an east Austin flood plain soil (Bergstrom silty clay loam) and a west Austin limestone/stone dominated soil (Speck clay loam). One and a half inches of the Speck clay loam was placed in the bottom of each lysimeter box, followed by 4 inches of the Bergstrom soil. This arrangement was performed to simulate the common practice of placing "topsoil" over the limited thickness of soil in west Austin. The 2" DilloDirt incorporation treatment was four inches of a mix comprised of equal volumes of Dillo Dirt and Bergstrom soil mixed in a cement mixer.

The following fertilizer treatments were used:


21-0-0 ammonium sulfate (synthetic) 8-2-4 (certified organic) 2" incorporated DilloDirt, (unlabeled natural) DilloDirt at N rate (unlabeled natural) 41-0-0 100% sulfur coated urea (synthetic) ½" DilloDirt-surface application (unlabeled natural) 13-13-13 (synthetic) 15-5-10 (synthetic) 9-1-1 (labeled, natural) Control-no fertilizer or grass planting, but wetted and incubated for 1 month prior to leaching event (See figures 1 and 6). Treatments 3,6 and 10 were performed at the beginning of the experiment and were not repeated. Treatment 10 was conducted to assess baseline nutrient loss potential (Table 1). All other treatments were performed every two weeks and based on application rates of 1 pound of nitrogen per 1000 square feet. This schedule was postponed by 1 week following each leaching to ensure favorable moisture conditions were present. Immediately prior to this fertilization, clippings were collected for yield and nutrient content. These clippings were not returned to the lysimeters. This removal of 1 week of grass re-growth following an event differed from the weekly clipping of the lysimeters where all clippings were allowed to accumulate on the soil surface. Prior to the application of all treatments (except #3), the lysimeter boxes were seeded to common Bermuda grass. Treatments were initiated upon 75% grass coverage (see Table 21). Initial plans to use St. Augustine grass were changed due to concerns over the heavy clay in sod available in the College Station area, as well as, limited previous success in greenhouse St. Augustine grass studies. The use of washed plugs was addressed, however it was decided that it would take 4-5 months for the slow growing St. Augustine grass to fill in.

A limitation of available space within the greenhouse prevented four replications of each treatment. Subsequently, two identical three replication studies were conducted in two different greenhouse rooms. While identical methodologies were used, the studies differed in time (two months) and temperature (due to time and greenhouse) (see Table 22) and therefore are reported separately as Study 1 and Study 2.

Each lysimeter was exposed to two, one-hour rainfall events 28 days following the initial fertilization and/or after the first fertilization following the previous rainfall event. These rainfall events were generated using overhead sprinkler nozzles with 300 percent overlap. Catch cups were placed between each lysimeter to monitor the amount of water reaching the lysimeter. Each rainfall event simulated a 1-inch per hour rainfall. No significant differences were observed in distribution output and collection.

The runoff and leachate from each lysimeter was measured for volume and total nitrogen, nitrate-N, phosphorus and other water-born elements. Extremely limited runoff was measured during these rainfall events; subsequently primary focus has been directed toward leachate data, however runoff nutrient masses were used when determining total nutrient mass losses. The limited runoff was likely due to 1) porosity of the loamy soils used and/or 2) the significant rooting system of the grass that formed root channels and macro-flow channels. The subsequent analysis of soil cores, showing distribution of nitrate-N and P with depth, taken for the center of the lysimeters helped alleviate concerns over sidewall flow. Furthermore, each lysimeter was maintained at near field capacity to prevent soil drying and creation of open spaces along lysimeter sidewalls. Weekly, supplemental water was added to each lysimeter based on theta-probe measurements.

Data reporting and report terminology

The following terms were selected and used during the development of this report:

Event - the runoff/leaching event, which includes two 1-hour time frames with samples collected on an hourly basis

Time - a 1-hour time frame during an event where individual samples were collected for analysis

Leachate - the water which was collected after passing though the bottom opening of the lysimeter

Runoff - the water which was collected after flowing across the weir installed on the down slope side of the lysimeter

N - nitrogen

P - phosphorus

nitrate-N - nitrogen in the nitrate form

ppm - parts per million

mg - milligrams

mg/L - milligrams per liter, used for expressing ppm in liquids

mg/kg - milligrams per liter, used for expressing ppm in solids

box - a lysimeter

TKN - total Kjeldahl nitrogen which includes all nitrogen components except nitrite-N and nitrate-N

Total N - summation of TKN and nitrate-N

Salinity - a 2 part water:1 part soil mixture which electrical conductivity is expressed as ppm salt

Soil test phosphorus - Texas Cooperative Extension routine soil phosphorus test using acidified ammonium acetate + EDTA

Total phosphorus - total phosphorus determined in a modified TKN method.

The report has been segmented into three major sections including; discussion, tables, and figures. This report has focused on highlighting statistical differences observed as a result of the nitrogen and phosphorus treatments. Significance was determined using ANOVA and Tukey mean separate test methods with a probability cutoff level of 0.05 (95% of difference addressed in the ANOVA).

The figures are simply a graphic representation of the data from the tables, and were included to better illustrate differences between treatments, events, time and depths

Data Discussion

Clipping Data

Clipping data is often used as an indicator of grass performance and health. Grass, which is not actively growing, is often assumed to be a risk for increased water runoff and decreased nutrient utilization. Since irrigation water was provided and soil moisture was maintained at appropriate levels, differences in grass growth is then attributed to nutrient or salinity imbalances. Excessive grass growth, often attributed to excess nitrogen availability, can also be a derogatory factor, as plant disease, environmental and homeowner maintenance issues may arise. Making the assumption that a well growing, properly fertilized lawn will produce 2.5 lbs of dry grass tissue per 1000 sq. ft. each week, we calculated that 1-week of re-growth should yield approximately 2.8-3.5 grams of dry tissue. Overall, we exceeded this growth rate (Table 2), with the 2" Dillo Dirt rate being the highest while the DDN and 41-0-0 treatments were the lowest.

Initially, off color grass was observed for the lower Dillo Dirt treatments and the 8-2-4 and 9-1-1 organic treatments (Table 23). The off color organic treatments were primarily limited to first month of growth. The subsequent three months of growth produced a relatively green grass. This initial lag period was likely an effect of immobilization of available N by soil microbes as decomposition of more resistant carbon in the Dillo Dirt occurs. By the fourth month, the organic treatments generally outpaced the inorganic treatments in grass color quality (Table 23).

Clipping phosphorus levels were also evaluated. While overall, the levels were adequate, two issues were observed. First, the P concentrations from the later clippings from the 21-0-0 treatments were below the desired 1900-4500 mg P/kg range. Part of the lower P concentrations in this treatment may be caused by increased soil salinity observed in the post-study soil analysis (Tables 20-22). While these levels are not excessively high, it is likely that high sulfate concentrations may have limited phosphorus uptake in this ammonium sulfate fertilizer treatment. Chemical analysis of grass clippings suggests that no micro-nutrient deficiencies occurred during the study.

Leaching and Runoff Data

The data from the control plots (treatment 10) is illustrated in Figures 1 and 6. The boxes were only leached during the first two events. Based on the very low nitrate-N and non-detectable P levels during these two events and high hand weeding requirements, it was decided to abandon these boxes. An additional set of boxes with only bermuda grass (no nitrogen) were lost to poor grass growth, likely to insufficient N.

No significant differences were found in leachate volumes. While additions of large amounts of organic matter have been shown to increase water holding capacity, the need to maintain the lysimeters at near field-capacity (to minimize sidewall shrink) negated any potential observations differences. The limited volume of runoff failed to provide any significant trends.

The loss of nitrate-N has been of particular interest to the City of Austin. While the concentrations of nitrate-N (Tables 7-8) leaving the system are of importance, the overall mass loss of nitrate-N (Table 11-14) leaving a system is needed when for use in watershed models and determination of watershed loading. Over both studies, the use of inorganic N fertilizer resulted in significantly higher nitrate-N and total N losses. Overall, significantly lower nitrate-N and total N losses were observed when an organic fertilizer was used, with the exception of the 2" Dillo Dirt treatment in study 2. For comparison purposes, each 110 mg of nutrient lost represents 0.10 lbs of nutrient per 1000 sqft.

The N rates for this study were based on the application of 2 lbs of N per 1000 sqft, split applied over a period of two week, for each event. The leachate and total N loss data suggest this rate was excessive, even for the ideal growing conditions of a greenhouse. Our initial calculations/N rates were based on previous published nutrient need factors for greenhouse/field research. The return of clippings and the subsequent N cycling did lessen N needs, however the wide spread adoption of mulching mowers and the Don't Bag It Program and results of this study suggests that historical turf grass N recommendations need renewed scrutiny. Additionally, the lack of differences between 100% sulfur coated urea (41-0-0) and 21-0-0 suggests that fertilizers with sulfur coatings may not provide significant slow release benefits.

The loss of phosphorus from the lysimeters was largely related to the application rate of P. The initial soils used tested very high (Bergstrom) and moderate (Speck) for plant available P. The use of P containing fertilizer (Tables 15-18) significantly increased P losses. A review of Table 20 illustrates the influence/change of soil test P levels caused by the different treatments. While the absolute mass losses of P are significantly lower than those for nitrate-N and total N, the target level for P in slow moving streams is approximately 75-100 times lower than nitrate-N. Additionally, existing research suggests that once soil P levels become excessively high, potential phosphorus mobility may exist for years.

Soil Analysis

The analysis of soil samples following the completion of the greenhouse study was conducted via 4" cores which were segmented into 1 inch increments and prepared for laboratory analysis. A time frame of 12 days from the last rainfall event was required for drying before sampling/coring could occur. During this time frame, some additional mineralization of organic carbon/nitrogen sources did occur and likely skewed the nitrate-N levels upward for the Dillo Dirt treatments.

Overall, total N levels (Table 19) increased based on application rate. The inorganic fertilizers, 8-2-4 and 9-1-1 were not significantly different from one another, however the higher loading rates for the Dillo Dirt treatments were reflected in the TKN levels. Similar trends were observed for the phosphorus containing treatments. Tables 21 and 22 illustrate the mean soil test data for each treatment x depth increment. The soil test phosphorus data shows significant enrichment of the soil surfaces following addition of P containing fertilizers. The downward enrichment of lower soil layers, while not always statistically significant at the P-0.05 level, does strongly suggest against continual application of medium to high phosphorus containing materials to already high and very high testing soils.

Conclusions and Future Management

This study has resulted in the redirecting of some turf grass fertility research by Texas Cooperative Extension faculty. While significant questions continue to exist, enough information exists to direct homeowners, landscape professions, retailers and municipal employees on methods toward reducing potential offsite movement of nutrients. The Texas Cooperative Extension Soil, Water and Forage Testing Laboratory is developing new recommendations for environmentally sensitive lawns and urban landscape areas. For these environmentally sensitive urban landscapes, N and P fertilizer rates will be based on the results of a recent soil test followed by the following alterations:

Reduce nitrogen application rates to ½ lb per 1000 sq. ft. per application. Reduce annual nitrogen application rates for turf from 3-4 lbs per 1000 sq. ft. to 1 lb per 1000 sq. ft. If soil test nitrate levels exceed 20 ppm the first initial spring fertilization is skipped. If soil test nitrate levels exceed 40 ppm, no nitrogen fertilization is recommended that year. The first nitrogen application (½ lb N per 1000 sqft) should occur at or very near greenup in the spring. In general, this should be near the first of April. d. The relief of clippings from lawn will provide additional nutrients. However, if during the summer the grass lacks vigor, a second application could take place in July. Reducing the amount of nitrogen present in fall will reduce grass growth and lessen the potential for brown patch. For soils with very high testing phosphorus levels, avoid the use of phosphorus containing fertilizers. No phosphorus free organic fertilizer is currently available. If selecting organic fertilizers for a very high phosphorus testing lawn, select a product with as high of a nitrogen to phosphorus ratio possible. A high ratio will reduce additional phosphorus buildup. List of Figures Figure 1 Concentrations of nitrate-N in leachate recovered during leaching first two events. Figure 2 Concentration of nitrate-N in leachate recovered during leaching last two events. Figure 3 Mass of nitrate-N in leachate recovered during leaching first two events. Figure 4 Mass of Nitrate-N in leachate recovered during leaching last two events. Figure 5 Total N recovered in leachate and runoff from all events. Figure 6 Concentrations of phosphorus in leachate recovered during leaching first two events. Figure 7 Concentrations of phosphorus in leachate recovered during leaching last two events. Figure 8 Mass of phosphorus in leachate recovered during leaching first two events. Figure 9 Mass of phosphorus in leachate recovered during leaching last two events. Figure 10 Total phosphorus recovered in leachate and runoff from all events. Figure 11 Mass of clipping removed from treatments. Figure 12 Nitrogen content of clippings removed from treatments. Figure 13 Nitrogen yields by clipping removed from treatments. Figure 14 Phosphorus content of clippings removed from treatments. Figure 15 Phosphorus yields by clipping removed from treatments. Figure 16 Soil TKN levels at conclusion of study. Figure 17 Influence of treatment on TKN values and soil depth. Figure 18 Soil total phosphorus levels at conclusion of study. Figure 19 Influence of treatment on total phosphorus values with depth. Figure 20 Soil test phosphorus levels at conclusion of study. Figure 21 Influence of treatment on soil test phosphorus values with depth. Figure 22 Soil nitrate-N levels at conclusion of study. Figure 23 Influence of treatment on soil nitrate-N values with depth. Figure 24 Soil salinity levels at conclusion of study.

Austin City Connection - The Official Web site of the City of Austin Contact Us: or 512-974-2446. Legal Notices | Privacy Statement © 1995 City of Austin, Texas. All Rights Reserved. P.O. Box 1088, Austin, TX 78767 (512) 974-2000


Back to Organic Records Search