University of Florida

Best Management Practices & Water Resources

For more information contact Samira Daroub.

Previous Research

Particulate P Demonstration Farm

The University of Florida's Everglades Research & Education Center created a BMP demonstration sugarcane farm at the research center in Belle Glade, Florida. The BMP farm was established to demonstrate to growers the operational differences between an optimized BMP sugarcane farm and a conventional BMP sugarcane farm.  The main objectives of the demonstration farm are to apply flow velocity and floating aquatic weed controls to effectively demonstrate to EAA growers the resultant P load reductions. 

Two hydraulically isolated sugarcane blocks of 125 and 200 acres each were segregated and equipped with identical drainage pumps and monitoring instrumentation to record rainfall, flow, canal levels and to collect discrete hourly drainage water samples.  Demonstration farm data, i.e., drainage volume, P species concentrations, total suspended solids, canal levels, flow velocities, and rainfall are collected and analyzed to assess and demonstrate the effectiveness of floating aquatic weed management and drainage canal velocity control on reducing particulate, dissolved, and total P farm loads.

Return to top

BMP Consultation Program

Project researchers consult with growers in the S-5A sub-basin on their farm specific BMP implementation techniques during the period of May 5 to June 6. The S-5A basin is comprised of a wide variety of growers, so this will provide us a diversified group to develop and improve our operational expertise.  There are currently 82 permitted structures and 53 farm basins in the S-5A basin.  From June 6 onward we will begin consultations with growers farming in the S-6 sub-basin.

The goal of the program is to achieve further basin P load reductions by improving implementation of existing BMPs. The consultation methodology consists of an initial contact and program explanation, one or more field visits to complete a farm BMP appraisal, a discussion with the grower of issues that confront implementation of BMPs on their farm, an analysis of information provided by grower, and a follow up visit for conveying recommendations. Extension materials developed by UF/IFAS that explain in detail the BMPs in the SFWMD's table are distributed when appropriate.  Summary reports of all consultations will be compiled, reviewed by a BMP advisory committee, and included in the project's annual report.

Return to top

Loxahatchee National Wildlife Refuge: Impact of Canal Sediments on Water Quality

This project integrates with existing monitoring projects in the Loxahatchee National Wildlife Refuge performed by other scientists.  This project is divided into two tasks: Task 1 is designed to monitor changes in the surface elevation of sediments within the L-40 canal prior to and following commencement of discharge from the STA-1E discharge pump and Task 2 is designed to study changes in water quality by sampling water as it flows along the L-40 canal.  It is conjectured that water flow induced by the STA-1E pump station or other sources may, at times, entrain sediments resulting in erosion of the sediment surface and deterioration of the water quality at downstream locations.  These studies will provide a better understanding of conditions that can result in deterioration of water quality.  These studies and associated activities may identify the need for adaptation of STA discharge operations or for additional structural changes (e.g. dredging).

Task 1: Hydrographic surveys on L-40 canal near the STA-1E discharge.

Five canal cross-sections, above and below the STA-1E discharge have been selected for this study.  Relative surface water elevation is determined by tapedown measurements from a mark on the interior permanent post to the water surface at each section during each visit.  Depth to sediment from the water surface is determined from a boat.  Initial measurements across each transect have bee done at 2.5 ft increments.  Subsequent surveys are being measured at five standardized locations (30, 40, 50, 60, and 70 ft from the interior post), across each transect.  Measurements will be performed at least six times per year over three years.

Task 2: Synoptic water quality research studies on l-40 canal.

This task will survey water quality and other relevant parameters in the L-40 canal on selected dates.  Each water quality survey will sample sites upstream and downstream of the STA-1E outfall.  These data will be used to evaluate the relationship of water quality to canal water velocity, as well as effluent and marsh water quality.  Where feasible the water quality survey will follow the Lagrangian (plug-flow) sampling design that makes observations and draws samples based on time-of-travel to each sampling site.

Return to top

Farm Verification of BMP Efficacy

Best management practices have been extremely effective in reducing P concentrations and loads emanating from farms in the EAA.  The farm-level reductions appear to be reflected in basin-level monitoring data produced by the SFWMD.  The BMP studies have shown that water management and crop rotation practices have the greatest effects on farm drainage water P concentrations and loads.  Relative to water management, it appears that achieving uniform drainage across a farm is an extremely effective practice.  Additionally, hydraulically blocking farms and using booster pumps can reduce P concentrations and loads by aiding in achieving uniform drainage, avoiding over-drainage, and disallowing the direct off-farm discharge of water from fields with higher P concentrations.  The crop rotation BMP is tied closely to the water management BMPs.  Rotating crops between hydraulically separated blocks allows for the redistribution of waters with high P concentrations to sugarcane fields.  It also ensures that crops that require more intensive water management are adequately cared for while not over- or under-draining the rest of the farm.

Rainfall adjusted farm-level load reductions, expressed as the SFWMD adjusted unit area load (AUAL), averaged 44.8% for the project sites.  Unit area loads (UALs) per unit of rainfall increased by an average of 10.7% based on WY 1993-1994.  Drainage pumping volume per unit of rainfall decreased by an average of 11.2%.  Total-P concentrations decreased by an average of 7.2%.  While these reductions were occurring at the farm-level, EAA basin-level AUALs decreased by approximately 50% and total-P concentrations declined by 7.8%.  Continued research leading to the development and implementation of BMPs is a requisite for compliance with Rule 40E-63.

Return to top

Effects of BMPs on Crops and Soils

A large-scale lysimeter demonstration project was started in December 1997 to understand the efficacy of various proposed BMP strategies and potential impacts of BMP implementation on long-term soil fertility and crop production trends.  The lysimeter site included 25 lysimeters: 11 smaller units dedicated to vegetable (crisphead lettuce) and rice cropping systems and 14 larger units planted to sugarcane.  The sugarcane lysimeter assessment was designed to demonstrate the effects of higher than traditional water table (WT) levels (that occur under BMP implementation) on 3 popular sugarcane cultivars as well as the effects of delivering nutrient-rich drainage waters (P-fertigation) to sugarcane.  The vegetable/rice lysimeter study was designed to demonstrate short- and long-term soil fertility and crop nutrient uptake trends for different vegetable/rice/flooded fallow crop rotations.  Drainage waters from the vegetable/rice lysimeters served as the P-fertigation source into select sugarcane lysimeters.

Averaged across all lysimeter treatments for the entire duration of the 36-month study period the vegetable/rice P export (122 lbs P/ac) was almost eleven times greater than the P amount exported by sugarcane (11 lbs P/ac). This large difference arose from the great difference in fertilizer P input into the two systems.  The vegetable/rice treatment received approximately 6.6 times greater fertilizer P input than the sugarcane treatments.

Water extractable P soil test values in the sugarcane lysimeters did not change appreciably over the course of the project and the treatment that included vegetable irrigation waters did not show any marked increase in soil test P. Water extractable soil test P levels for the vegetable/rice treatments increased while fertilizer P was being applied at high rates, but once water extractable soil P levels reached a plateau (-40 lbs P/ac), resultant fertilizer P additions were subsequently reduced, soil P levels decreased to approximately 30 lbs P/ac.

Depth of water table had no effect on P load exported from sugarcane grown under two water table regimes, 18 to 24 inch and 14 to 20 inch water tables.  The sugarcane treatment that received vegetable drainage water exported 2.7 times more P in drainage waters than sugarcane that did not receive vegetable drainage water. From this field lysimeter assessment it appears that routing vegetable and rice drainage waters through sugarcane fields is an effective practice to lower vegetable drainage water P loads, but is somewhat limited by the timing and intensity of the specific rainfall event and the stage of growth of the sugarcane receiving the drainage water.

Return to top

Specific Conductance in EAA Farm Canals

Specific conductance was monitored at ten EAA farms (12 pump structures). All data were collected using Hydrolab DataSonde (series 3, 4, and 4a) multi-parameter water quality data loggers. In order to identify the specific ions and ion ratios that comprise specific conductance, weekly grab samples were taken in 2001 and 2002 from eight farms (10 pump structures) and analyzed for ionic composition. Summary statistics showed that mean specific conductance above 1.275 mS/cm occurred at only two out of the ten farms monitored. Higher concentrations of sodium (Na+) and chloride (Cl-) were also observed at these two farms. Of the two farms, one also showed high levels of sulfate (SO42-).

Potential sources of specific conductance were evaluated. These included geological influences, drainage pumping, irrigation water and fertilizer application. Comparing average specific conductance data points of the study sites to historical Cl- concentration maps of shallow groundwater revealed that the highest average readings occurred on a farm located over historically high Cl- concentrations in 20-50 ft ground water. The farm with the second highest average specific conductance readings was located in an area that has wells of high Cl- concentration. The Na/Cl ratio in the farm canals ranged from 0.57 to 0.78. The Na/Cl ratio in seawater is 0.55. It has been reported that connate seawater underlies the area and exchanges with the surface water where canals are cut into the limestone. Shallow ground water hydrology and quality has a major impact on specific conductance in the EAA.

The effect of drainage pumping on specific conductance was variable and site specific. There was a low correlation between drainage pumping and conductance when all the sites were combined. Irrigation had a low negative correlation with specific conductance. Statistical analysis of the daily average specific conductance at three intensively monitored farms indicated that drainage pumping increased specific conductance at two farms, but not at the third. Irrigation decreased specific conductance at all three farms.

Previous research in the EAA indicated that potassium chloride (KCl) fertilizer application contributed less than 3% to the total dissolved solids (TDS) concentrations in canal waters. It is also reported that a sugarcane crop at harvest takes up more P and K from the soil than that applied by fertilizers. Our results show KCl fertilizer application in one of the high conductance farms with mixed cropping systems contributed less than 6.5% of the TDS in drainage water. This was calculated assuming that all the KCl fertilizer ended up in the drainage water which is highly unlikely as crops take up K+ and Cl- in large quantities.

Specific conductance in the EAA canals is strongly influenced by the composition of the shallow ground water, historically reported to be high in Na+ and Cl- due to connate seawater entrapment and the mixing of surface and ground water. The effect of drainage pumping was variable and site specific. Irrigation, in general, decreased specific conductance. Canal specific conductance is governed mainly by the quality and the hydrology of the underlying shallow ground water, which is farm specific. Fertilizers contributed a very small percentage to the total dissolved solids in the drainage water therefore had no substantial contribution to specific conductance in the EAA. Current P load reduction BMPs have reduced specific conductance in some locations in the EAA. It was the conclusion of this study that no further BMPs can be identified by additional research that would provide abatement of specific conductance in the discharge in the EAA. The issue of specific conductance in the EAA is a geological one, and shallow ground water is the major factor controlling the level of specific conductance in the EAA farm canals.

Return to top

Particulate Phosphorus Transport and Control

Particulate P accounts for 20% to 70% of the total P load exported from EAA farms and is frequently the cause of spikes in farms total P loads. The conclusions from our earlier studies suggests that a significant fraction of the particulate P in the EAA originates from in-stream biological growth rather than from field soil erosion. Recently deposited biological sediment material such as settled plankton, filamentous algae, and macrophyte detritus is the fraction that contributes the most to particulate P export. Exported solids may also be contributed directly from loosely bound material detached by turbulent shear forces of floating aquatic vegetation. Other contributions to particulate P loads come from submerged aquatic vegetation and planktonic growth. One of the primary goals of this study was to identify conditions that cause increased particulate P load rates, and analyze those conditions to determine operating procedures that may be optimized to reduce particulate P export, and therefore overall P export at the farm level. Load rate is the product of flow and concentration over a given unit time period.

The particulate P demonstration study was conducted on three farms in the EAA: a sugarcane farm in the northern EAA (UF9200A), a mixed-crop operation in the eastern EAA (UF9206A&B), and a sugarcane farm in the western EAA (UF9209A). Each pump station was fully instrumented, and data was continually recorded for key parameters such as rainfall, pump flow rates, and inlet and outlet water levels. All pump stations are equipped with ISCO 3700 portable automatic samplers that collect water samples every 15 or 30 minutes and composite them into one- or two-hour discrete samples for analysis. All collected samples are analyzed for total suspended solids (TSS), total P, and total dissolved P (TDP). Particulate P is calculated as the difference between total P and TDP. The complexity and the diversity of the systems included are considerable. The approach that has been adopted here was to conduct various forms of cluster analysis to attempt to identify primary parameters that have had the most impact on particulate P transport at the study farms. Event analysis was conducted on the fraction of the total P load contributed by particulate P for each pump station over the four-year study period. The annual contributions from the particulate P loads to the total P loads have decreased in two of the three farms in 2003. Particulate P at UF9200A decreased from an average of 50% over the last three years (2000-2002) to 28% in 2003. The particulate P load contributions of UF9206A increased from 26% in year 2000 to 36% in years 2001 and 2002, and decreased to 27% in 2003. Particulate P load contributions from farm UF9206B decreased from 40% in 2000 to an average contribution of 36% during the last three years. At UF9209A the contribution from particulate P to total P load was almost constant, around 67% in 2001 and 2002. In 2003, UF9209A pumped its canals lower and longer than previous years, causing more sediments to be dislodged from the bottom of the canal and transported out of the farm, resulting in a particulate P contribution of 80% to the total P load.

Load Distribution Analysis of the cumulative hydraulic and particulate P loads generated for each farm and year, showed that 50% of the annual particulate P loads was contributed by less than 25% of the hydraulic load. Process Distribution Analysis was conducted to determine the most probable mechanism for particulate P transport in the sub-events that contribute most to the annual loads, i.e. those in the top 50% of the load distribution. The objective of this analysis is to identify conditions that give rise to the increased particulate P transport events. The most distinctive pattern observed from this analysis is the number of farm-years that were dominated by few events. Data over the four-year study shows that six of the 15 farm-years sampled had a single event that contributed 30% or more to the top 50% particulate P load. Three farm-years had two events that contributed a total of 30% or more. Three farm-years had three events that contributed a total of 30% or more. Only two of the 15 farm-years had their load rates distributed such that it took more than three events to contribute a total of 30% or more to the top 50%.

Periodically, large volume (500-1000 liter) composite samples were taken at each of the study farms. These samples were concentrated by sedimentation in the field. The sediment solids were collected and further concentrated in the lab, after which they were analyzed for the same physical and chemical properties as the farm sediments, including bulk density, solids content, particle specific gravity, organic matter content, and P content. Selected samples were analyzed for particle size distribution and settling velocity distribution. The analysis from the concentrated suspended solids (bulk samples) showed that the exported suspended solids volume is relatively small when present in its settled state. However, the contribution to the total annual P load of the farm could be significant. Thus the importance of the suspended solids on the overall water quality of the farm must be considered when solids removal and control plans are being evaluated.

Farm Sediment Surveys were conducted with the objective of determining the P storage in the main canal sediments, to evaluate farm sediments properties, and to monitor changes in sediment character and inventory over time. Quarterly inventories were conducted of canal sediment volume, mass, and P content at each farm. A number of transect locations were set up at regular intervals upstream of the pump station at each farm. Canal sediment surface elevation and depth was determined at each location. Core samples of the sediment were taken at each transect, sectioned and analyzed for key physical and chemical parameters, including bulk density, solids content, particle specific gravity, organic matter content, and P content. The surveys reported cover the 22-month period from November 2000 through August 2002 for UF9200A, UF9206B, and UF9209A. It appears that there was a trend toward sediment accumulation over the study period at UF9206B and UF9209A, while at UF9200A sediment depth remained relatively constant. The P content (on a dry weight basis) typically decreases as depth increases, but the bulk density of the sediment increases as depth increases.

To understand the transport of particulate P in farm canals in the EAA it is necessary to identify and state the primary processes of movement. Major processes identified affecting particulate P movement were first flush, cumulative high velocity, restart flush, particulate phosphorus spike, and pump cycling.

First Flush

The first flush includes biological material accumulated during the quiescent period between pumping events. This highly mobile material causes an increase in the concentration of suspended solids during the first hours of pump events.

Cumulative High Velocity

Cumulative high velocity produces a steadily increasing discharge concentration of suspended solids as the water farther upstream has a longer time to accumulate eroded suspended solids as it moves downstream to the discharge pump station.

Restart Flush

Restart flush is similar to first flush. When pumping is terminated, suspended solids in the canal system settle out in place. If there has been a significant concentration of suspended solids in the downstream reaches of the canal system at shutdown, there will be a high initial concentration in the discharge when the pump is restarted.

Particulate Phosphorus Spike

Particulate P spikes occur occasionally. A particulate P spike is defined when the particulate P concentration for a particular sample is more than twice that of either the preceding or succeeding samples. The spike is assumed to originate from a random release of particulate material from upstream sources, such as a collection of floating macrophytes or a removal of a flow obstruction.

Pump Cycling

Pump Cycling differs from pump restart in that the pump cycles through on-off oscillations over relatively short time periods, e.g. 30 minutes to two hours. This condition occurs when a farm pump is on automatic on-off control that is tied to canal level.

The diversity of the farms has allowed a number of observations to be made regarding the importance of various operating parameters affecting particulate P loads. Dominant events started when pumping operations deviated from typical practices, but these deviations were specific to each particular farm. Following is a short description of recommendations to reduce particulate P loads.

Velocity in Canals

Velocity is a key control parameter for reducing particulate P export. Recommended velocities are relative, in that they must be within the operating framework of the configuration of the farm. Velocities should be as low as possible, and velocity excursions should be avoided, regardless of the average or typical velocity of the canal system. Velocities greater than 0.4 m/s (1.3 ft/sec) have been associated with greater transport rates at the study farms. Given the parabolic relationship between velocity and erosion, "slow and long periods" is preferred than "fast and short periods" for pumping a given volume of water.

Pump Cycling and Reduced Run Times

Long-run period cycling of about 8-16 hours, which reduces continuous pumping duration, has been shown to be beneficial in interrupting continued high velocity transport. This was evidenced on farms where the response time of the farm hydraulic system (i.e., the time required from pump start-up to the time when the equivalent of one volume of farm canal water is exported) is greater than the pump cycling period. Short period cycling of one hour or less is detrimental and should be avoided.

Level Control

Control of canal water levels is critical in avoiding major velocity excursions, and also to stay away from large deviations of the normal farm canal velocities. Lack of level control or major changes in minimum canal levels have resulted in dominant events at the two farms that did not practice strict canal water level control. Canal levels should be controlled to give minimum canal depths that do not exceed the maximum velocity recommendation.

Aquatic Weed Control

Weed control programs in the main canals is one of the most productive techniques in reducing the supply of high P content biomass. Physical removal along the entire length of the main canals is expensive to implement and not practical. For that reason, installation of weed-retention booms is recommended to be located at a distance >300 m (984 ft) upstream the main pump station. Spot spraying of weeds closest to the pump station is also recommended. Chemical treatment of major weed infestations will lead to the accumulation of transportable material into the bottom of the canal and is not recommended.

Return to top