Runoff Characteristics from Row Crop Farming in Florida

A four-year monitoring project documents wet-detention pond efficiency at removing nutrients, metals, pesticides, and bacteria from agricultural runoff.

Agricultural runoff has been identified as a significant source of water pollution in theUnited States. Uncontrolled, runoff removes topsoil, nutrients, pesticides, and organic materials and carries them to water bodies where they become pollutants. Given the problems associated with agricultural runoff, staff at the Southwest Florida Water Management District (SWFWMD) initiated a research project to study the effectiveness of a wet-detention pond in treating runoff from farm fields. Two types of BMPs are generally used to reduce the impact of agricultural runoff: source reduction and end-of-pipe treatment. Common source reduction methods, such as conservation tillage and cover crops, were not observed at this site, but plastic mulch was a common technique employed during the growing season. Plastic mulch with raised beds is a conservative irrigation technique that reduces the need for herbicides and fertilizers, thus reducing contamination by these chemicals. But farmers still use biocides to control fungi, bacteria, and insects, and there is increased runoff caused by the impermeable plastic and compacted soil between rows. Another source control method observed was shallow swales at the end of the planted rows that collected runoff before it was diverted into the ditch. Ditches can slow down runoff water and help diminish the water's force, anchor the soil, and filter out sediments while detention ponds can provide final treatment by storing runoff in a controlled manner and treating runoff through natural processes between storm events. Both the ditch and the pond were monitored in this study.

This article presents the results from over four years of data collection (85 storm events as well as ambient grab samples). The project was designed to characterize an agricultural stormwater treatment system, and the data presented here include (1) measuring the reduction (or increase) of pollutants in storm runoff treated by a wet-detention pond, (2) determining water quality concentrations in a pretreatment ditch, and (3) comparing water quality and sediment samples for different years. The complete study, including additional data, is available from the author by request.

Method

Site Description

This stormwater monitoring project evaluated agricultural runoff in southern Hillsborough County near Ruskin, FL, and adjacent to Cockroach Bay, a part of the Tampa Bay estuary. The fields are irrigated using groundwater, and the crops were winter vegetables such as tomatoes, peppers, and lettuce. The detention pond is actually two ponds (cells) in series that are not connected until the pond level is above 0.61 m NGVD (2.0 ft.) and only solidly connected when the pond begins to discharge at 0.76 m NGVD (2.5 ft.). NGVD (National Geodetic Vertical Datum) essentially designates water levels in relation to mean sea level (Figure 1).

Instrumentation and Sample Collection

The hydrology of the basin was characterized by recording rainfall with a tipping bucket rain gauge and measuring surface water levels with float and pulleys and bubbler-type flow meters. Flow-weighted composite water quality samples were collected at the outflow by converting water level to flow with weir formulas. At the inflow, velocity meters installed in the large pipe were not always able to accurately measure flow, so composite timed samples were collected based on rainfall and flow was validated later using a water budget method. Sensors were connected to data loggers that stored the data and averaged the measurements at 15-minute intervals. Storm samples were collected automatically and stored in refrigerated units until retrieved, preserved with acid, and transported on ice to the laboratory for analysis.

Rainfall water quality was collected with a wet/dry precipitation collector that sampled wetfall only, and

the sample was immediately drained into a refrigerator until collected as previously described. Sample analysis used standard methods and followed the SWFWMD-approved quality assurance plan.

Sediment samples for nutrients, metals, and pesticides were collected in July 1997 at seven locations in the ditch, pond, and marsh and again in July 1998 for 10 locations. In July 2000 and July 2002, pesticide samples were analyzed for 10 sites, but nutrients and metals were not analyzed. Samples for each location were extracted intact from the sediments with a 2 in. diameter hand driven corer by using several cores in close proximity. These were well mixed in the field and then placed in EPA approved jars to be sent on ice to the Florida Department of Environmental Protection laboratory in Tallahassee.

Efficiency of the pond for removing pollutants was calculated by two methods, but only the results for loads are presented in this article. The load efficiency was calculated by adding the individual storm events together for a given time period (in this case yearly) using the following formula.

Loads are calculated by using Event Mean Concentrations (mg/lit. or µg/lit.) multiplied by flow volumes (m³) multiplied by conversion factors. These were calculated for each storm event and added together for SOL and usually have units of kg/yr. During the four years of study, over 85 storm events were sampled for load calculations and they represented greater than 80% of the storm runoff into the pond.

Results and Discussion

The wet-detention pond built to treat runoff from agricultural fields was monitored to evaluate its ability for removing pollutants and to better understand the dynamics of the system. Ambient samples taken between storm events quantified conditions in the ditch. Grab samples of chlorophyll and fecal coliform were taken in the ditch and the pond. Sediment samples were collected four times and represent the entire stormwater system.

Pollutant Concentrations

Individual Event Mean Concentrations collected at the inflow of the pond compare rain amounts and evaluate relationships between pollutants (Figure 2). El Niño produced above-average rainfall from September 1997 through March 1998, and La Niña was implicated in the below-average rainfall the following year. Although record droughts continued through 2001, several large storms and hurricanes helped make up the deficit at this site during the summer rainy season.

In general, all constituent concentrations tended to increase in early 1998 in response to the deluges produced by El Niño. Another spike was produced by increased agricultural activity in 2000-2001. Unlike the other constituents, nitrogen concentrations exhibit a much greater increase with more-intensive agricultural activity than during the El Niño storms. All the metals of concern (copper, iron, lead, cadmium, zinc) were correlated with each other (r=0.60 to 0.80) and are represented in Figure 2 by copper.

Metals and phosphorus concentrations were orders of magnitude greater during some of the El Niño storms. Part of the increase can be attributed to a breach through the berm between the shallow swale that usually collected water in the fields before it was rerouted and discharged to the ditch. Of special note for phosphorus are the high concentrations for all samples (avg. 1 to 2 mg/lit./yr.). It might not be surprising that phosphorus concentrations are much higher at the inflow than measured in urban development in Florida where untreated runoff usually ranges between 0.1 and 0.6 mg/lit. (Harper 1994, Rushton et al. 1997, and others), but phosphorus concentrations are even higher than other agricultural studies. Other studies measured average total phosphorus concentrations of 0.48 mg/lit. discharging from pastures, 0.14 mg/lit. from citrus, and 0.40 mg/lit. from row crops (Harper 1994).

Pollutant Loads

An analysis of pollutant concentrations is valuable for understanding processes taking place in the pond, but pollutant loads, which include both concentrations and water volumes, are a better measure for predicting impacts to receiving waters and calculating percent reduction of pollutants. Differences between years were affected by (1) the unseasonable amount of rainfall and pollutant concentrations induced by the El Niño rains in 1998, (2) the lower flows and pollutant concentrations measured in 1999, and (3) the increased flow and concentrations caused, in part, by more-intensive agricultural activity in 2000 and 2001 (Figure 3).

A comparison of the hydrology is shown in the first figure to demonstrate how much greater the impact of pollutant loads is than can be attributed to flow alone. The large pond area was able to reduce the volume of runoff water by 17 to 29% from the inflow to the outflow and 26 to 45% when rainfall on the pond is included as an input. The greater amount of rainfall in 1998 increased loads at both the inflow and the outflow. Another point of interest is the large percentage of nitrogen that enters the pond directly in rainfall. For ammonia, it is often a greater load than the loads discharged at the outflow. The increase in agricultural activity including irrigation contributed to the increased nitrogen and phosphorus loads in 2000 and 2001. The unseasonable rainfall caused by El Niño during the growing season contributed to the total suspended solids loads during 1998. The results show the pond is especially effective in reducing high loads before discharge to Cockroach Bay (Figures 2 and 3).

Percent Efficiency

The reduction of pollutant loads from the inflow to the outflow is shown in Figure 4. One of the goals of the State of Florida Water Policy (Chapters 62-40 FAC) is an 80% removal of pollutant loads by stormwater systems, especially of total suspended solids. This goal of 80% reduction is met for many metals in 1998, 2000, and 2001 and nearly so for inorganic nitrogen. However, organic nitrogen shows poor efficiency, and phosphorus only meets the 80% goal for 1999, the year of reduced agricultural activity. Copper meets the goal for all years but also fails to meet the water quality state standard (Chapters 62 - 302 FS) for all years. None of the years shows consistent results, and the wide variation in pollution removal demonstrates the differences that can occur in wet-detention ponds between years as hydrologic, agricultural, and other conditions change.

Ambient Samples in the Ditch and Pond

Water quality grab samples were collected between storm events for various constituents that could not be sampled using automatic equipment. Both the ditch and the pond were sampled for chlorophyll and coliform bacteria at the same time (Figure 5). There are similar patterns of reduced concentrations as water flows through the system for both chlorophyll and fecal coliform bacteria with much-higher concentrations measured in ditch stations 3 and 4 than in the other locations (see Figure 1 for sampling locations). Even though fecal coliform bacteria were decreased substantially by the time they reached the outflow, the concentrations still exceeded the levels considered safe for the propagation and harvesting of shellfish. The increased concentrations of chlorophyll at the pond inflow may be an artifact of increased total phosphorus and total organic nitrogen in the sediments at this location (as will be seen later).

Similar patterns were seen for phosphorus, suspended solids, and metals as were measured for chlorophyll and fecal coliform and are represented in this article by phosphorus and copper in Figure 5. Even though the data for metals and nutrients are not strictly comparable, because the storm samples in the pond were compared to the ambient samples in the ditch, the same pattern is noted. The trend would probably have been even more pronounced if grab samples had been taken in the pond at the same time as the ditch samples, because Event Mean Concentrations during storm events tend to have higher concentrations than grab samples.

Although the concentrations at the pond outflow were demonstrably reduced by a significant amount, nevertheless, the pond discharge water still did not meet water quality goals. For example, fecal coliform, iron, and copper were discharged above state standards, and chlorophyll and phosphorus indicated that the pond discharge water was measured in the eutrophic to hypereutrophic lake classification.

Sediment Samples

Concentrations in the sediments also showed some of the same patterns as the water concentrations. Samples were collected in July 1997 right after the pond was constructed and the ditch was recontoured, and again one year later in July 1998. Samples were also collected for pesticides, but not for nutrients and metals, in July 2000 and July 2001. All of the nutrients tended to have high concentrations in ditch stations 3 and 4. It should be noted that stations 3 and 4 were not measured in 1997. The elevated metal concentrations measured in the ditch can be partially explained by the poor flushing of water through that part of the ditch, the low dissolved-oxygen levels, and the use of pesticides to increase yield. The metals in the ditch that were measured above the possible toxicity level were copper, mercury, cadmium, and zinc, but none of these concentrations was above the probably toxic levels except for copper. Other trends of interest include the tendency for phosphorus and iron to concentrate in the center of the two ponds. Also of note was the dramatic increase of orthophosphorus from being barely detected in 1997 (1 mg/kg) to an average of 37 mg/kg measured in 1998 (Figure 6).

Advertisement

Pesticide measurements in the sediments (Table 1) identified chlordane and DDT derivatives as present in almost all soil samples.

Summary of Results

• Rainfall directly on the pond contributed 17 to 29% of the hydrologic input to the pond.
• The large storage capacity of the pond at the beginning of the rainy season helped reduce the annual volume that left the pond by 11 to 25% (26 to 45% when rainfall is considered an input).
• Significant differences in pollution concentrations were noted between years with extremely high concentrations of metals and pesticides flushed through the system during the El Niño storms in 1997 - 1998 and higher concentrations of nutrients and metals of concern with increased agricultural activity in 2000 - 2001.
• Concentrations of nutrients and metals were greatly reduced as water flowed through the ditch and out of the pond.
• Average concentrations in the ditch were more than three times greater for fecal coliform and more than five times greater for total coliform than samples measured at the inflow of the pond.
• Fecal coliform bacteria in the pond discharge water exceeded concentrations considered safe for the propagation and harvesting of shellfish at the outflow.
• The chlorophyll concentrations were reduced as water flowed from the ditch through the pond and over the outflow structure, but the pond still discharged water intothe marsh in the eutrophic to hypereutrophic range.
• High concentrations of metals were measured in the sediments in the ditch, and some concentrations, especially for copper, were possibly toxic.
• Pesticides measured in the sedi-ments identified chlordane and DDT derivatives as present in almost all soil samples.

Author's Bio: Betty Rushton, Ph.D., is an environmental scientist with the Southwest Florida Water Management District's Resource Management Department in Brooksville, FL.