Pesticide Research Results - 1993 to 1996

In the following discussion and series of graphs (figures 1-4), we will attempt an overview of the complex research results.  NB! Each pesticide discussed has a separate sheet of graphs with four mini-diagrams (a,b,c,d) on the sheet (figures 1-4). 

On each sheet :

• The graph 'a' shows measured amount of pesticide residue dissolved in the creek water (in nanograms per litre which is indicated by the stars and dotted lines). These levels are compared to the creek's water flow volumes from March to November  (in cubic meters per second indicated by the solid bars in the background).
• The graph 'b' shows precipitation amounts (in milimetres) measured during the rainfall season.
• The graph 'c' shows the measured amount of pesticide residue in the rainfall.
• The graph 'd' shows measured amount of pesticide residue in air.
• We have also provided sheets, (figures 5 , 6) which compare the measured amount of pesticide residue dissolved in the creek water, similar to 'a' (above).  The four mini-graphs on each sheet of "A" - "D" represent 1996 -1993 respectively.

Over four field seasons, from 1993 to 1996, the temporal trends of three phenoxy- herbicides (2,4-D, dichlorprop and MCPA) and one phenolic herbicide (bromoxynil) were determined for the South Tobacco Creek watershed. These were examined in the surface water, air and rainfall during this period.

1) Herbicides in Surface Water
Runoff conditions are important to the analysis of herbicides in the surface water. After the spring snowmelt in 1994, the field season was relatively dry, with no runoff during the growing season. However, after the field season — in October — two significant rainfalls produced overland flow within the watershed.

The 1995 field season was not as dry as the previous year, with some flow in South Tobacco Creek except for two short intervals when there was no flow. South Tobacco Creek was flowing for the entire 1996 sampling period as well as for the 1993 season.

From 1993 to 1995, 18-litre water samples were collected either weekly or biweekly over the period from spring breakup to autumn freeze-up. In 1996, two-liter samples were collected every second day during the peak local pesticide application period. This method allowed the researchers to capture and identify a number of peak concentration occurrences that may have been averaged out, or missed altogether, using the weekly or biweekly sampling techniques. Concentrations in herbicide levels in creek water can change dramatically within a few days. The more frequent sampling procedure provided a more accurate monitoring of the timing and magnitude of changes in herbicide concentrations in surface water. 

Surface Water Results
All four herbicides monitored in the study were present in surface water, but in concentrations well below the Canadian water quality guidelines for the protection of freshwater aquatic life.
 

Maximum Pesticide Concentrations of Four Herbicides Detected in South Tobacco Creek, 1993-1996   Versus - Canadian Water Quality Guidelines   (CWQG) (Measured in ng/L or parts per trillion)
Chemical	Freshwater Aquatic Life Guidelines	Max. Concentration Observed
Bromoxynil	5000	41.6
2,4-D	4000	680
Dichlorprop	(None Established)	18.3
MCPA	4200	44.2

Each of the four major herbicides was found at elevated levels during spring runoff in March. In 1994 and 1996, the levels dropped to the point where they were nearly undetectable in April (figures 1 - 4). However, this pattern was not observed in 1995. Herbicide levels in the 1995 spring runoff were unusually low. In fact, MCPA residues were below the detection level. This was likely due to late October rainfall in 1994 which may have washed or flushed the residual herbicides out of the watershed in the fall (figures 5, 6).

As might be expected, the maximum herbicide concentrations in surface water were generally found during and closely following local application times. Rapid changes to the amount of herbicide residue in surface water were observed beginning in late May or early June and well into July (figures 1- 6).

A correlation between herbicide concentrations in surface water and residues in the air and rainfall was determined. Elevated levels of all the herbicides -- 2, 4-D, dichlorprop, MCPA and bromoxynil -- in creek water in 1993-1996 were observed during periods when there was not enough precipitation to produce overland runoff. However, during this period of maximum concentrations of herbicide residues in the creek water, elevated levels in herbicide concentrations were also observed in air and precipitation samples (figures 1- 4).

Much higher concentrations of MCPA were observed in the precipitation in 1996 as compared to previous years (figure 7). As well in 1996, the creek water herbicide levels were higher, similar to 1995. This is likely related to the new higher sampling frequency which enabled peak levels to be identified and measured which otherwise may have passed through the system unrecorded (hit and miss), using the weekly or bi-weekly sampling technique (figure 5,6). 
 

Initially, it was estimated that the quantities of herbicides discharged from the watershed via South Tobacco Creek was less than 0.01 percent of the total amount applied throughout the watershed. However, after applying the higher frequency sampling techniques in 1996, estimates were revised. Based on the new sampling, estimated discharges were increased to about one percent (1.0%) of the quantities applied in the watershed.

Click on any of the charts below for full details.

Figure 1: Dichlorprop Levels	Figure 2: 2,4 - D Levels
Figure 3: MCPA Levels	Figure 4: Bromoxynil Levels
Figure 5	Figure 6
Figure 7: Precipitation MCPA Concentration Levels

2) Herbicides in the Air and Rainfall
Concentrations of 2, 4-D, dichlorprop, MCPA and bromoxynil were at their maximum in both rainfall and air samples during periods of local use. However, these chemicals persisted for only short periods, although bromoxynil remained in the atmosphere longer at a higher level than the others (figure 1 - 4).

Phenoxyacid and phenolic herbicides found in air samples were not related to the origin of the air mass over the South Tobacco Creek watershed during the sampling events, but more closely related to local application activities.

The origin and movement of the air masses over the study area were determined for two events — June 13 and 19, 1994 — when the air sampling showed elevated phenoxyacid and bromoxynil levels. During the June 13 sampling, the air mass over the watershed was slow moving and had been situated over the area during the 12 hours prior to sampling (figure 8). However, the air mass over the watershed on June 19 had moved into the area from South Dakota during the 12 hours previous to sampling (figure 9).  Since the concentrations of phenoxyacid herbicide residues were high in both instances even though the air masses were from different locations, it is logical to assume that while long-range transport may contribute to phenoxyacid concentrations in the air, local sources appear to be the major contributor.

Atrazine
Atrazine was detected in the air, rainfall and creek water during the study. (figure 10)  Atrazine was also detected in water samples at the Twin Watersheds in October 1994, even though it was not applied within the watershed (figure 11). The most likely explanation for this persistent detection is through dry deposition, gas exchange and rainfall contacting the soil and water surfaces. 

Work conducted from 1993 to 1995 found that residues of atrazine in the South Tobacco Creek watershed were associated with the origin of the air mass over the sampling point during the sampling. For example, during the period of maximum atrazine detection in the air samples in 1995, the air masses over the region had moved in from the Gulf of Mexico, across the U.S. Midwest corn belt and into the study area (figure 13). Atrazine is commonly used for weed control on cornfields, and its heaviest use is in the U.S. corn belt.  Lower atrazine concentrations were generally observed when the air mass over the sampling sites had entered the area from non-source or lower use regions (figure 12).




Click on any of the charts below for full details.

Figures 8 and 9	Figure 10
Figure 11	Figure 12 & 13


 
 
 

Spring runoff begins in South Tobacco Creek