CICEET Progress Report for the period 8/01/03 through 2/01/04

Project Title: Atmospheric Deposition of Currently Used Pesticides to the St. Jones River Reserve and Upper Delmarva Peninsula
Principal Investigator(s): Alba Torrents and Laura McConnell

Accomplishments
Scheduled Tasks:
For the reporting period, we had proposed three main tasks:

1. Extraction and analysis of remaining air and rainfall samples
2. Elucidation of effect of local pesticide usage on wet deposition
3. Investigation on the effect of temperature and relative humidity on vapor-particle partitioning of pesticides.

Progress on Tasks
Task 1 was conducted as scheduled. The sampling terminated on 9/16/03. All the rainfall and the air samples collected till that time have been extracted. All the samples have been analyzed for the presence of herbicides.

Task 2 has been attempted for one site­pesticide combination i.e., metolachlor for Horn Point.

Progress on Task 3 is being made. A GC has been adapted and columns manufactured to perform vapor-particle partitioning determinations. We are in the process of calibrating the experimental setup.

Additionally to the above mentioned tasks, in November 2003 we also conducted a seminar/workshop at Lewis were all our collaborators and researchers from the 3 different sites were present. We discussed the next steps on data analysis so it would be useful to on-site researchers, information transfer, and we intent to have another workshop sometime next fall in Delaware.

Difficulties Encountered
No major difficulties were encountered during this period. The sample extraction went smoothly. Only the insecticide analysis on the rain samples and air filters from 2003 and air PUFs from 2001-2003 remains. The vapor-particle partitioning experimental setup posed some challenge with regards to getting the parts together and working.

Anticipated Success in Meeting Project Objectives in Scheduled Project Period
Work is being conducted as planned.

Preliminary Results
Sampling and data Analysis
Analysis of herbicides for Horn Point from 2000-2003 reveals some interesting results. In terms of rainfall during the April-mid September, 2002 was the driest year whereas 2003 was the rainiest. However, 2001 had the least number of rainfall events. The average annual rainfall for Maryland Eastern shore is about 57 cm from April-September. This accounts for about 52% of the total rainfall that typically falls in a year. Thus, 2000 and 2003 were above average years in terms of rainfall whereas 2001 and 2002 were below average (See Table 1). As an approximation, there were roughly twice as many rain events from April to Sept in 2001 and 2002 than in 2000 and 2003.

The herbicides were generally not detected in precipitation after September. However, pendimethalin was detected in two samples in October 2001 and metolachlor in a few samples in October 2002. The herbicides were frequently detected in the samples from March, 2002 which suggests that either the herbicides are being applied earlier than believed or influence from sources further south. Metolachlor and atrazine were the most frequently detected herbicides. Overall, in all of the 193 samples collected from April-mid September over 4 years, metolachlor was the most frequently detected herbicide (in 70% samples) followed by atrazine (47%), alachlor (26%) and acetochlor (18%) (See Figure 1).

In order to estimate the amount of pesticides being deposited to the region by wet deposition, the fluxes of pesticides in rainfall were computed as the mass of pesticide falling per unit area from April- mid September. Figure 2 gives an overview of the herbicide flux in different months over the four years. Trifluralin has not been analyzed thus far for samples from 2003. It is seen that the most herbicide fluxes are confined to the months of April to July. Metolachlor and atrazine dominate the fluxes. 2001 and 2002 had much smaller herbicide fluxes than 2000 and 2003. Assuming that application patterns and amounts applied are more or less constant over the years; wet deposition appears to be strong function of rainfall amount. That is to say, the more it rains, the higher the wet deposition of the pesticide will be. The total herbicide fluxes also follow the sequence of number of rain events from April-September in a given year: 2003 > 2000 > 2002 > 2001.

Comparison of metolachlor wet deposition flux to the estimated usage in the Maryland counties on the eastern shore reveals a high correlation (See Figure 3). This suggests that the source of the herbicide in wet deposition is mainly local.

Pesticides in air over the Peninsula
From the 2000 air data from Horn Point, it is seen that most organochlorine pesticides (endosulfans, chlordanes etc.) are seen in the air throughout the year. Some compounds are more persistent than others. One way to ascertain the persistence of a compound in the atmosphere is to determine the half-life of the compound of interest. Half-lives can vary widely even for sites situated relatively close and can reflect the difference in the removal mechanisms. The atmospheric half-lives of the g-HCH have been found to range from 2.2 yr at Lake Michigan to 7.3 yr at Lake Ontario. The half-life of g-chlordane was 3.2 yr at Lake Michigan and that of trans-nonachlor was 9.2 yr at Lake Ontario (Cortes et al., 1998). Owing to the higher temperatures in the Delmarva Peninsula (as compared to the Great Lakes region), the half-lives of pesticides no longer in use (e.g. chlordanes) should be smaller than that observed in the Great Lakes region.

For most pesticides currently in use (atrazine, endosulfans etc.) the concentrations peak during the application periods and are considerably lower during the rest of the year. The concentration pattern suggests that temperature is not the only factor governing the atmospheric concentrations. There is considerable non-linearity and scatter in the Clausius Clapeyron plot for a-endosulfan (see Figure 4) and temperature only explains about 40 % variation in the concentration. As has been suggested (Cortes et al., 1999) the atmospheric concentration for pesticides currently in use is highly dependent on the agricultural cycle. We are in the process of determininng the half-lives of the pesticides frequently detected in the air the Peninsula, taking into account the agricultural usage in the region using the model developed by Cortes et al. (1999) in order to determine the atmospheric half-lives is the first step in quantifying the effect that agricultural usage has on atmospheric concentrations.

Tasks and activities for next reporting period

Tasks for the next reporting period
The focus for the next reporting period shall be to analyze the remaining air and rainfall samples for insecticides. This includes the rainfall and air filter samples from 2003 and the air PUFs from 2001-2003. PAH analysis shall also be carried out on select air samples.

The other objective is to further analyze the air concentration data so as to obtain useful information about the fate of pesticides in the atmosphere of the Delmarva Peninsula. The air concentration data will be used to determine the vapor-particle partitioning of select pesticides under field conditions and compare them to the predictions of the Junge-Pankow model (Junge, 1977; Pankow, 1987). This will involve the use of both filter PUF concentration data (Bidleman and Harner, 2000). The gaseous air concentration data from Horn Point shall be used to determine the influence of local usage on aerial concentrations of currently used pesticides and the half-lives of pesticides no longer in use. To achieve this objective the air concentration data of pesticides will be analyzed as per the approach developed by Cortes et al. (1998, 1999) to link the observed aerial concentration to temperature, time and agricultural pesticide application cycle.

The vapor-particle partitioning experiments shall be conducted to elucidate the influence of temperature and relative humidity on vapor-particle partitioning.

Work plan to accomplish tasks
The PhD student is working full time on this project. Sample and data analysis will be conducted as scheduled.

Cortes et al. (1998, 1999) have developed a method to analyze the gas-phase partial pressures of semivolatile organic compounds (SVOCs) which involves modifying the basic Clausius Clapeyron equation to account for temperature, decay and agricultural cycle. The agricultural cycle is approximated by a Laurentzian function. The data for the Horn Point site for select pesticides shall be analyzed by the techniques developed by these authors. For the determination of field partitioning coefficients, the data form Horn Point for 2002 and 2003 shall be used and analyzed as per the technique suggested by Bidleman and Harner (2000).

The experimental set up for the laboratory experiments on vapor-particle partitioning shall be completed and the experiments will be conducted for select pesticides as per the approach developed by Goss (1992, 1993).

Concerns or difficulties
At this point, we have no concerns or difficulties.

Expenditures
We are currently in a no-cost extension and expenditures are as budgeted. Furthermore, data analysis and manuscript preparation will continue for one additional year upon the completion of this grant as the PhD student has been awarded a one-year University of Maryland fellowship to concentrate in data analysis and thesis write-up.

References
Bidleman, T.F. and Harner, T. In Handbook of property estimation methods for chemicals. 2000, CRC Press, pp 233-260.

Cortes, D.R., Basu, I., Sweet, C.W., Brice, K.A., Hoff, R.M. and Hites, R.A. Environ. Sci. Technol., 1998, 32(13), 1920-1927.

Cortes, D.R.; Hoff, R.M.; Brice, K.A. and Hites, R. A. Environ. Sci. Technol., 1999, 33(13), 2145-2150.

Goss, K.-U. Environ. Sci. Technol. 1992, 26, 2287-2294.

Goss, K.-U. Environ. Sci. Technol. 1993, 27(10), 2127-2132.

Junge, C.E. In Fate of pollutants in the air and water environments: Part 1. Mechanisms of interactions between environments and mathematical modeling and the physical fate of pollutants”, ed. Suffet, I.H., 1977, Wiley, New York, pp 7-25.

Pankow, J.F. Atmos. Environ. 1987, 21(11), 2275-2283.