**2. Field Methods**

Two stretches of Florida's First Coast beaches were selected for field observation (Figure 1):


Teams of two walked the stretches of beach, each 6 km in length (Figure 1). The team would arrive at the site, in the middle of the 6-km-long beach stretch, split up so that one walked north 3 km and the other walked south 3 km, and then turn around and back-track the 3-km-long walk to arrive back at the site. If the tide was going out, the team would walk the first track along the debris line and walk the second track (the back-track) along the waterline, and vice versa for an incoming tide.

**Figure 1.** (**a**) The two stretches of Florida's First Coast beaches, the north site being north of the inlet and the south site being south of the inlet, selected for field observation. (**b**) View of the beach for the north site. (**c**) View of the beach for the south site.

Both sites were visited roughly once a week over two months (Table 1). Each site visit required, on average, 3 h of walking the beach stretch. For the beach walks, the following flotsam was photographed with timestamp and GPS location: Trash; seashells; seaweed; and oil tar balls. In addition, the start and end of each beach walk was photographed with timestamp and GPS location. In the field, suspected oil tar balls were recognized by their deep black color, shiny and sooty appearance with varying range of sizes and consistencies (from smaller dime-sized pellets to larger racquetball-sized blobs). Additionally, the tidal condition (high/flood or low/ebb), weather (wind/storm) and wave climate (flat, knee-high, head-high or overhead-high) were noted for each beach walk; however, those data will not be used in this present study and are being reserved for future work related to understanding the temporal patterns of oil tar ball wash-up.

All samples were handled with non-powdered gloves, individually stored in separate air-tight baggies, brought to the laboratory and collectively stored in an air-conditioned environment for subsequent physical and chemical analyses. The GPS data, including the photographs of the samples *in situ*, were organized into a GIS database for subsequent analysis.


**Table 1.** Schedule of the beach walks.

#### **3. Laboratory Testing**

Physical analysis involved categorizing all of the samples according to their suspected composition. Table 2 lists the seven different categories of physical composition that cover the range of the collected samples: coal; bone; lava; wood; rock; tar ball; and other. Figure 2 shows samples typifying the physical character of coal, bone, lava, wood, rock and tar ball.

The literature is widely varying per how to characterize oil tar balls using chemical tests [8–12], but nonetheless guided us to conduct three chemical tests on the collected samples. The objectives of the chemical tests were to provide a first-order (low-level) identification of the collected samples, which was due to the minimal budget of the project, in that for future work and when the associated higher costs can be afforded, higher order (more detailed) chemical tests will be conducted. Thus, three chemical tests (each detailed below) were attempted to characterize the chemical makeups of the collected samples and identify them as containing oil (tar ball) or as not containing oil (anything other than a tar ball; examples, in addition to coal, bone, lava, wood and rock, would include shells, jellyfish, man-o-war, fish kill, and netting/string entanglements). Additional samples

obviously containing oil were obtained from Navarre Beach (Florida Panhandle) and Ponte Vedra Beach (Florida's First Coast), where these samples were collected on an *ad hoc* basis for the purpose of serving as control measures in the chemical tests.


**Table 2.** Categorization of the collected samples based on physical composition.

The first test attempted was gas chromatography-mass spectrometry (GC-MS) and gas chromatography/flame ionization detector (GC/FID). The GC-MS and GC/FID tests failed because, with the low-budget instrument used, temperatures were not capable of gasifying the sample compositions and producing a chemical signature. Only solvent peaks for samples obviously containing oil were obtained, which was most likely due to insufficient temperatures of the GC columns used. Columns suggested by manufacturers to characterize crude oil sludge were beyond the available budget of the study. It is hypothesized that the majority of oil obtained from tar balls are the high molecular weight portions of crude oil, commonly referred to as tar. Compounds of high molecular weight have higher melting points than lighter compounds. In order for GC techniques to be applied, the instrument must reach sufficient temperatures for the compound of interest to reach a gaseous state, pass through the column and produce a reading.

The second test attempted was solubilization of samples in DCM (dichloromethane) which is a commonly used solvent for isolating oils from particulate solid masses. A small piece from a sample in each of the categories of physical composition (Table 2) was placed into a 20-mL scintillation vial containing roughly 10 mL of DCM. This test was applied by observing if any oil residue became isolated from the sample and floated to the top of the vial. No oil residue was observed to come off any sample except for all the samples that belonged to the tar ball category where there were obvious collections of oil residue floating in the vial. Results of this test allow us to conclusively state that the members of the five categories not suspected to be tar balls indeed do not contain crude oil. However, extraction of a substance using DCM does not lead to a certain conclusion that the samples in the tar ball category do indeed contain crude oil, since DCM is a commonly used organic solvent and will dissolve many nonpolar compounds. As a result, this test can only be used to eliminate possible tar ball samples, not confirm them.

The third test attempted was evaluation of melting point. In hopes to confirm the presence or absence of crude oil in the samples, we had desired to measure a melting point of bits of "oil" from the samples in the tar ball category. For all the samples in the tar ball category, a portion of each sample was powderized and the "oil" particles were separated from the larger pieces of rock and debris via mechanical processing. Separation of the "oil" from the surrounding sand particles was impractical (although it was attempted, again via mechanical processing). It was eventually determined that the "oil" particles would melt at a much lower temperature than the sand, therefore the presence of the sand should not make a difference. The "oil"/sand mixture was heated in a test tube using a Bunsen burner which quickly revealed that the "oil" simply softened slightly and turned the sample into a gooey mess, accompanied with the smell of hot asphalt. While these results strongly suggest the sample was indeed a tar ball, the lack of quantitative evidence leaves it undetermined.

**Figure 2.** Example samples typifying the physical character of (**a**) coal, (**b**) bone, (**c**) lava, (**d**) wood, (**e**) rock and (**f**) tar ball. Disregard the "tar ball" script on the forensic ruler.

#### **4. Spatial Analysis**

The maps shown in Figure 3 are direct outputs of the developed GIS database showing the GPS-based locations of the furthest extents of the survey, all the collected samples and just the samples that were categorized as tar balls. Figure 3a shows the spatial distribution of all the collected samples, as an indication of all the flotsam that washes up on Florida's First Coast beaches, with additional points indicating the furthest extent of ground covered by the team over the full duration of the fieldwork. Figure 3b shows the spatial distributions of all the samples that were categorized as tar balls, as an indication of baseline conditions of oil tar ball wash-up on Florida's First Coast beaches, again with same extent of survey points for reference.

**Figure 3.** GPS-based locations of (**a**) all the collected samples and (**b**) just the samples that were categorized as tar balls.
