Prevalence and distribution of fecal indicator ...

Marine Pollution Bulletin 54 (2007) 1472–1482 www.elsevier.com/locate/marpolbul

Prevalence and distribution of fecal indicator organisms in South Florida beach sand and preliminary assessment of health effects associated with beach sand exposure Tonya D. Bonilla a,*, Kara Nowosielski a, Marie Cuvelier a, Aaron Hartz a, Melissa Green b, Nwadiuto Esiobu b, Donald S. McCorquodale a, Jay M. Fleisher c, Andrew Rogerson a,1 a

Oceanographic Center of Nova Southeastern University, Dania Beach, FL 33004, USA Department of Biological Sciences, Florida Atlantic University, Davie, FL 33314, USA College of Osteopathic Medicine, Nova Southeastern University, Fort Lauderdale, FL 33328, USA b

c

Abstract Fecal indicator levels in nearshore waters of South Florida are routinely monitored to assess microbial contamination at recreational beaches. However, samples of sand from the surf zone and upper beach are not monitored which is surprising since sand may accumulate and harbor fecal-derived organisms. This study examined the prevalence of fecal indicator organisms in tidally-affected beach sand and in upper beach sand and compared these counts to levels in the water. Since indicator organisms were statistically elevated in sand relative to water, the study also considered the potential health risks associated with beach use and exposure to sand. Fecal coliforms, Escherichia coli, enterococci, somatic coliphages, and F+-specific coliphages were enumerated from sand and water at three South Florida beaches (Ft. Lauderdale Beach, Hollywood Beach, and Hobie Beach) over a 2-year period. Bacteria were consistently more concentrated in 100 g samples of beach sand (2–23 fold in wet sand and 30–460 fold in dry sand) compared to 100 ml samples of water. Somatic coliphages were commonly recovered from both sand and water while F+-specific coliphages were less commonly detected. Seeding experiments revealed that a single specimen of gull feces significantly influenced enterococci levels in some 3.1 m2 of beach sand. Examination of beach sand on a micro-spatial scale demonstrated that the variation in enterococci density over short distances was considerable. Results of multiple linear regression analysis showed that the physical and chemical parameters monitored in this study could only minimally account for the variation observed in indicator densities. A pilot epidemiological study was conducted to examine whether the length of exposure to beach water and sand could be correlated with health risk. Logistic regression analysis results provided preliminary evidence that time spent in the wet sand and time spent in the water were associated with a dose-dependent increase in gastrointestinal illness. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Beach sand; Marine beach; Water quality; Interstitial water; Fecal indicator microorganisms; Enterococci; Fecal coliforms; Escherichia coli; Coliphage; Public health; Swash zone; Recreational beach; Exposure; Health risk

1. Introduction

* Corresponding author. Present address: Department of Infectious Diseases and Pathology, University of Florida, Gainesville, FL 32611, USA. Tel.: +1 352 392 4700x5839; fax: +1 352 392 9704. E-mail address: [email protected]fl.edu (T.D. Bonilla). 1 Present address: Department of Biology, Marshall University, Huntington, WV 25755, USA.

0025-326X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2007.04.016

The numbers of fecal indicator organisms in shoreline water are enumerated to assess the level of fecal contamination at recreational beaches, and hence the associated health risks of exposure to fecal-derived microbial pathogens (USEPA, 1986; Pruss, 1998). Fecal indicator concentrations in beach sand are not routinely measured despite the possibility that beach sand may act as an important

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

reservoir for microbial contaminants (USEPA, 1999). Importantly, sand is a natural filter that traps environmental particulates, organic matter and microorganisms (Ahammed and Chaudhuri, 1996; Hua et al., 2003; Hijnen et al., 2004) and beach sand naturally filters water washing ashore by waves or from the land after rain events. Beach sand also accumulates microorganisms shed from humans and animals (Papadakis et al., 1997; Elmir et al., 2007). The large surface area of sand grains and the unique microhabitats within the cracks and crevices provide microbes with a variety of potentially suitable environments for growth and/or enhanced survival (USEPA, 1999). Furthermore, fecal indicator bacteria can persist and potentially multiply in tropical soil and sand (Carrillo et al., 1985; Davies et al., 1995; Byappanahalli and Fujioka, 1998; Solo-Gabriele et al., 2000; Craig et al., 2004; Anderson et al., 2005). The accumulation of fecal bacteria in sand has two potential consequences for beach users. The washout of bacteria from the sand into nearshore waters might complicate the task of water quality managers’ intent on monitoring the quality of bathing water. Moreover, if fecal indicators are being concentrated in beach sand, fecal-born pathogens may also be accumulating raising the question of whether contact with sand poses additional health risks related to beach use. The environmental occurrence of fecal indicator bacteria in beach sand has been documented (Ghinsberg et al., 1994; Alm et al., 2003; Whitman et al., 2003; Shibata et al., 2004). However, the extent of which direct exposure to beach sand rich with fecal microorganisms affects the health of beachgoers has been largely unexplored. In this study, levels of enterococci, Escherichia coli, fecal coliforms, somatic coliphages and F+-specific coliphages in the water column, tidally-affected beach sand, and upper beach sand were measured at three popular bathing-beaches in South Florida (Hollywood Beach, Fort Lauderdale Beach, and Hobie Beach). The relationship between the different indicators and the influence of physical and chemical parameters on indicator levels was examined to understand how readily each fecal pollution indicator could be influenced by environmental conditions in sand. In addition, the variation of bacterial indicator densities in beach sand was examined on a micro-spatial scale. A Pilot Epidemiological Study was conducted looking into possible health effects of exposure to beach sand. Since this is the first epidemiological study of its kind, we focused on the occurrence of associated illnesses with length of exposure and not on specific exposures to individual indicator organisms. 2. Methods 2.1. Study sites and sample collection Three South Florida beaches [Hobie (25°44 0 22.500 N, 80°10 0 18.700 W), Hollywood

Beach Beach

1473

(26°02 0 02.5600 N, 80°06 0 50.3600 W), and Ft. Lauderdale Beach (26°07 0 17.3500 N, 80°06 0 14.2400 W)] were sampled bimonthly between July 2001 and July 2002. Hobie and Hollywood Beaches were additionally sampled monthly through July 2003. Hobie Beach is situated on Biscayne Bay in Dade County, and both Hollywood and Ft. Lauderdale Beaches form part of the Broward County coastline along the Atlantic Ocean. Samples were obtained from three sites along a transect line during low tide: (1) water site (5 m from shore at knee deep), (2) wet sand site (midway between the current water level and the high tide line), and (3) dry sand site (5 m above the high tide line). Each site was sampled in triplicate at three locations perpendicular to the shoreline and 2 m apart. A single sand sample was a composite of cores spanning surface sand to sand at a depth of 12 cm. Similarly, for each of the three water locations, three 500 ml samples were collected in sterile collecting bottles. All samples were immediately placed on ice for transport to the laboratory. All samples were processed within 6 h of collection. 2.2. Enumeration of bacteria Sand samples (200 g) were shaken vigorously in 500 ml of phosphate-buffered saline (USEPA, 2000) for 1 min to suspend bacteria. After shaking, large particulates in the extract were briefly allowed to settle. Water samples were also shaken vigorously by hand for 1 min to dislodge bacteria from suspended aggregates. Samples were processed by membrane filtration (APHA, 1999). Typically three different samples volumes were passed through a Millipore 0.45 lm mixed cellulose ester membrane filter by gentle vacuum filtration, and the filter with collected bacteria was placed on the agar surface of culture media (MEI for enterococci, M-FC for fecal coliforms, and MTEC for E. coli) as described in USEPA (2000). MEI cultures were incubated at 41 °C for 24 h and colonies of any color surrounded by a blue halo were scored as enterococci. After a 2 h incubation at 35 °C (to resuscitate stressed cells) M-FC and MTEC cultures were incubated at 44.5 °C for 22 h in a waterbath. Blue colonies on M-FC media were counted as fecal coliforms, and colonies from MTEC media that remained yellow after the membrane was transferred to an absorbent pad saturated with urea broth were counted as E. coli. 2.3. Enumeration of coliphage Somatic and male-specific (F+) coliphages were extracted from sand by vigorous shaking in 3% beef extract (pH 7.0). Aliquots of the sand extract or seawater were added to molten Tryptic Soy Agar containing exponentially growing E. coli and antibiotic for selection in a single-layer plaque assay (USEPA, 2001). E. coli CN-13 (ATCC 700609) and naladixic acid (10 lg/ml) were used for the propagation of somatic coliphages. E. coli Famp

1474

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

(ATCC 700891) and streptomycin/ampicillin (1.5 lg/ml) were used to propagate F+-specific coliphages. Plaques were counted after 4–6 h and after 24 h of incubation at 37 °C. An additional set of samples were enriched prior to enumeration by plaque assay as described above. Here, sand and water samples were pre-incubated for 3 h at 37 °C in the presence of half-strength tryptic soy broth (BD diagnostics), 20 mM MgCl2, 0.5% exponential phase E. coli, and antibiotic. Following the 3 h incubation, samples were placed on ice and subsequently analyzed by a single-layer plaque assay as described above. 2.4. Influence of bird feces on beach sand and translocation of bacterial sized particles by beach users Since both gull excrement and the movement of people on the beach may influence the spread of fecal organisms in dry sand two experiments were conducted to assess the scale of these factors. 2.4.1. Impact of gull excreta Samples of fresh gull feces (n = 16) were collected from Hollywood Beach between May and August 2003. Samples consisted of the feces and any contaminated surface sand (ca. 1 g sample in total). The fecal mass was transported on ice to a private beach on the grounds of the Oceanographic Center of Nova Southeastern University and buried at the private beach 2 cm below the sand surface at a site 5 m above the high tide line. These seeding experiments were conducted at a site free of beach users. Samples of sand were collected prior to seeding with the fecal mass to determine the background enterococci levels at the beach. After 24 h, samples of sand (50 g at 2 cm depth) were collected at distances of 0, 0.5. 1.0, 1.5, and 2.0 m from the inoculation site. The experiment was repeated on four separate occasions and four sand samples were collected at each distance. All experiments were conducted in the absence of rainfall. Enterococci counts were determined as previously described. 2.4.2. Translocation of fluorescently labeled beads by beach users Approximately 1 g of sand was mixed with a high concentration (ca. 108) of 0.6 lm green-fluorescent beads (Polysciences). Beads were buried 2 cm below the surface in dry sand at sites of high traffic (close to showers, trash can and sidewalk). Deployments were on dry, sunny days when the beach was well used. After 4 h, samples of sand (ca. 10 g) were collected at distances from the source (10 cm, 50 cm, 1 m, 2 m, 3 m, 5 m). Samples were shaken with distilled water and the supernatant filtered on a 0.45 lm filter and examined by epifluorescence microscopy for the presence of beads. Triplicate samples were collected and scored as positive or negative for beads. The experiment was replicated three times.

2.5. Microspatial distribution of enterococci in sand To investigate the variation of counts in the sand on a small scale, samples from both wet (tidally influenced) and dry sand were collected between August 2002 and January 2003. Sand samples (less than 1 g) were collected along a 2 m transect (on each outing, 20 samples were collected every 10 cm along the transect). In all cases, the samples were taken from 2 cm below the surface and placed in individual sterile, pre-weighted, micro-centrifuge tubes. The maximum weight of sand processed was 0.8 g. To remove bacteria, 200 ll of phosphate-buffered saline was added and the tubes were vortexed on maximum for 1 min. Suspended bacteria were collected on membrane filters and enterococci were cultured on mEI agar as previously described. All counts were normalized to CFU 100 g1. Over the duration of the study, 400 samples of dry sand and 60 samples of wet sand were analyzed. 2.6. Physical and chemical parameters On each outing the temperature of the air, sand, and water, was recorded. To determine water content of the sand, a 10 g aliquot was oven-dried at 105 °C for 24 h. The sample was weighed before and after drying and the water weight was adjusted for salt content. The salinity of seawater and sand extract (100 g in 100 ml distilled water) was measured using a probe meter. The turbidity of seawater and sand extract was measured as NTUs. Data on the local rainfall accumulation was obtained from the National Weather Service (NOAA http://www.weather. gov/climate/index.php). 2.7. Statistical analysis of indicator organism densities Fecal indicator data from routine monitoring are reported as the geometric mean of colony-forming units isolated from three grab samples of water or from threecomposite samples of sand. The significance of indicator level differences in sand and water was verified by a twosample t-test assuming unequal variance. The Spearman’s Rho statistic was used to analyze the correlation between fecal indicator organism levels using individual data points. The relationship between physical/chemical parameters and fecal indicator levels was examined through multiple linear regression tests using individual data points that were log-transformed. Statistical analyses were performed in SPSS 14.0. 2.8. Epidemiological study A Prospective Cohort Study was conducted to yield preliminary estimates of possible health effects of exposure to the three locations studied; exposure to water, wet sand and dry sand. Since each beach user that entered the water also had to have some exposure to both the dry and wet sand, we were faced with the issue of multiple exposures

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

of individual participants. Thus, we chose to investigate the number of hours beach users spent in the water, and wet sand and the dry sand and related these measures to possible risks of illness. Time spent in the water included the activities of swimming, snorkeling, diving while activities in the sand included sitting, playing, or lying in the sand. On each trial day participants were made aware of the study and if willing were given a survey form and asked to complete the form 4 days after their beach visit. The questionnaire focused on activity on the beach, duration of activity, and illness derived during the 4 days post beach interval. Fourteen symptoms were listed on the questionnaire but these were grouped into four categories for analyses: gastroenteritis (nausea, diarrhea, stomach pain or cramps), dermatological (skin rash, cuts that became infected, earache, eye irritation), upper respiratory (coughing, nasal congestion, runny nose, sore throat), and constitutional (fever, chills). Since gastroenteritis and upper respiratory illness has been shown to be associated with swimming in recreational waters contaminated with domestic sewage (Kay et al., 1994; Fleisher et al., 1996), we chose these two illnesses as Outcome Illnesses. Exposure endpoints were defined as (1) people exposed to dry sand only, (2) people exposed to wet sand only, and (3) people entering the water without significant exposure to either wet or dry sand. Our control group consisted of non-beachgoers randomly chosen from the general population. The entry criterion for the control group was that they should not have visited a beach in at least 9 days. The number of respondents who participated in the study was 882 for the experimental group and 609 for the control group. 2.8.1. Statistical analysis of epidemiological study The average time spent by study location was calculated. A correlation analysis was conducted in order to determine if time of exposure was correlated between the three exposure locations. Unconditional multiple logistic regression modeling was used to assess exposure locations to each outcome illness. All statistical analyses were conducted using the SAS system of statistical analysis. 3. Results Over the course of the 2-year study, 288 seawater samples and 576 sand samples were analyzed by membrane filtration for enterococci, fecal colifoms, and E. coli at three popular bathing-beaches in South Florida (Ft. Lauderdale Beach, Hollywood Beach, and Hobie Beach). Levels of enterococci, fecal coliform, and E. coli levels were consistently more concentrated in beach sand compared to the seawater at all three study beaches (Fig. 1a). Levels in tidally-affected wet sand (100 g) were 2–23 times greater than in 100 ml water (p < 0.01), and levels in upper beach dry sand (100 g) were 30–460 times greater than levels in 100 ml water (p < 0.001). The concentration of bacteria in sand is even more evident when a comparison is made between the numbers of organisms from the water column

1475

(100 ml) to the number of organisms in 100 ml of interstitial water in beach sand (Fig. 1b). With this comparison, bacteria were concentrated some 100-fold in wet sand and over 1000-fold in dry sand relative to water. The highest bacteria levels were seen at Hobie Beach, especially in the case of bacteria in the upper beach sand. There were no clear temporal trends in the data over the 34 sampling events although bacterial counts in the sand at Hobie Beach were generally lower in samples collected over events 12–16 (i.e. winter months when temperatures and rainfall were lower). Fig. 2 shows the temporal sampling data for enterococci, fecal coliforms, and E. coli from Hobie Beach. Enterococci were most frequently isolated while E. coli was the least encountered. This finding was similar at Ft. Lauderdale and Hollywood Beaches. However, in contrast to findings from Hobie Beach, at Ft. Lauderdale and Hollywood Beaches some of the highest bacterial counts were observed during winter months. It is of interest to note that beach usage at both Ft. Lauderdale and Hollywood Beaches was also highest during the winter months. In a recent study by Elmir et al. (2007) it was demonstrated that bather shedding of microorganisms can affect beach quality. At the microspatial level, variation in the numbers of enterococci in sand over short distances was considerable. In the course of counting bacteria in 0.8 g or less of dry sand (n = 400) the overall mean count was equivalent to 2429 CFU 100 g1 (S.E. 892), a value close to the levels found when processing 100 g aliquots of sand. On a given day, the difference between the highest and lowest counts of enterococci in samples of dry sand that were collected every 10 cm along a 2 m transect averaged greater than 15,000 CFU g1. The greatest variation along a single 2 m transect of dry sand was in October 2002 when counts ranged from below detection to 49,344 CFU 100 g1. Variation of bacterial counts in wet sand was equally large. The micro-sampling method for wet sand yielded an overall mean of 1900 CFU 100 g1 (S.E. 1088; n = 60). As was found for the dry sand, on a given day counts of enterococci in wet sand varied markedly on a micro-scale (on average > 10,000 CFU g1). On one occasion in September 2002 samples along a single 2 m transect of wet sand ranged from below detection to 17,672 CFU 100 g1. The ability of fecal enterococci to radiate out (and in part account for the microspatial variation noted above) from a fecal pellet inoculated into dry sand was determined by counting the numbers of enterococci at distances from a fecal inoculation (i.e. 1 g of fresh gull feces with associated sand). After 24 h, the average count of enterococci (less the background level) at the site of inoculation was 20,708 CFU 100 g1 (±S.E. 5029). After 24 h at 0.5 m from inoculation, counts were 16,083 ± 2254 CFU 100 g1. At 1 m and 1.5 m the counts fell to 4734 ± 1645 CFU 100 g1 and 1750 ± 914 CFU 100 g1, respectively. At 2.0 m from the inoculation site, counts after 24 h were still above background (747.5 ± 219 CFU 100 g1) indicating that a single gull dropping can influence some 3.1 m2 of dry beach sand

1476

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

a Log CFU / 100 ml or CFU / 100 g

107

Ft. Lauderdale Beach Hollywood Beach Hobie Beac h

106 105 104 103 102 101 100

b

Water Wet Dry Fecal coliforms

Water Wet Dry E. coli

Water Wet Dry Fecal coliforms

Water Wet Dry E. coli

Water Wet Dry Enterococci

10 7

Log CFU / 100 ml

10 6 10 5 10 4 10 3 10 2 10 1 10 0

Water Wet Dry Enterococci

Fig. 1. Average numbers of fecal coliforms, E. coli and enterococci from 30 sampling occasions at the three study beaches over the entire 2-year sampling period. In (a) counts are as reported as Log CFU per 100 ml of water or log CFU per 100 g of sand and (b) is the same data with the sand counts normalized to the interstitial water content of the sand, where both water and sand counts are reported as Log CFU per 100 ml.

in the absence of any noticeable disturbance by traffic or weather. The bead translocation experiment investigated the movement of bacterial sized particles from a point source (analogous to a bird dropping) in dry sand (Table 1). Fluorescent beads were consistently moved 1 m within 4 h in high traffic areas. In one case, beads were detected in the sand 5 m from the source. Overall, beads were moved an average of 1.6 m in just 4 h in high traffic areas. A single-layer plaque assay was used for the detection of somatic and F+-specific coliphage. The detection of coliphages was greatly enhanced by including a preliminary enrichment step prior to setting up the plaque assay (Table 2). Somatic coliphages were commonly detected in both seawater and sand, whereas F+-specific coliphages were seldom recovered. Between the three beaches, F+-specific coliphages were most often isolated from Hobie Beach, in particular from the wet sand. Levels of both somatic and F+-specific coliphages were more concentrated in sand (100 g) compared to water (100 ml) and coliphage detection

using the direct method occurred primarily in the cooler winter months (data not shown). The fecal indicator counts were compared by correlation analyses (Table 3). Consistent among the three beaches, fecal coliforms were the best correlated indicator across beach zones (i.e. between water and wet sand, wet sand and dry sand, water and dry sand). At Ft. Lauderdale Beach, E. coli was also well correlated across beach zones. In general, enterococci populations were not consistently well correlated across areas of beach and this was finding was similar for somatic coliphages. The levels of different indicator bacteria (enterococci, fecal coliforms and E. coli) fluctuated similarly within a single water or sand sample. Regression analysis was used to compare the biological data with the physical/chemical measurements at the time of sampling. At Hobie Beach, the most significant factor to impact indicator levels considered in this study was rainfall. Rainfall was positively correlated with bacterial indicator levels and inversely correlated with coliphage levels

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

Water Wet Sand Dry Sand

Enterococci

106 105 104 103 102 101 100

Fecal coliforms

107

1 2 3

4 5

6

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

1 2 3

4 5

6

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

1 2 3

4 5

6

7 8 9 10 11 12 13 14 15 16 17 18 19 92

106 105 104 103 102 101 100

E. coli

Log CFU / 100ml water and CFU / 100 g sand

1477

106 105 104 103 102 101 100

J

A

S

O

N

D

J

F

M

A

21 22 23 24 25 26 27 28 29 30 31 32 33 34 M

J

J

S O N D J

F M A M

Fig. 2. Seasonal counts of enterococci, fecal coliforms, and E. coli from Hobie Beach. Data is reported as the log geometric mean of 3 samples in CFU per 100 ml (water samples) or Log CFU per 100 g (sand samples). Error bars represent the standard deviation. X-axis numbers 1–25 correspond to bimonthly sampling dates in the time period of July 31, 2001–July 23, 2002 and numbers 26–34 correspond monthly sampling dates between September 24, 2002 and June 12, 2003. Month is listed below.

Table 1 Distances moved by bacterial sized fluorescent beads (0.6 lm) inoculated into dry sand Distance from source (m)

Beach site Shower

0.1 0.5 1.0 2.0 3.0 5.0

1 +++ +++ ++ + – –

Trash can 2 +++ +++ + + – –

3 +++ +++ + + – –

1 +++ +++ + – – –

Sidewalk 2 +++ + + – – –

3 +++ + + + + +

1 +++ +++ +++ + – –

2 +++ +++ ++ – – –

3 +++ +++ + + – –

Samples were collected after 4 h. Translocation was a result of movement of people on the beach at sites of high traffic (close to a shower, trash can, and sidewalk). +++ signifies beads in all three replicate samples, ++ beads in two samples, and + indicates beads in just one sample. The experiment was replicated 3 times (trials 1, 2 and 3). (–) no beads.

both in the water column and sand. Most notably, regression analysis revealed that at Hobie Beach rainfall could predict a significant proportion of the variance observed with enterococci in dry sand (R2 = 21%, F1,23 = 7.67, p < 0.05) and rainfall and turbidity together could predict 36% of the variance observed in somatic coliphages from the water column (F2,20 = 4.15, p < 0.05). Increased rainfall levels were strongly associated with a decrease in turbidity

(p < 0.001) and salinity (p < 0.01) of water and wet sand; however no correlation was observed between these parameters in the dry sand. At Hollywood Beach both rainfall and temperature were important predictors where rainfall was positively correlated and temperature was inversely correlated with indicator levels. While not robust, the effects of rainfall and temperature were most evident in the sand at Hollywood

1478

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

Table 2 Frequency of somatic coliphage and male-specific (F+) coliphage detection using a single-layer plaque assay Hobie Beach

Somatic enriched Somatic direct Male-specific (F+) enriched Male-specific (F+) direct

Hollywood Beach

Ft. Lauderdale Beach

Water

Wet

Dry

Water

Wet

Dry

Water

Wet

Dry

18/27 9/23 5/24 0/23

15/27 9/23 7/24 0/23

11/27 3/23 3/24 3/23

19/28 1/24 5/25 0/23

13/28 11/23 0/25 1/24

13/28 10/23 4/25 1/24

11/19 2/15 1/16 0/15

12/19 2/15 0/16 0/15

6/19 4/15 0/16 0/15

Samples from nearshore water, intertidal wet sand and upper beach dry sand at three South Florida beaches were analyzed directly and following an enrichment step.

Table 3 Correlation coefficients (q) of fecal indicator organisms in nearshore water (wt), intertidal wet sand (ws), and upper beach dry sand (ds) FCa

Hobie Beach FC wt FC ws FC ds EC wt EC ws ENT wt

ds

wt

ws

0.61** –

0.56** 0.53** –

0.67** 0.43*

0.40*

Ft. Lauderdale Beach FC wt 0.52** FC ws – FC ds EC wt EC ws EC ds a

c d * **

ENTc

ws

Hollywood Beach FC wt 0.59** FC ws – FC ds EC wt EC ds Ent ws Col ds

b

ECb ds

COLd

wt

ws

ds

ws

0.65** 0.48**

0.43* 0.56** 0.36*

0.45**

0.39*

– –

–

0.39* 0.50**

0.44*

0.35* 0.49** –

**

0.43 –

0.43** 0.35*

0.37* **

0.47* *

0.54

0.35 0.56**

– –

0.43** 0.41* 0.43*

0.43* –

0.81** 0.70**

0.62** 0.74**

–

0.73** –

0.54** 0.72** 0.45* 0.56** –

0.46*

0.56** 0.55**

0.43* 0.52** 0.69**

Fecal coliforms. E. coli. Enterococci. Somatic coliphages. Correlation is significant at the 0.05 level (2-tailed). Correlation is significant at the 0.01 level (2-tailed).

Beach where regressions indicated that rainfall and temperature could predict a significant proportion of the variance observed with fecal coliforms (R2 = 36%, F2,23 = 6.15, p < 0.01) and E. coli (R2 = 26%, F2,23 = 3.78, p < 0.05) in dry sand, and enterococci in wet sand (R2 = 31%, F2,23 = 5.05, p < 0.05). At Ft. Lauderdale Beach the effects of temperature remained important; however, turbidity and salinity also emerged as important predictors of indicator levels. Turbidity was positively correlated with bacteria and inversely correlated with coliphages, and salinity was inversely correlated with bacteria. Most notably, 55% of the variance observed with coliphages in dry sand could be predicated by the temperature and turbidity of dry sand (F2,20 = 8.12, p < 0.01). Thirty- four percent of E. coli var-

iance in wet sand could be attributed to temperature and turbidity (F2,20 = 4.08, p < 0.05) and 30% of enterococci variance in wet sand could be attributed to turbidity (F1,20 = 7.25, p < 0.01). In dry sand, temperature and salinity significantly influenced E. coli levels (R2 = 32%, F2,23 = 5.28, p < 0.05). The results of the epidemiological study showed crude rates for GI illness among beachgoers versus population controls as 8.5/100 and 15.3/100 respectively (p < 0.0001). The observation that beachgoers had lower rates of illness than the control group begs the question: are those who attend the beach systematically different from those who do not with regard to health status, as has been questioned in previous studies (Seyfried et al., 1985b,a; Balarajan et al., 1991). Because of this finding and the theoretical

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482 0.20

Probability of Gastroenteritis

question raised, we restricted further analysis to the beachgoers only. Excluding the control group from further analysis seemed warranted. The mean time beachgoers reported spending time in each of the three exposure locations was as follows: Water 34.6 ± 27.7 min, wet sand 36.6 ± 33.5 min, and dry sand 44.7 ± 27.7 min, p = 0.47. The correlation analysis revealed no substantial correlation between the time beachgoers spent at each exposure location (greatest R-square value = 0.12). Logistic Regression Analysis showed only time spent in the wet sand and time spent in the water were associated with increased GI illness. These estimates in terms of Odds Ratios were 1.008 (95% CI 1.001–1.015) and 1.009 (95% CI 1.000–1.018) per 10 min of exposure, respectively. The probability of GI illness per 10 min intervals of exposure calculated from the results of the multiple logistic regression are plotted in Figs. 3–5.

1479

0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0

20

40

60

80

100

120

140

Time of Exposure in Minutes Minutes Exposed vs Probability GI Wet Sand Minutes Exposed vs Probability GI Water Fig. 5. Dose–response plots showing the relationship between the minutes beachgoers were exposed to water/wet sand and reports of gastrointestinal illness.

Time Spent in Water vs Probability of GI Illness

Probability of GI illness

0.16

4. Discussion

0.14 0.12 0.10 0.08 0.06 0.04 0

20

40

60

80

100

Minutes spent in water Fig. 3. Dose–response plots showing the relationship between the minutes beachgoers were exposed to water and reports of gastrointestinal illness.

Time on Wet Sand vs Probability of GI Illness 0.20

Probability of GI Illness

0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0

20

40

60

80

100

120

140

Minutes spent on wet sand Fig. 4. Dose–response plots showing the relationship between the minutes beachgoers were exposed to wet sand and reports of gastrointestinal illness.

The potential of waterborne disease transmission at recreational beaches is related to the abundance of specific indicator microbes in the water column. When indicator levels are high (greater than regulatory standards) fecal contamination may be present leading to an increased risk of encountering disease-causing pathogens. The present study examined fecal indicator bacteria in beach sand and the water column to determine whether exposure to beach sand might pose a health concern. Beach sand from three South Florida beaches (Ft. Lauderdale, Hollywood, and Hobie Beaches) was sampled over a two-year period. When 100 ml water column samples were compared with 100 g sand samples (more or less equivalent amounts), the levels of fecal indicator microorganisms in wet sand were some 10-fold higher and levels in dry sand were on average 100-fold higher than levels detected in the water column. The sand data was normalized for water content to allow bacterial counts from 100 ml water samples to be directly compared with 100 ml of interstitial water. With this comparison, the levels of fecal indicator organisms were on average 100 times greater in wet sand and 1000 times greater in dry sand relative to water. The high concentration of fecal indicator bacteria in the sand interstices raises the question of whether enteric pathogens are similarly concentrated in the sand, and ultimately whether exposure to beach sand poses a health hazard. While the variability of bacterial indicator densities among individual samples from the water column was relatively low, the variability among discrete samples of sand was consistently high. Analysis of beach sand on a microspatial scale revealed that culturable enterococci were distributed in a very irregular or ‘patchy’ fashion, particularly in dry sand. However, in 100 g samples of sand, the pooled indicator bacteria density was consistently high. The higher

1480

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

bacterial densities observed in dry sand may in part be attributable to less predation. The dry sand contained approximately half the water content of the intertidal wet sand leading to a reduced water film surrounding sand grains. Macroinvertebrate consumers and larger protozoa in particular may not be active in this environment. The impact of bird populations on the microbial quality of bathing waters has already been documented (Oshiro and Fujioka, 1995; Fogarty et al., 2003; Wither et al., 2005). The extent to which bacteria from fecal droppings can radiate out from the source was illustrated in the present study. One fecal pellet could significantly influence the indicator count on 3 m2 patch of undisturbed beach. The movement of people on the beach may also contribute to the abundance of indicator bacteria and their distribution in the sand. In high traffic areas, bacterial sized particles were translocated on average 1.6 m in just 4 h. The mechanisms for translocation of bacteria in beach sand were not explored but most likely include the dissemination of bacteria throughout the sand from loci (such as bird feces), and transport via people. In fact, in situ bacterial replication may be enhanced as a result of sand bioturbation (Kinzelman et al., 2003). Furthermore, direct shedding of bacteria from beach users may impact indicator levels (Papadakis et al., 1997; Elmir et al., 2007). Wet sand contained fewer indicator bacteria than the dry sand but still significantly more than the water column. A recent study at Hobie Beach found that enterococci levels were correlated with tidal cycles and the highest concentrations of indicator bacteria in shoreline water were detected during high tide (Shibata et al., 2004). Loss due to tidal action explains the lower populations in wet sand relative to dry sand. Additionally, predation may be an important sink for indicator microbes (Bomo et al., 2004; Boehm et al., 2005) as more intense predation can be expected in the wet sand where consumers are not limited by the thin water film typical of the upper beach sand. Bacteriophages of E.coli have been recommended as alternate indicators of fecal pollution (Borrego et al., 1987; Calci et al., 1998) and in this study two different bacteriophages were isolated and monitored: somatic coliphages and F+-specific coliphages. Data from this study reveals that somatic coliphages are commonly isolated by culture in marine beach sand where as F+-specific coliphages are not. The detection of coliphages is enhanced upon pre-incubation of samples in growth medium containing host E. coli prior to assaying for plaque formation, implying that these viruses remain viable at levels that are below detection by direct culture. Enrichment was particularly effective for viral recovery in samples that contained higher salt content. Our finding that coliphages are inversely correlated with salinity supports previous studies (Paul et al., 1993). Similar among the three beaches studied here was the finding that bacterial fecal indicator levels are elevated in beach sand relative to water. This was true for two different beaches directly facing the Atlantic Ocean (Ft. Lauderdale

and Hollywood Beaches) and a beach bordering a sheltered bay (Hobie Beach). Others have reported similar findings at both freshwater and marine beaches (Alm et al., 2003; Whitman and Nevers, 2003; Shibata et al., 2004). Fecal coliforms, E. coli, and enterococci counts were generally well correlated in the same sample of water or sand. Compared to fecal coliforms, enterococci were less correlated between areas of the beach (i.e. between the water, wet sand and dry sand) suggesting that the population characteristics of enterococci vary according to beach location. This notion is supported by a recent study comparing the species composition of enterococci in different shore sites (Bonilla et al., 2006). Using multiple linear regression analysis, at each beach a different set of physical/chemical parameters emerged as the most robust predictors of indicator levels. At Hobie Beach rainfall and turbidity were important; at Hollywood Beach rainfall and temperature were important while at Ft. Lauderdale Beach temperature, salinity and turbidity were important. However, no set of predictors could account for the variation in indicator density in both the water and sand at a particular beach. Micro-spatial sampling of enterococci revealed that these bacteria have a highly patchy distribution in sand. The lack of ubiquitous affects of physical/chemical conditions on different areas of the beach and patchy distribution of bacteria indicate that conditions that promote survival and replication exist on a small scale. Logistic regression analysis showed only time spent in the wet sand and time spent in the water were associated with increased GI illness. These estimates in terms of Odds Ratios were 1.008 (95% CI 1.001–1.015) and 1.009 (95% CI 1.00–1.018) per 10 min of exposure respectively. The finding of higher rates of GI illness in the control group is somewhat troubling. This may be due to the possibility that people who choose to go to the beach might be systematically different than those who do not. Thus those on the beach may less likely to be ill than the general population controls we used although selection Bias cannot be ruled out. This new finding is very preliminary and must be verified by future studies. The findings that time spent in the wet sand and time spent in the water showed clear and independent dose responses with GI illness is indeed interesting. The failure of time spent in dry sand to impact health lends support to the theory that both indicator organisms and associated pathogens re-suspend from the wet sand to cause a health effect. This finding, however, needs to be validated in future, more comprehensive, epidemiologic studies. 5. Conclusions In conclusion, fecal indicator organisms were more concentrated in beach sand compared to nearshore water at three South Florida beaches. The highest indicator densities were found in upper beach sand (5 m above the intertidal zone). Micro-spatial analysis revealed that the

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

distribution of indicator bacteria in sand was irregular or ‘patchy’, supporting the notion that indicator bacteria can replicate in environmental soils. Moreover, the general physical and chemical status of sand and water could not routinely account for the observed variation in indicator densities suggesting that conditions that promote survival and replication exist on a small scale. This study demonstrated that a single fecal pellet could largely influence the indicator counts in surrounding sand and that bacterial sized particles were routinely translocated through beach use. Importantly, findings from a pilot epidemiological study correlated exposure to water and intertidal sand with gastrointestinal illness in beach users, while exposure to upper beach sand did not produce a health effect. These preliminary findings suggest that water is an important factor for exposure to pathogens at the beach, and although indicator levels are high in dry upper beach sand, the lack of regular lofting of microorganisms by water decreases the risks of exposure. Acknowledgements The authors thank the United States Environmental Protection Agency (Grant R828830). The research results have not been subjected to the EPA’s peer review and therefore do not necessarily reflect the views of the Agency. No official endorsement should be inferred. We thank Andrea Echeverry for field and technical assistance. References Ahammed, M.M., Chaudhuri, M., 1996. Sand-based filtration/adsorption media. Journal of Water Supply Research and Technology-Aqua 45, 67–71. Alm, E.W., Burke, J., Spain, A., 2003. Fecal indicator bacteria are abundant in wet sand at freshwater beaches. Water Research 37, 3978– 3982. Anderson, M.L., Whitlock, J.E., Harwood, V.J., 2005. Persistence and differential survival of fecal indicator bacteria in subtropical waters and sediments. Applied and Environmental Microbiology 71, 3041–3048. (APHA), A.P.H.A., 1999. Standard Methods for the Examination of Water and Wastewater, 20th edition. Balarajan, R., Soni Raleigh, V., Yuen, P., Wheeler, D., Machin, D., Cartwright, R., 1991. Health risks associated with bathing in sea water. BMJ 303, 1444–1445. Boehm, A.B., Keymer, D.P., Shellenbarger, G.G., 2005. An analytical model of enterococci inactivation, grazing, and transport in the surf zone of a marine beach. Water Research 39, 3565–3578. Bomo, A.M., Stevik, T.K., Hovi, I., Hanssen, J.F., 2004. Bacterial removal and protozoan grazing in biological sand filters. Journal of Environmental Quality 33, 1041–1047. Bonilla, T.D., Nowosielski, K., Esiobu, N., McCorquodale, D.S., Rogerson, A., 2006. Species assemblages of Enterococcus indicate potential sources of fecal bacteria at a South Florida recreational beach. Marine Pollution Bulletin 52, 807–810. Borrego, J.J., Morinigo, M.A., Devicente, A., Cornax, R., Romero, P., 1987. Coliphages as an indicator of fecal pollution in water – its relationship with indicator and pathogenic microorganisms. Water Research 21, 1473–1480. Byappanahalli, M.N., Fujioka, R.S., 1998. Evidence that tropical soil environment can support the growth of Escherichia coli. Water Science and Technology 38, 171–174.

1481

Calci, K.R., Burkhardt, W., Watkins, W.D., Rippey, S.R., 1998. Occurrence of male-specific bacteriophage in feral and domestic animal wastes, human feces, and human-associated wastewaters. Applied and Environmental Microbiology 64, 5027–5029. Carrillo, M., Estrada, E., Hazen, T.C., 1985. Survival and enumeration of the fecal indicators Bifidobacterium-adolescentis and Escherichia-coli in a tropical rain-forest watershed. Applied and Environmental Microbiology 50, 468–476. Craig, D.L., Fallowfield, H.J., Cromar, N.J., 2004. Use of macrocosms to determine persistence of Escherichia coil in recreational coastal water and sediment and validation with in situ measurements. Journal of Applied Microbiology 96, 922–930. Davies, C.M., Long, J.A., Donald, M., Ashbolt, N.J., 1995. Survival of fecal microorganisms in marine and freshwater sediments. Applied Environmental Microbiology 61, 1888–1896. Elmir, S.M., Wright, M.E., Abdelzaher, A., Solo-Gabriele, H.M., Fleming, L.E., Miller, G., Rybolowik, M., Peter Shih, M.T., Pillai, S.P., Cooper, J.A., Quaye, E.A., 2007. Quantitative evaluation of bacteria released by bathers in a marine water. Water Research 41, 3–10. Fleisher, J.M., Kay, D., Salmon, R.L., Jones, F., Wyer, M.D., Godfree, A.F., 1996. Marine waters contaminated with domestic sewage: nonenteric illnesses associated with bather exposure in the United Kingdom. American Journal of Public Health 86, 1228–1234. Fogarty, L.R., Haack, S.K., Wolcott, M.J., Whitman, R.L., 2003. Abundance and characteristics of the recreational water quality indicator bacteria Escherichia coli and enterococci in gull faeces. Journal of Applied Microbiology 94, 865–878. Ghinsberg, R.C., Dov, L.B., Sheinberg, Y., Nitzan, Y., Rogol, M., 1994. Monitoring of selected bacteria and fungi in sand and sea-water along the Tel-Aviv coast. Microbios 77, 29–40. Hijnen, W.A., Schijven, J.F., Bonne, P., Visser, A., Medema, G.J., 2004. Elimination of viruses, bacteria and protozoan oocysts by slow sand filtration. Water Science and Technology 50, 147–154. Hua, J.M., An, P.L., Winter, J., Gallert, C., 2003. Elimination of COD, microorganisms and pharmaceuticals from sewage by trickling through sandy soil below leaking sewers. Water Research 37, 4395– 4404. Kay, D., Fleisher, J.M., Salmon, R.L., Jones, F., Wyer, M.D., Godfree, A.F., Zelenauch-Jacquotte, Z., Shore, R., 1994. Predicting likelihood of gastroenteritis from sea bathing: results from randomised exposure. Lancet 344, 905–909. Kinzelman, J.L., Whitman, R.L., Byappanahalli, M., Jackson, E., Bagley, R.C., 2003. Evaluation of beach grooming techniques on Escherichia coli density in Foreshore Sand at North Beach, Racine, WI. Lake and Reservoir Management 19, 349–354. NOAA. National Weather Service Climatic Data. . Oshiro, R., Fujioka, R., 1995. Sand, soil, and pigeon droppings – sources of indicator bacteria in the waters of Hanauma Bay, Oahu, Hawaii. Water Science and Technology 31, 251–254. Papadakis, J.A., Mavridou, A., Richardson, S.C., Lampiri, M., Marcelou, U., 1997. Bather-related microbial and yeast populations in sand and seawater. Water Research 31, 799–804. Paul, J.H., Rose, J.B., Jiang, S.C., Kellogg, C.A., Dickson, L., 1993. Distribution of viral abundance in the reef environment of Key Largo, Florida. Applied Environmental Microbiology 59, 718–724. Pruss, A., 1998. Review of epidemiological studies on health effects from exposure to recreational water. International Journal of Epidemiology 27, 1–9. Seyfried, P.L., Tobin, R.S., Brown, N.E., Ness, P.F., 1985a. A prospective study of swimming-related illness. I. Swimming-associated health risk. American Journal of Public Health 75, 1068–1070. Seyfried, P.L., Tobin, R.S., Brown, N.E., Ness, P.F., 1985b. A prospective study of swimming-related illness. II. Morbidity and the microbiological quality of water. American Journal of Public Health 75, 1071– 1075. Shibata, T., Solo-Gabriele, H.M., Fleming, L.E., Elmir, S., 2004. Monitoring marine recreational water quality using multiple microbial

1482

T.D. Bonilla et al. / Marine Pollution Bulletin 54 (2007) 1472–1482

indicators in an urban tropical environment. Water Research 38, 3119– 3131. Solo-Gabriele, H.M., Wolfert, M.A., Desmarais, T.R., Palmer, C.J., 2000. Sources of Escherichia coli in a coastal subtropical environment. Applied Environmental Microbiology 66, 230–237. USEPA. 1986. Ambient water quality criteria for bacteria. Office of Water, Washington, DC. USEPA. 1999. Action plan for beaches and recreational waters. Office of Research and Development and Office of Water, Washington, DC. USEPA. 2000. Improved Enumeration Methods for the Recreational Water Quality Indicators: Enterococci and Escherichia coli. Office of Science and Technology, Washington, DC.

USEPA. 2001. Method 1602: Male-specific (F+) and Somatic Coliphage in Water by Single Agar (SAL) Procedure. Office of Water, Washington, DC. Whitman, R.L., Nevers, M.B., 2003. Foreshore sand as a source of Escherichia coli in nearshore water of a Lake Michigan beach. Applied Environmental Microbiology 69, 5555–5562. Whitman, R.L., Shively, D.A., Pawlik, H., Nevers, M.B., Byappanahalli, M.N., 2003. Occurrence of Escherichia coli and enterococci in Cladophora (Chlorophyta) in nearshore water and beach sand of Lake Michigan. Applied and Environmental Microbiology 69, 4714–4719. Wither, A., Rehfisch, M., Austin, G., 2005. The impact of bird populations on the microbiological quality of bathing waters. Water Science and Technology 51, 199–207.