Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Microbiological and metal contamination of watercress in the Wellington region, New Zealand—2000 survey

Microbiological and metal contamination of watercress in the Wellington region, New Zealand—2000... Abstract Objectives: To investigate potential microbiological and metal contamination of watercress and to assess the public health risks associated with harvesting and consumption of watercress. Method: During March and April 2000, samples were taken from 11 known or potential watercress collection sites in the Wellington region. Microbiological testing included bacterial counts for presumptive and faecal coliforms (w atercress); total coliforms (growing water); and Escherichia C. Edmonds Hutt Valley District Health Board, New Zealand R. Hawke School of Earth Sciences, Victoria University of Wellington, New Zealand coli (E. coli) and presence/absence tests for Campylobacter species (g rowing w ater and watercress). Watercress concentrations of a range of metals were also measured. Results: All of the sites showed significant levels of E. coli in samples of both watercress and water. The E. coli levels in water were well above recommended freshwater recreational contact safety guidelines at most sites. Campylobacter was detected in the growing waters at all sites (80% of the samples) and in 11% of the watercress samples. Mean metal concentrations in watercress did not exceed the NZ Food Regulations (1984) levels at any of the sites. However, lead concentrations at the urban sites and one of the semi-urban sites would have exceeded the new Australia New Zealand Food Standards Code maximum le vels (2003). Conclusions and Implications: The consumption of raw watercress contaminated with enteric pathogens could potentially cause serious gastrointestinal illness (e.g. campylobacteriosis) and people gathering watercress could also be at risk of infection from contact with contaminated surface waters. ( Aust N Z J Public Health 2004; 28: 20-6) atercress (Nasturtium spp.) is an aquatic perennial herb that grows wild along the margins of slow-moving rivers, streams, ditches and drains. It is common in New Zealand, especially in the North Island. 1 Watercress is a popular food and is traditionally ser ved cooked; however, it is increasingl y being consumed uncooked in salads or as a gar nish and is being sold as a salad vegetable in restaurants/cafes. At present (2000) watercress is harvested both for personal use and for commercial sale in New Zealand. Since watercress grows in the aquatic environment (sometimes completely submerged), it is subject to potential chemical and microbiological contamination from the water/sediment in which it grows. Sites favourable for watercress usually have comparatively small volumetric flows, which do not allow for dilution of the pollutants they receive and are often either in urban or intensive agricultural areas. Many streams in New Zealand are subject to faecal contamination from urban and rural runoff (both diffuse and point sources), which can introduce enteric pathogens e.g. Campylobacter, into the receiving waters. 2 The consumption of watercress contaminated with enteric pathogens could be a risk factor for gastrointestinal illness and therefore could be contributing to New Zealand’ s high incidence of certain enteric diseases, e.g. campylobacteriosis.3 Persons g athering watercress could also be at risk of infection from contact with contaminated surface waters. Despite the f act that salad vegetables are generally regarded as low-risk foods they can harbour a range of pathogens. Since these foods are not necessarily cooked before consumption they can pose a health risk to the consumer.4 Watercress may pose a signif icantly g reater risk than other common salad vegetables (e.g. lettuce) as it g rows in waterways subject to waterbor ne contaminants. Although watercress has not formally been linked to any enteric disease outbreak in New Zealand, many overseas outbreaks of human gastroenteritis have been linked to the consumption of contaminated fresh vegetables,5 and watercress is included in a list of salad vegetables that have caused extensive outbreaks of salmonellosis.6 Increasingl y, enteric disease research is focusing on environmental f actors and disease and water quality datasets are being linked.7 In ter ms of metal contamination, a number of studies undertaken on central North Island rivers, e.g. the Waikato River, have revealed levels of arsenic in watercress above the World Health Organization limit for foodstuffs. This is attributed to the geothermal activity in the area and the impact of geothermal power stations.8,9 Overseas studies have also demonstrated the ability of watercress to absorb cer tain metals from contaminated water/sediments.10 Submitted: April 2003 Revision requested: August 2003 Accepted: September 2003 Correspondence to: Dr Richard Hawke, School of Ear th Sciences, Victor ia University of Wellington, PO Box 600, Wellington, New Zealand. Fax: +64 4 463 5186; e-mail: richard.hawke@vuw.ac.nz AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ The aims of this 2000 study were: • to investigate potential microbiological and metal contamination of watercress and growing waters from a number of streams in the Wellington and Wairarapa regions; and, • to assess public health risks regarding harvesting and consumption of watercress from these sites. To get a good representation of different growing sites it was decided to sample watercress and growing waters from a number of surface water sites in the greater Wellington region. Watercress and growing waters were sampled for microbiological contaminants and watercress was also sampled for metals. Samples were taken during five weeks in March and April 2000. To assess potential microbiological contamination, the presence of faecal indicator organisms (e.g. E. coli) and a common waterborne pathogen (Campylobacter) was tested. Following discussions with environmental monitoring and health agencies, a number of metals were selected for analysis. These metals could be present in growing water/sediment from industrial discharges or runoff11 and the New Zealand Food Regulations (1984) and the new Australia New Zealand Food Standards Code (2003) specify maximum permissive levels for metals.12,13 Figure 1: Water and watercress sampling locations, Wellington Region, New Zealand. Method Eleven streams in the greater Wellington region were selected as watercress sampling sites (see Figure 1). The sites covered a range of urban, semi-urban and rural conditions and a range of water quality and sediment characteristics. The sites were also selected to be representative of catchment types generally found elsewhere in New Zealand and typical watercress growing environments. Sites of known collection and consumption were targeted in addition to a control site. Of the 11 sites selected for this study, five were known to be sites where watercress was gathered in 2000 on a regular basis either for commercial sale or personal consumption. Sampling technique Watercress was gathered (weekly) from each waterway on five occasions from March 2000. The f ive samples were to allow the temporal variability in potential contamination to be assessed. In the first week, five 300 g watercress samples (healthy whole plants excluding the roots) were collected from each of the sites for individual microbiological analysis in order to be able to compare microbiological results against the Ministr y of Health’s Microbiological Reference Criteria for Food (1995).14 However, for the first week only one sample was collected from the Waiwhetu Stream because there was insufficient watercress available. For the remaining four weeks, five samples (if available) were taken from each site and combined by the laborator y into one sample (for each site) for analysis. Simultaneously, each week, a onelitre water sample was taken from each of the sites for microbiological analysis. Microbiological analysis Microbiological testing of the watercress included bacterial 2004 VOL. 28 NO . 1 counts for presumptive coliforms, faecal colifor ms, Escherichia coli (E. coli) and presence/absence tests for Campylobacter species. Microbiological testing of the growing waters included bacterial counts for total coliforms, E. coli and presence/absence tests for Campylobacter species. The microbiological anal ysis was undertaken by the Institute of Environmental Science and Research Limited (ESR), which is accredited by International Accreditation New Zealand (IANZ) for the test methods: water samples were tested for total coliforms and E. coli using Colilert MPN (IDEXX Laboratories Inc, Maine, USA); Campylobacter species presence/absence was tested for in both watercress and water samples using an adapted method;15 detection and isolation of E. coli, faecal coliforms and presumptive coliforms followed the APHA standard methods.16 Since watercress is commonly eaten as a salad vegetable the results were compared with the Microbiological Reference Criteria for Food (1995) section 5.25 (salads). This reference recommends limits for f aecal coliforms of n=5, c=2, m=102, M=103 where: • n=number of samples. • c=maximum allowable number of defective samples. • m=an acceptable le vel and values above it are marginally acceptable (i.e. 2 samples) or unacceptable (i.e. >2 samples). • M=a le vel which separates marginally acceptable quality from defective quality. Any sample above M is unacceptable. Overseas studies suggest that these criteria may be inappropriate for plant material. Bacteria such as Klebsiella and Enterobacter species, which are normal inhabitants of plant material, gi ve a positive faecal coliform test, but their presence in foods of plant origin may not necessarily be associated with faecal contamination.17-19 These authors recommend that the faecal coliform test be replaced by analysis for E. coli when testing plant-derived material. Therefore, in this study (as in the study by Little et al. 199720) the f aecal coliform limits recommended in the Reference Criteria for Foods (5.25) were used with E. coli rather than f aecal colifor ms. For the first week all sites (except Waiwhetu) had five samples AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article analysed and could be directly compared with the reference criteria. For subsequent w eeks where samples were combined, the result was considered to be an average and was considered non-compliant if greater than M. There are no criteria for Campylobacter in the Microbiological Reference Criteria for Food (5.25) relating to salads. However, the criteria for Campylobacter in Foods – Cooked Ready-To-Eat, (5.8b) “Some components not cooked in manufacturing process (e.g. sandwiches)”, is 0/10 g Campylobacter. This criterion defines that Campylobacter contamination is unacceptable in foods that will not be cooked prior to consumption. The presence/absence test for Campylobacter was therefore considered sufficient to determine the safety of raw watercress for human consumption. zinc and mercury) to give an indication of potential contamination. Five combined watercress samples in total were tested for metals from each site for five weeks. Metal analyses, b y ICP-MS, were undertak en byAgriquality New Zealand Limited using methods that have been accredited by International Accreditation New Zealand (IANZ). Results Results of microbiological analyses of the watercress samples are summarised in Table 1 (E. coli ) and Table 2 (Campylobacter). For the f irst week, none of the sites met the acceptable reference criteria for salads. Six sites were marginally acceptable and four sites were unacceptable. During the remaining four weeks, the composite samples were above M on six occasions at f ive sites. There did not appear to be any difference between catchment types. For example, the Golf Course Stream, Black Stream, Owhiro Stream and Papawai Stream were all classif ied non-complying in the f irst week and included streams flowing through urban areas, streams flowing through rural areas and streams flowing through native forest (bush) areas. The intra-site variability observed from the f ive samples collected in the first week showed considerable variation was possible. Campylobacter was present in 11% of the watercress samples. There did appear to be a difference between catchment types with Campylobacter being found in r ural catchments (e.g. Opaki, Papawai, Ohariu), semi-urban catchments (e.g. Mazengarb) and the Golf Course Stream site, but not in samples from catchments with significant urban development. Campylobacter was more common in the water samples than in the watercress samples: Metal analysis The concentrations of a range of metals in the watercress were measured (i.e. arsenic, chromium, lead, cadmium, nickel, copper, Table 1: Microbiological (E. coli) results of watercress samples by site compared with Microbiological Reference Criteria for Food – section 5.25 (salads). a Growing site Acceptable b Week 1 (five samples) Subsequent four weeks Marginally NonNonacceptablec complyingd complyinge Hulls Creek (urban) Black Stream (urban) Waiwhetu Stream N/A (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) Papawai Stream (rural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) Yes Yes N/A Yes Yes Yes Yes Yes N/A Table 2: Microbiological (Campylobacter) results of watercress samples by site compared with Microbiological Reference Criteria for Food – cooked ready-to-eat (5.8b).a Analysis Campylobacter presence (watercress 1st week) (0/5) (0/5) (0/1) (3/5) (0/5) (0/5) (0/5) (0/5) (0/5) (2/5) (0/5) Campylobacter presence (watercress total) (0/9) (0/9) (0/5) (4/9) (0/9) (1/9) (1/9) (0/9) (0/9) (3/9) (1/9) Hulls Creek (urban) Yes Yes Yes Black Stream (urban) Waiwhetu Stream (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (r ural) Papawai Stream (r ural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) Notes: (a) The new Food Standards Australia New Zealand (2001) Guidelines for the microbiological examination of ready-to-eat foods state that the presence of E. coli in ready-to-eat foods is undesirable and levels exceeding 100 per gram are unacceptable and indicate a level of contamination that may have introduced pathogens.21 (b) Acceptable: E. coli count first week (5 samples <100/g). (c) Marginally acceptable: E. coli count first week (2 samples >100/g and <1000/g). (d) Non-complying: E. coli count first week (3 or more samples >100/g or any sample >1000/g). (e) Non-complying: E. coli count subsequent 4 weeks (>1000/g, composite sample). Notes: (a) The new Food Standards Australia New Zealand (2001) Guidelines f or the microbiological examination of ready-to-eat foods state that Campylobacter should not be present in ready-to-eat foods as consumption of foods containing this pathogen may result in food-bor ne illness.21 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ Campylobacter was present in 11% of the watercress samples compared with 80% of the water samples. Differences in sample size and physical proper ties between water and watercress make a direct comparison difficult. For example, Campylobacter presence on watercress is likely to be affected by se veral factors including: the surface area of the watercress sample and the resulting water v olume retained; water temperature; and the retention and accumulation of bacteria on plant surfaces. Since the Campylobacter data is only presence/absence it gives no indication of the actual numbers in each sample, i.e. whether heavily or lightly contaminated. Some degree of Campylobacter contamination would be expected in catchments subject to faecal contamination as Campylobacter is excreted by a wide variety of animals and birds.22 However, the Microbiological Reference Criteria def ine any level of Campylobacter contamination as being unacceptable in foods that will not be cooked prior to consumption. Figure 2: Water vs. watercress E. coli counts: Owhiro Stream. 3000 E. coli (Water MPN/100ml; Watercress MPN/g) Water E. coli Watercress E. coli 2500 2000 1500 1000 500 0 4-Apr 9-Apr 14-Apr 19-Apr 24-Apr 29-Apr Sampling Date Water samples Applicab le standards or guidelines for acceptable levels of E. coli or Campylobacter in growing waters could not be found. Therefore the presence/absence test for Campylobacter in growing waters w as considered appropriate as any lev el of Campylobacter in growing waters could potentially contaminate watercress. Gathering watercress e xposes people to the risk of infection; for example, through cuts and abrasions, splashes to the eyes and mucus membranes, hand-to-mouth activity (e.g. eating and smoking) and exposure to aerosols. Hence the Ministry for the Environment’s 1998 Bacteriological Water Quality Guidelines for Marine and Fresh Water23 were used to assess potential health risks from water contact through harvesting watercress. These guidelines specify two categories: Alert/Amber Mode II if a single sample is greater than 273 E. coli/100 mL, and Action/Red Mode if a single sample is greater than 410 E. coli/100 mL. Results of microbiological analyses of the water samples are summarised in Table 3. Consistent with previous work, there was no apparent relationship between the presence of Campylobacter and the E. coli counts at the sites.24 The urban streams had greater numbers of elevated E. coli results (e.g. above 2,400 MPN/100 mL) than the predominantly semi-urban and rural streams. This may be due to the run-off characteristics of sealed surfaces; that is, in rural areas, soil and plants may retain micro-organisms before they reach surface waters. As outlined above with Camplyobacter, differences in sample size and physical properties between the water and the watercress samples make a direct comparison difficult. In general, the water E. coli count was higher than the watercress count, but this was not always the case (see Figure 2). All the sites, with the exception of the Golf Course Stream site, exceeded the Action/Red Mode on one or more occasions (see Table 3). Hence, there is a potential health risk from water contact when collecting watercress from surface waters in most catchment types. Table 3: Microbiological results of water samples by site. Analysis E. coli 100 mL Campylobacter >273 >410 >2,400 presence Alert/Amber Action/Red MPN/100mL Mode IIa Modea Hulls Creek (urban) Black Stream (urban) (5/5) (4/5) 1 Waiwhetu Stream (3/5) (urban) Mazengarb (4/5) Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) (4/5) (5/5) Papawai Stream (5/5) (rural) Parkvale Stream (5/5) (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) (1/5) (5/5) (3/5) Watercress metals results The sample results were compared with the New Zealand Food Regulations (1984) permissible proportion levels (in ppm) for Note: (a) Ministry for the Environment Bacteriological Water Quality Guidelines.23 2004 VOL. 28 NO . 1 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article selected metals. There are no regulation levels for chromium and nickel (see Table 4). In this study all mean metal concentrations were below the maximum permitted limits in the New Zealand Food Regulation (1984) levels. It is noted, however, that mean concentrations of lead observed in the urban streams and one of the semi-urban streams (see Table 4) would exceed the replacement limit in the Australia New Zealand Food Standards Code.13 Watercress samples from the urban streams and one of the semi-urban streams had higher mean metal concentrations of zinc, copper and lead than the other sites. Discussion Microbiological Campylobacter was uniformly present in waters over a range of catchment types. Thus these sites are unsuitable for gathering watercress for consumption unless thoroughly cooked (e.g. boiling). The presence of Campylobacter on 11% of the watercress samples is of public health significance. Campylobacter has been shown to survive long enough on ready-to-eat (RTE) produce to cause infection.5 The Microbiological Reference Criteria for Food (section 5.8b) for cook ed ready-to-eat foods is 0/10 g Campylobacter. There is also the potential for cross contamination of other foods and preparation surfaces/utensils from processing contaminated raw watercress. All the watercress and growing water samples at each site showed variable levels of E. coli contamination. High E. coli counts would be expected in slow-moving streams where watercress grows in sufficient quantities for collection. Little et al. (1997) describes the results of a large salad survey in England and Wales in which 2,552 samples were examined and used the same limits that were used in this study (i.e. E. coli, n=5, c=2, m=102, M=103).20 That survey found only 1% of salads had E. coli counts of 102 cfu or more per gram. In this study, 46% of watercress samples had E. coli counts of 10 2 cfu or more per gram. This indicates significant faecal contamination of watercress from contaminated growing waters. While it is not possible from the tests used to know the origin of the E. coli, sources such as livestock or birds can act as reservoirs of micro-organisms that can cause human disease. If faecal contamination (as indicated by E. coli results) is present on watercress then human pathogens may also be present, posing a risk of infection. Pathogens such as Salmonella , Camp ylobacter, Giardia and Cryptosporidium and viruses are potential contaminants of watercress especially in rural situations, where livestock can car ry and excrete patho gens. There is no formal data available on the incidence of microbial hazards from watercress grown in New Zealand. However, many overseas outbreaks of human gastroenteritis (e.g. Shigella (USA)), have been linked to the consumption of contaminated fresh vegetables. Watercress is included in a list of salad vegetables that have caused extensive outbreaks of salmonellosis.5,6 Studies have shown that thorough washing and treatment of produce with chlorinated water can reduce the populations of pathogenic micro-organisms on fresh produce but it cannot eliminate them. Beuchat (1997) suggests that the reduction of risk for human illness associated with raw produce can be better achieved through controlling points of potential contamination e.g. during growing, harvesting, processing and distribution. In regard to raw watercress, strictly monitored controls on the quality of the g rowing waters would be the most appropriate method of reducing health risks for consumers. 25 Table 4: Metal results by site (mean concentrations a). Nickel (ppm) Permissible le vels Food Regulations (1984)b Hulls Creek (urban) Black Stream (urban) Waiwhetu Stream (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) Papawai Stream (rural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) 0.12 0.13 0.07 0.06 0.05 0.05 0.06 0.05 0.07 0.05 0.24 Cadmium (ppm) 1 ppm Copper (ppm) 30 ppm Zinc (ppm) 40 ppm Chromium (ppm) Arsenic (ppm) 2 ppm Lead (ppm) 2 ppm Mercury (ppm) 0.03 ppm Notes: (a) Results are repor ted by wet weight. (b) The A ustralia New Zealand Food Standards Code was adopted in New Zealand in Februar y 2001 and took full effect on 20 December 2002.13 The code has maximum levels of metal contaminants in food for lead (0.1 mg/kg) and cadmium (0.1 mg/kg) that can be applied to watercress; these levels are lower than the le vels in the New Zealand Food Regulations (1984). While all the results for cadmium were below this level, mean lead levels at the urban sites and one of the semi-urban sites exceed the new limit. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ Testing for pathogens as an indicator of watercress or water quality is not cost effective because of the potential number of pathogens that could be present. It is more appropriate to use a suitable faecal indicator organism e.g. E. coli, as an indicator of the growing water quality and apply an appropriate standard (e.g. New Zealand Drinking Water Quality Guidelines (2000) standard of less than 1 E. coli /100 mL), to ensure minimal potential for watercress contamination. 26 Fasciolosis Liver flukes (Fasciola hepatica) are parasitic in the bile ducts of mammals. Cercariae encyst on aquatic vegetation (e.g. watercress) and the cysts (metacercariae) are then swallowed by the final host (humans, sheep, goats, cattle) while feeding. Wild watercress is reported as the main source of infection in Europe, where there is a high rate of endemic f asciolosis in domestic animals. 27 Over the past 30 years or so, liver fluke infection of livestock has become considerably more widespread in New Zealand, largely in association with the spread of the exotic snail host, Lymnaea columella.28 Although there have been no documented cases of fasciolosis in New Zealand related to consumption of watercress gathered from New Zealand waterways, there is potential for cases of fasciolosis through consumption of raw watercress. For example, a case of fasciolosis was recently reported in Australia and infection was attributed to the consumption of watercress.29 Any future risk of fasciolosis would be minimised by ensuring that raw watercress for human consumption is grown only under strictly controlled conditions. 27 However, the mean concentrations of lead observed in the urban streams and one of the semi-urban streams (see Table 4) exceeded the maximum per mitted levels in its replacement, the Australia New Zealand Food Standards Code.13 Overseas and NZ studies have shown that watercress can bio-accumulate metals to levels that could pose health risks. For example, studies undertaken in areas of the central North Island have revealed high levels of arsenic (in excess of the WHO limit for arsenic in foodstuffs) in the watercress, probably because of the geothermal activity in the area.9,30 A specif ic site risk assessment would need to be undertaken to determine potential metal contamination before the safety of consuming watercress gathered from a site could be assessed. Recommendations 1. Watercress harvested from any uncontrolled surface w ater source in New Zealand should not be consumed unless the watercress is thoroughly cooked (e.g. in boiling water) to destroy potential human pathogens. 2. Watercress should not be eaten raw unless it can be demonstrated that the growing environment is strictl y controlled and effectively monitored to ensure the water source is of suitable standard e.g. controlled hydroponic cultivation. A suitably strict standard should be applied to the growing waters e.g. New Zealand Drinking W ater Quality Guidelines (2000) E. coli standard of less than 1 E. coli/100 mL, to monitor acceptable watercress microbiological quality.26 3. Further research should be under taken to assess the risk of contracting waterborne illnesses through contact with contaminated water when gathering watercress from uncontrolled surface waters. 4. The potential for fasciolosis in New Zealand from consumption of wild watercress should be the subject of a targeted study. This would involve further research on such f actors as the geographic distribution of the host snails and the geographic distribution of Fasciola hepatica in livestock. 5. There are areas in New Zealand, specifically the central Nor th Island, where studies have revealed high levels of arsenic in the watercress as a result of geothermal activity. In these areas it is recommended that watercress collected from local waterways is not consumed. 6. To minimise potential health risks, watercress should not be gathered for consumption from waterways subject to significant historical/current industrial discharges. Note: On 12 June 2000, as a result of the preliminar y results from this study, Regional Public Health issued a public health media release and the Director-General of Health issued a ‘privileged statement’ warning the public that due to bacteriological contamination, they should not eat or serve watercress harvested from creeks, ri vers or streams unless it was washed and cooked thoroughly in boiling water. The statements also advised people selling watercress to inform their customers of this safety warning. Potential health risks from water contact while gathering watercress At many growing locations, gathering watercress e xposes people to contact with contaminated water. To give an indication of potential health risks through water exposure, the E. coli levels in the growing w aters at each site were compared with the 1998 Ministry for the Environment Bacteriolo gical W ater Quality Guidelines for Fresh Water . The measured E. coli levels were compared against the single sample maximum limits in the guidelines. Alert Mode II and Action/Red Modes are triggered when a single bacteriological sample exceeds the predetermined level. At the Action Mode the guidelines recommend that the local authority and health authorities should warn the pub lic through the media that the water body is unsafe and arrange for the local authority to erect signs war ning the public of a health danger.23 All of the sites except for Golf Course Stream exceeded the Action/Red mode level on one or more occasions. Persons gathering watercress from these or similar sites could therefore be at risk of illness from contact (e.g. hand-to-mouth contact, direct skin contact, aerosols) with contaminated water. Health risks from metal contamination Mean metal concentrations in watercress at all the sites were within the New Zealand Food Re gulations (1984) levels. 2004 VOL. 28 NO . 1 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article Acknowledgements This study was supported by funding from the Ministry of Health. Thanks to Hutt Valley District Health Board for additional funding and support. We would like to thank B. Tuau, A. Bichan, S. Giles, E. Kearney, G. O’Leary, B. Johnson and C. Hayes (health protection officers) and the staff at ESR for their assistance. Finally, thanks to Professor Philip Weinstein for his review of the draft paper. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Microbiological and metal contamination of watercress in the Wellington region, New Zealand—2000 survey

Loading next page...
 
/lp/wiley/microbiological-and-metal-contamination-of-watercress-in-the-w0dsUjalK2

References (45)

Publisher
Wiley
Copyright
Copyright © 2004 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1467-842X.2004.tb00627.x
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives: To investigate potential microbiological and metal contamination of watercress and to assess the public health risks associated with harvesting and consumption of watercress. Method: During March and April 2000, samples were taken from 11 known or potential watercress collection sites in the Wellington region. Microbiological testing included bacterial counts for presumptive and faecal coliforms (w atercress); total coliforms (growing water); and Escherichia C. Edmonds Hutt Valley District Health Board, New Zealand R. Hawke School of Earth Sciences, Victoria University of Wellington, New Zealand coli (E. coli) and presence/absence tests for Campylobacter species (g rowing w ater and watercress). Watercress concentrations of a range of metals were also measured. Results: All of the sites showed significant levels of E. coli in samples of both watercress and water. The E. coli levels in water were well above recommended freshwater recreational contact safety guidelines at most sites. Campylobacter was detected in the growing waters at all sites (80% of the samples) and in 11% of the watercress samples. Mean metal concentrations in watercress did not exceed the NZ Food Regulations (1984) levels at any of the sites. However, lead concentrations at the urban sites and one of the semi-urban sites would have exceeded the new Australia New Zealand Food Standards Code maximum le vels (2003). Conclusions and Implications: The consumption of raw watercress contaminated with enteric pathogens could potentially cause serious gastrointestinal illness (e.g. campylobacteriosis) and people gathering watercress could also be at risk of infection from contact with contaminated surface waters. ( Aust N Z J Public Health 2004; 28: 20-6) atercress (Nasturtium spp.) is an aquatic perennial herb that grows wild along the margins of slow-moving rivers, streams, ditches and drains. It is common in New Zealand, especially in the North Island. 1 Watercress is a popular food and is traditionally ser ved cooked; however, it is increasingl y being consumed uncooked in salads or as a gar nish and is being sold as a salad vegetable in restaurants/cafes. At present (2000) watercress is harvested both for personal use and for commercial sale in New Zealand. Since watercress grows in the aquatic environment (sometimes completely submerged), it is subject to potential chemical and microbiological contamination from the water/sediment in which it grows. Sites favourable for watercress usually have comparatively small volumetric flows, which do not allow for dilution of the pollutants they receive and are often either in urban or intensive agricultural areas. Many streams in New Zealand are subject to faecal contamination from urban and rural runoff (both diffuse and point sources), which can introduce enteric pathogens e.g. Campylobacter, into the receiving waters. 2 The consumption of watercress contaminated with enteric pathogens could be a risk factor for gastrointestinal illness and therefore could be contributing to New Zealand’ s high incidence of certain enteric diseases, e.g. campylobacteriosis.3 Persons g athering watercress could also be at risk of infection from contact with contaminated surface waters. Despite the f act that salad vegetables are generally regarded as low-risk foods they can harbour a range of pathogens. Since these foods are not necessarily cooked before consumption they can pose a health risk to the consumer.4 Watercress may pose a signif icantly g reater risk than other common salad vegetables (e.g. lettuce) as it g rows in waterways subject to waterbor ne contaminants. Although watercress has not formally been linked to any enteric disease outbreak in New Zealand, many overseas outbreaks of human gastroenteritis have been linked to the consumption of contaminated fresh vegetables,5 and watercress is included in a list of salad vegetables that have caused extensive outbreaks of salmonellosis.6 Increasingl y, enteric disease research is focusing on environmental f actors and disease and water quality datasets are being linked.7 In ter ms of metal contamination, a number of studies undertaken on central North Island rivers, e.g. the Waikato River, have revealed levels of arsenic in watercress above the World Health Organization limit for foodstuffs. This is attributed to the geothermal activity in the area and the impact of geothermal power stations.8,9 Overseas studies have also demonstrated the ability of watercress to absorb cer tain metals from contaminated water/sediments.10 Submitted: April 2003 Revision requested: August 2003 Accepted: September 2003 Correspondence to: Dr Richard Hawke, School of Ear th Sciences, Victor ia University of Wellington, PO Box 600, Wellington, New Zealand. Fax: +64 4 463 5186; e-mail: richard.hawke@vuw.ac.nz AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ The aims of this 2000 study were: • to investigate potential microbiological and metal contamination of watercress and growing waters from a number of streams in the Wellington and Wairarapa regions; and, • to assess public health risks regarding harvesting and consumption of watercress from these sites. To get a good representation of different growing sites it was decided to sample watercress and growing waters from a number of surface water sites in the greater Wellington region. Watercress and growing waters were sampled for microbiological contaminants and watercress was also sampled for metals. Samples were taken during five weeks in March and April 2000. To assess potential microbiological contamination, the presence of faecal indicator organisms (e.g. E. coli) and a common waterborne pathogen (Campylobacter) was tested. Following discussions with environmental monitoring and health agencies, a number of metals were selected for analysis. These metals could be present in growing water/sediment from industrial discharges or runoff11 and the New Zealand Food Regulations (1984) and the new Australia New Zealand Food Standards Code (2003) specify maximum permissive levels for metals.12,13 Figure 1: Water and watercress sampling locations, Wellington Region, New Zealand. Method Eleven streams in the greater Wellington region were selected as watercress sampling sites (see Figure 1). The sites covered a range of urban, semi-urban and rural conditions and a range of water quality and sediment characteristics. The sites were also selected to be representative of catchment types generally found elsewhere in New Zealand and typical watercress growing environments. Sites of known collection and consumption were targeted in addition to a control site. Of the 11 sites selected for this study, five were known to be sites where watercress was gathered in 2000 on a regular basis either for commercial sale or personal consumption. Sampling technique Watercress was gathered (weekly) from each waterway on five occasions from March 2000. The f ive samples were to allow the temporal variability in potential contamination to be assessed. In the first week, five 300 g watercress samples (healthy whole plants excluding the roots) were collected from each of the sites for individual microbiological analysis in order to be able to compare microbiological results against the Ministr y of Health’s Microbiological Reference Criteria for Food (1995).14 However, for the first week only one sample was collected from the Waiwhetu Stream because there was insufficient watercress available. For the remaining four weeks, five samples (if available) were taken from each site and combined by the laborator y into one sample (for each site) for analysis. Simultaneously, each week, a onelitre water sample was taken from each of the sites for microbiological analysis. Microbiological analysis Microbiological testing of the watercress included bacterial 2004 VOL. 28 NO . 1 counts for presumptive coliforms, faecal colifor ms, Escherichia coli (E. coli) and presence/absence tests for Campylobacter species. Microbiological testing of the growing waters included bacterial counts for total coliforms, E. coli and presence/absence tests for Campylobacter species. The microbiological anal ysis was undertaken by the Institute of Environmental Science and Research Limited (ESR), which is accredited by International Accreditation New Zealand (IANZ) for the test methods: water samples were tested for total coliforms and E. coli using Colilert MPN (IDEXX Laboratories Inc, Maine, USA); Campylobacter species presence/absence was tested for in both watercress and water samples using an adapted method;15 detection and isolation of E. coli, faecal coliforms and presumptive coliforms followed the APHA standard methods.16 Since watercress is commonly eaten as a salad vegetable the results were compared with the Microbiological Reference Criteria for Food (1995) section 5.25 (salads). This reference recommends limits for f aecal coliforms of n=5, c=2, m=102, M=103 where: • n=number of samples. • c=maximum allowable number of defective samples. • m=an acceptable le vel and values above it are marginally acceptable (i.e. 2 samples) or unacceptable (i.e. >2 samples). • M=a le vel which separates marginally acceptable quality from defective quality. Any sample above M is unacceptable. Overseas studies suggest that these criteria may be inappropriate for plant material. Bacteria such as Klebsiella and Enterobacter species, which are normal inhabitants of plant material, gi ve a positive faecal coliform test, but their presence in foods of plant origin may not necessarily be associated with faecal contamination.17-19 These authors recommend that the faecal coliform test be replaced by analysis for E. coli when testing plant-derived material. Therefore, in this study (as in the study by Little et al. 199720) the f aecal coliform limits recommended in the Reference Criteria for Foods (5.25) were used with E. coli rather than f aecal colifor ms. For the first week all sites (except Waiwhetu) had five samples AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article analysed and could be directly compared with the reference criteria. For subsequent w eeks where samples were combined, the result was considered to be an average and was considered non-compliant if greater than M. There are no criteria for Campylobacter in the Microbiological Reference Criteria for Food (5.25) relating to salads. However, the criteria for Campylobacter in Foods – Cooked Ready-To-Eat, (5.8b) “Some components not cooked in manufacturing process (e.g. sandwiches)”, is 0/10 g Campylobacter. This criterion defines that Campylobacter contamination is unacceptable in foods that will not be cooked prior to consumption. The presence/absence test for Campylobacter was therefore considered sufficient to determine the safety of raw watercress for human consumption. zinc and mercury) to give an indication of potential contamination. Five combined watercress samples in total were tested for metals from each site for five weeks. Metal analyses, b y ICP-MS, were undertak en byAgriquality New Zealand Limited using methods that have been accredited by International Accreditation New Zealand (IANZ). Results Results of microbiological analyses of the watercress samples are summarised in Table 1 (E. coli ) and Table 2 (Campylobacter). For the f irst week, none of the sites met the acceptable reference criteria for salads. Six sites were marginally acceptable and four sites were unacceptable. During the remaining four weeks, the composite samples were above M on six occasions at f ive sites. There did not appear to be any difference between catchment types. For example, the Golf Course Stream, Black Stream, Owhiro Stream and Papawai Stream were all classif ied non-complying in the f irst week and included streams flowing through urban areas, streams flowing through rural areas and streams flowing through native forest (bush) areas. The intra-site variability observed from the f ive samples collected in the first week showed considerable variation was possible. Campylobacter was present in 11% of the watercress samples. There did appear to be a difference between catchment types with Campylobacter being found in r ural catchments (e.g. Opaki, Papawai, Ohariu), semi-urban catchments (e.g. Mazengarb) and the Golf Course Stream site, but not in samples from catchments with significant urban development. Campylobacter was more common in the water samples than in the watercress samples: Metal analysis The concentrations of a range of metals in the watercress were measured (i.e. arsenic, chromium, lead, cadmium, nickel, copper, Table 1: Microbiological (E. coli) results of watercress samples by site compared with Microbiological Reference Criteria for Food – section 5.25 (salads). a Growing site Acceptable b Week 1 (five samples) Subsequent four weeks Marginally NonNonacceptablec complyingd complyinge Hulls Creek (urban) Black Stream (urban) Waiwhetu Stream N/A (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) Papawai Stream (rural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) Yes Yes N/A Yes Yes Yes Yes Yes N/A Table 2: Microbiological (Campylobacter) results of watercress samples by site compared with Microbiological Reference Criteria for Food – cooked ready-to-eat (5.8b).a Analysis Campylobacter presence (watercress 1st week) (0/5) (0/5) (0/1) (3/5) (0/5) (0/5) (0/5) (0/5) (0/5) (2/5) (0/5) Campylobacter presence (watercress total) (0/9) (0/9) (0/5) (4/9) (0/9) (1/9) (1/9) (0/9) (0/9) (3/9) (1/9) Hulls Creek (urban) Yes Yes Yes Black Stream (urban) Waiwhetu Stream (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (r ural) Papawai Stream (r ural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) Notes: (a) The new Food Standards Australia New Zealand (2001) Guidelines for the microbiological examination of ready-to-eat foods state that the presence of E. coli in ready-to-eat foods is undesirable and levels exceeding 100 per gram are unacceptable and indicate a level of contamination that may have introduced pathogens.21 (b) Acceptable: E. coli count first week (5 samples <100/g). (c) Marginally acceptable: E. coli count first week (2 samples >100/g and <1000/g). (d) Non-complying: E. coli count first week (3 or more samples >100/g or any sample >1000/g). (e) Non-complying: E. coli count subsequent 4 weeks (>1000/g, composite sample). Notes: (a) The new Food Standards Australia New Zealand (2001) Guidelines f or the microbiological examination of ready-to-eat foods state that Campylobacter should not be present in ready-to-eat foods as consumption of foods containing this pathogen may result in food-bor ne illness.21 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ Campylobacter was present in 11% of the watercress samples compared with 80% of the water samples. Differences in sample size and physical proper ties between water and watercress make a direct comparison difficult. For example, Campylobacter presence on watercress is likely to be affected by se veral factors including: the surface area of the watercress sample and the resulting water v olume retained; water temperature; and the retention and accumulation of bacteria on plant surfaces. Since the Campylobacter data is only presence/absence it gives no indication of the actual numbers in each sample, i.e. whether heavily or lightly contaminated. Some degree of Campylobacter contamination would be expected in catchments subject to faecal contamination as Campylobacter is excreted by a wide variety of animals and birds.22 However, the Microbiological Reference Criteria def ine any level of Campylobacter contamination as being unacceptable in foods that will not be cooked prior to consumption. Figure 2: Water vs. watercress E. coli counts: Owhiro Stream. 3000 E. coli (Water MPN/100ml; Watercress MPN/g) Water E. coli Watercress E. coli 2500 2000 1500 1000 500 0 4-Apr 9-Apr 14-Apr 19-Apr 24-Apr 29-Apr Sampling Date Water samples Applicab le standards or guidelines for acceptable levels of E. coli or Campylobacter in growing waters could not be found. Therefore the presence/absence test for Campylobacter in growing waters w as considered appropriate as any lev el of Campylobacter in growing waters could potentially contaminate watercress. Gathering watercress e xposes people to the risk of infection; for example, through cuts and abrasions, splashes to the eyes and mucus membranes, hand-to-mouth activity (e.g. eating and smoking) and exposure to aerosols. Hence the Ministry for the Environment’s 1998 Bacteriological Water Quality Guidelines for Marine and Fresh Water23 were used to assess potential health risks from water contact through harvesting watercress. These guidelines specify two categories: Alert/Amber Mode II if a single sample is greater than 273 E. coli/100 mL, and Action/Red Mode if a single sample is greater than 410 E. coli/100 mL. Results of microbiological analyses of the water samples are summarised in Table 3. Consistent with previous work, there was no apparent relationship between the presence of Campylobacter and the E. coli counts at the sites.24 The urban streams had greater numbers of elevated E. coli results (e.g. above 2,400 MPN/100 mL) than the predominantly semi-urban and rural streams. This may be due to the run-off characteristics of sealed surfaces; that is, in rural areas, soil and plants may retain micro-organisms before they reach surface waters. As outlined above with Camplyobacter, differences in sample size and physical properties between the water and the watercress samples make a direct comparison difficult. In general, the water E. coli count was higher than the watercress count, but this was not always the case (see Figure 2). All the sites, with the exception of the Golf Course Stream site, exceeded the Action/Red Mode on one or more occasions (see Table 3). Hence, there is a potential health risk from water contact when collecting watercress from surface waters in most catchment types. Table 3: Microbiological results of water samples by site. Analysis E. coli 100 mL Campylobacter >273 >410 >2,400 presence Alert/Amber Action/Red MPN/100mL Mode IIa Modea Hulls Creek (urban) Black Stream (urban) (5/5) (4/5) 1 Waiwhetu Stream (3/5) (urban) Mazengarb (4/5) Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) (4/5) (5/5) Papawai Stream (5/5) (rural) Parkvale Stream (5/5) (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) (1/5) (5/5) (3/5) Watercress metals results The sample results were compared with the New Zealand Food Regulations (1984) permissible proportion levels (in ppm) for Note: (a) Ministry for the Environment Bacteriological Water Quality Guidelines.23 2004 VOL. 28 NO . 1 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article selected metals. There are no regulation levels for chromium and nickel (see Table 4). In this study all mean metal concentrations were below the maximum permitted limits in the New Zealand Food Regulation (1984) levels. It is noted, however, that mean concentrations of lead observed in the urban streams and one of the semi-urban streams (see Table 4) would exceed the replacement limit in the Australia New Zealand Food Standards Code.13 Watercress samples from the urban streams and one of the semi-urban streams had higher mean metal concentrations of zinc, copper and lead than the other sites. Discussion Microbiological Campylobacter was uniformly present in waters over a range of catchment types. Thus these sites are unsuitable for gathering watercress for consumption unless thoroughly cooked (e.g. boiling). The presence of Campylobacter on 11% of the watercress samples is of public health significance. Campylobacter has been shown to survive long enough on ready-to-eat (RTE) produce to cause infection.5 The Microbiological Reference Criteria for Food (section 5.8b) for cook ed ready-to-eat foods is 0/10 g Campylobacter. There is also the potential for cross contamination of other foods and preparation surfaces/utensils from processing contaminated raw watercress. All the watercress and growing water samples at each site showed variable levels of E. coli contamination. High E. coli counts would be expected in slow-moving streams where watercress grows in sufficient quantities for collection. Little et al. (1997) describes the results of a large salad survey in England and Wales in which 2,552 samples were examined and used the same limits that were used in this study (i.e. E. coli, n=5, c=2, m=102, M=103).20 That survey found only 1% of salads had E. coli counts of 102 cfu or more per gram. In this study, 46% of watercress samples had E. coli counts of 10 2 cfu or more per gram. This indicates significant faecal contamination of watercress from contaminated growing waters. While it is not possible from the tests used to know the origin of the E. coli, sources such as livestock or birds can act as reservoirs of micro-organisms that can cause human disease. If faecal contamination (as indicated by E. coli results) is present on watercress then human pathogens may also be present, posing a risk of infection. Pathogens such as Salmonella , Camp ylobacter, Giardia and Cryptosporidium and viruses are potential contaminants of watercress especially in rural situations, where livestock can car ry and excrete patho gens. There is no formal data available on the incidence of microbial hazards from watercress grown in New Zealand. However, many overseas outbreaks of human gastroenteritis (e.g. Shigella (USA)), have been linked to the consumption of contaminated fresh vegetables. Watercress is included in a list of salad vegetables that have caused extensive outbreaks of salmonellosis.5,6 Studies have shown that thorough washing and treatment of produce with chlorinated water can reduce the populations of pathogenic micro-organisms on fresh produce but it cannot eliminate them. Beuchat (1997) suggests that the reduction of risk for human illness associated with raw produce can be better achieved through controlling points of potential contamination e.g. during growing, harvesting, processing and distribution. In regard to raw watercress, strictly monitored controls on the quality of the g rowing waters would be the most appropriate method of reducing health risks for consumers. 25 Table 4: Metal results by site (mean concentrations a). Nickel (ppm) Permissible le vels Food Regulations (1984)b Hulls Creek (urban) Black Stream (urban) Waiwhetu Stream (urban) Mazengarb Stream (semi urban) Owhiro Stream (semi urban) Ohariu Stream (rural) Papawai Stream (rural) Parkvale Stream (rural) Manaia Stream (rural) Opaki Stream (rural) Golf Course Stream (bush) 0.12 0.13 0.07 0.06 0.05 0.05 0.06 0.05 0.07 0.05 0.24 Cadmium (ppm) 1 ppm Copper (ppm) 30 ppm Zinc (ppm) 40 ppm Chromium (ppm) Arsenic (ppm) 2 ppm Lead (ppm) 2 ppm Mercury (ppm) 0.03 ppm Notes: (a) Results are repor ted by wet weight. (b) The A ustralia New Zealand Food Standards Code was adopted in New Zealand in Februar y 2001 and took full effect on 20 December 2002.13 The code has maximum levels of metal contaminants in food for lead (0.1 mg/kg) and cadmium (0.1 mg/kg) that can be applied to watercress; these levels are lower than the le vels in the New Zealand Food Regulations (1984). While all the results for cadmium were below this level, mean lead levels at the urban sites and one of the semi-urban sites exceed the new limit. AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH VOL. 28 NO. 1 Environmental Health Contamination of watercress in Wellington, NZ Testing for pathogens as an indicator of watercress or water quality is not cost effective because of the potential number of pathogens that could be present. It is more appropriate to use a suitable faecal indicator organism e.g. E. coli, as an indicator of the growing water quality and apply an appropriate standard (e.g. New Zealand Drinking Water Quality Guidelines (2000) standard of less than 1 E. coli /100 mL), to ensure minimal potential for watercress contamination. 26 Fasciolosis Liver flukes (Fasciola hepatica) are parasitic in the bile ducts of mammals. Cercariae encyst on aquatic vegetation (e.g. watercress) and the cysts (metacercariae) are then swallowed by the final host (humans, sheep, goats, cattle) while feeding. Wild watercress is reported as the main source of infection in Europe, where there is a high rate of endemic f asciolosis in domestic animals. 27 Over the past 30 years or so, liver fluke infection of livestock has become considerably more widespread in New Zealand, largely in association with the spread of the exotic snail host, Lymnaea columella.28 Although there have been no documented cases of fasciolosis in New Zealand related to consumption of watercress gathered from New Zealand waterways, there is potential for cases of fasciolosis through consumption of raw watercress. For example, a case of fasciolosis was recently reported in Australia and infection was attributed to the consumption of watercress.29 Any future risk of fasciolosis would be minimised by ensuring that raw watercress for human consumption is grown only under strictly controlled conditions. 27 However, the mean concentrations of lead observed in the urban streams and one of the semi-urban streams (see Table 4) exceeded the maximum per mitted levels in its replacement, the Australia New Zealand Food Standards Code.13 Overseas and NZ studies have shown that watercress can bio-accumulate metals to levels that could pose health risks. For example, studies undertaken in areas of the central North Island have revealed high levels of arsenic (in excess of the WHO limit for arsenic in foodstuffs) in the watercress, probably because of the geothermal activity in the area.9,30 A specif ic site risk assessment would need to be undertaken to determine potential metal contamination before the safety of consuming watercress gathered from a site could be assessed. Recommendations 1. Watercress harvested from any uncontrolled surface w ater source in New Zealand should not be consumed unless the watercress is thoroughly cooked (e.g. in boiling water) to destroy potential human pathogens. 2. Watercress should not be eaten raw unless it can be demonstrated that the growing environment is strictl y controlled and effectively monitored to ensure the water source is of suitable standard e.g. controlled hydroponic cultivation. A suitably strict standard should be applied to the growing waters e.g. New Zealand Drinking W ater Quality Guidelines (2000) E. coli standard of less than 1 E. coli/100 mL, to monitor acceptable watercress microbiological quality.26 3. Further research should be under taken to assess the risk of contracting waterborne illnesses through contact with contaminated water when gathering watercress from uncontrolled surface waters. 4. The potential for fasciolosis in New Zealand from consumption of wild watercress should be the subject of a targeted study. This would involve further research on such f actors as the geographic distribution of the host snails and the geographic distribution of Fasciola hepatica in livestock. 5. There are areas in New Zealand, specifically the central Nor th Island, where studies have revealed high levels of arsenic in the watercress as a result of geothermal activity. In these areas it is recommended that watercress collected from local waterways is not consumed. 6. To minimise potential health risks, watercress should not be gathered for consumption from waterways subject to significant historical/current industrial discharges. Note: On 12 June 2000, as a result of the preliminar y results from this study, Regional Public Health issued a public health media release and the Director-General of Health issued a ‘privileged statement’ warning the public that due to bacteriological contamination, they should not eat or serve watercress harvested from creeks, ri vers or streams unless it was washed and cooked thoroughly in boiling water. The statements also advised people selling watercress to inform their customers of this safety warning. Potential health risks from water contact while gathering watercress At many growing locations, gathering watercress e xposes people to contact with contaminated water. To give an indication of potential health risks through water exposure, the E. coli levels in the growing w aters at each site were compared with the 1998 Ministry for the Environment Bacteriolo gical W ater Quality Guidelines for Fresh Water . The measured E. coli levels were compared against the single sample maximum limits in the guidelines. Alert Mode II and Action/Red Modes are triggered when a single bacteriological sample exceeds the predetermined level. At the Action Mode the guidelines recommend that the local authority and health authorities should warn the pub lic through the media that the water body is unsafe and arrange for the local authority to erect signs war ning the public of a health danger.23 All of the sites except for Golf Course Stream exceeded the Action/Red mode level on one or more occasions. Persons gathering watercress from these or similar sites could therefore be at risk of illness from contact (e.g. hand-to-mouth contact, direct skin contact, aerosols) with contaminated water. Health risks from metal contamination Mean metal concentrations in watercress at all the sites were within the New Zealand Food Re gulations (1984) levels. 2004 VOL. 28 NO . 1 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Edmonds and Hawke Article Acknowledgements This study was supported by funding from the Ministry of Health. Thanks to Hutt Valley District Health Board for additional funding and support. We would like to thank B. Tuau, A. Bichan, S. Giles, E. Kearney, G. O’Leary, B. Johnson and C. Hayes (health protection officers) and the staff at ESR for their assistance. Finally, thanks to Professor Philip Weinstein for his review of the draft paper.

Journal

Australian and New Zealand Journal of Public HealthWiley

Published: Feb 1, 2004

There are no references for this article.