Cylindrospermopsin

As for microcystins, plant studies have addressed the uptake of cylindrospermopsin into leafy vegetables such as lettuce, arugula and mustard, sometimes at higher levels than microcystins (Table 5.5), with one study finding CYN in spinach at levels higher than other leafy greens, possibly resulting in exposure at levels relevant to health. Llana-Ruiz-Cabello et ah (2019) found high concentrations in spinach and lettuce, but only when applying CYN in concentrations of 25 pg/L together with 25 pg/L of MCs; applying 25 pg/L CYN alone resulted in fourfold lower concentrations in the plant material. However, such levels have not been reported from field studies or market acquired vegetables. In animals, CYN has been studied less than MCs, with accumulation reported from bivalves and crustaceans, in one case at levels potentially relevant to health (Table 5.5). For fish, studies with sufficiently selective methods (e.g., LC-MS/MS) are largely lacking; the review by Testai et al. (2016) includes three studies that found no or only very low concentrations in fish.

Saxitoxins

Saxitoxins (STXs) in food products are well documented in the marine environment, including numerous cases of human illness and death. To date, there have not been any reports of paralytic shellfish poisoning caused by freshwater cyanobacteria even though STX accumulation in freshwater mussels has been demonstrated (Negri & Jones, 1995). Freshwater fish, Oreocbromis niloticus and Geopbagus brasiliensis, were found to accumulate STXs from the environment, but not in concentrations that would lead to exposure in a health-relevant range (Table 5.5; Galvao et al., 2009). Testai et al. (2016) include one study finding up to 30.6± 14.5 pg/kg of PSP toxins in Cichlidae. Interestingly, intraperitoneal dosing of the tropical freshwater fish Hoplias malabaricus four times with STX at 800 pg/kg did not result in accumulation in muscle tissue (da Silva et al., 2011).

Anatoxins

Very little is known regarding the accumulation of anatoxin-a and/or homoanatoxin-a, with studies lacking on crops or invertebrates. One study has shown anatoxin-a to bioaccumulate in fish (Osswald et al., 2011), but others have shown it to rapidly eliminate from fish and mussels (Osswald et al., 2008; Colas et al., 2020).

Conclusions on exposure via food

In summary, as preliminary assessment considerering all types of cyanobac- terial toxins, the data available by 2019 do not point to a high level of shortterm exposure to cyanotoxins in crops or muscle tissue of fish and crayfish, whereas exposure may be more significant if viscera are eaten, as is the case for small fish, crustaceans and mussels. If for instance crops are sprayed or irrigated with lake water containing scums or high levels of cyanotoxins and in particular if foods are not sufficiently washed or prepared, risks may be higher. However, data obtained with reliable methods are insufficient for drawing clear conclusions.

Where crop irrigation with scum material is widespread (as described, e.g., in Li et al., 2014) or fish, mussels and crayfish from bloom-ridden waterbodies constitute staple foods, screening cyanotoxin concentrations in such foods is recommended, with attention to the methodological requirements described in section 5.3.4.

Hazard analysis for any of the above settings may indicate that when cyanotoxins in foods cannot be excluded because of - substantial - cyano- bacterial blooms in the waterbody used for the production of the food, a more detailed analysis becomes important. Checklist 5.2 provides guidance for conducting such an analysis.

CHECKLIST 5.2 FOR ASSESSING THE RISK OF CYANOTOXIN EXPOSURE THROUGH FOOD

1. Are blooms of potentially toxic cyanobacteria present in the waterbodies

used for collecting, producing or preparing food (see Chapters 4 and 8)?

• 1.1. Inspect these waterbodies to collect information on the presence of surface blooms or scums, strong greenish discoloration and turbidity.
• 1.2. Collect samples for species identification and quantification, particularly if these observations indicate cyanobacteria could be present.
• 1.3. Particularly if potentially toxic cyanobacteria are found and if feasible, have toxin content of the cells and bloom analysed (see point 2.3).
• 1.4. If cyanotoxins are present currently or were present during the previous month, further risk analysis in food becomes relevant. Clarify the time pattern of toxin occurrence - is it sporadic for a few days, or continuous for many weeks or months?
• 2. Are organisms (e.g., fish, shellfish, snails, bivalves) harvested for food

from the impacted waterbodies? If so,

• 2.1. Find out whether these species are likely to filter-feed particles, including cyanobacteria, and whether they have been reported to contain cyanotoxins.
• 2.2. Find out whether viscera and gonads are removed prior to consumption or whether the organisms are consumed whole.
• 2.3. Check whether analyses of their cyanotoxin content are feasible, and if so, together with experts derive a plan for sampling and analyses.
• 3. Are crops irrigated with water containing high amounts of cyanobacteria?
• 3.1. If so, check whether the use of alternative water sources, free of blooms, is feasible or run a programme of sampling and analyses to assess whether the practices used lead to cyanotoxins in the crop.
• 3.2. Investigate whether substantial amounts of cells cling to the surface of fruits or vegetables which are potentially consumed without sufficient treatment to remove them.

• 4. Are soils augmented with sediment dredged from systems containing high amounts of cyanobacteria? If so, check dredged material for cyano- toxins and - depending on the results - also the crop.
• 5. Find out whether other exposure pathways to these cyanotoxins are likely (drinking-water or recreation)? If so, estimate the dose from these and determine the proportion from food which is most appropriate for your setting.
• 6. Estimate the contribution of the affected foods to the local diet and the time spans of their contamination with cyanotoxins.
• 6.1. Is it consumed seasonally or year-round? On a daily basis, or occasionally? Are exposure patterns likely to be short term and occasional (justifying assessing exposure in relation to a short-term tolerable daily intake, TDI) or more likely to be continuous for many weeks on end and several times a week (necessitating application of a TDI for chronic lifetime exposure)?
• 6.2. Estimate the amounts consumed and the impact of local traditions for collecting and preparing these foods on exposure pathways.
• 7. Clarify the tolerable cyanotoxin dose from food in the local setting together with toxicologists, taking points 5 and 6 into account. Note that in deriving the WHO guideline values for chronic exposure via drinking- water, WHO apportioned 20% of intake to other sources, including food, while the short-term values are based on exposure only to drinking-water. As discussed above (see point 5 of this checklist), this apportionment may need to be adjusted locally, depending on other exposure routes and the contribution of foods containing cyanotoxins to the local diet.
• 7.1. From the results of local analyses and/or published data on the potential toxin content of these foods (see Table 5.5 and section 5.3.2) and the dose found tolerable for food in your setting, estimate how likely the cyanotoxin contents in the edible parts of these organisms are to exceed that tolerable dose and by how much.
• 7.2. If restricting access to fish, mussels and shellfish is considered, what are the consequences for overall local diet? Are suitable alternatives available, accessible and accepted?
• 7.3. If restricting access to fish, mussels and shellfish is considered and access to alternative protein food sources is poor or in question, how high is the uncertainty of the information base on cyanotoxin content in these foods? Does the information show a sufficiently substantial risk to justify the loss of this food source?
• 8. Are measures in place to control cyanotoxin contamination of food or exposure to potentially contaminated food (see Table 5.5)? Are they sufficient, or are further measures needed?