Risk & economic analyses

Assessing Risks, Regulatory Issues, and Socioeconomic Impacts of Proposed Mycotoxin Control Strategies



Theme 4 of the current project ties together the previous three themes in terms of understanding the overall impact of mycotoxin control through the methods developed in this project: economic, environmental, and regulatory. We aim to estimate the socioeconomic impacts of these mycotoxin control methods in terms of reduced mycotoxin levels in US corn, and corresponding economic impacts to corn growers. Additionally, we are assessing potential environmental impacts and regulatory requirements for these methods.


Economic Impact of Mycotoxins in the U.S. (See also Mitchell, et al 2016)

Economic impacts of mycotoxins within the US are multifaceted, and include:

  • Control and surveillance costs incurred by farmers
  • Handling and testing costs incurred by grain elevators
  • Efficiency and productivity losses incurred by livestock producers
  • Losses in the export market
  • Crop insurance pay-outs incurred by the government and/or other insurers 

However, reliable economic analyses and modeling of economic losses due to mycotoxin contamination are difficult, because state-by-state, publicly-available data on mycotoxin occurrence in US corn is almost non-existent. Modelling difficulties also arise as a result of high variability, which is an inherent characteristic of mycotoxin contamination of grains. Aflatoxin contamination levels in corn can be highly variable, across different farms, between individual grain lots from the same farm, and even among individual kernels within a lot. This heterogeneity makes sampling difficult; sampling variability accounts for the majority of variance in mycotoxin analyses (Johansson et al. 2000). Different aflatoxin test methods also vary in laboratory sample and test sample size, among other characteristics, which also adds a further degree of variability. Lastly, the baseline price per bushel of corn varies by location, as do discounts - both in terms of amounts and the contaminat range(s) for which disounts are assessed - for corn with aflatoxin present. 

For the purpose of this study, we were able to obtain three years (2011-2013) worth of data from Texas and Illinois, and one year (2012) of data from Iowa. Thus, it was necessary to extrapolate these concentration data to the corn crops in other climatologically-similar states as follows (see the map below): Iowa concentration data was applied to the Upper Midwest region (ND, MN, IA, WI and MI), while the Ohio Valley (MO, IL, IN, KY, TN, OH, WV and VA) were assumed to be similar to Illinois, and Texas results were used for the South/Southeast/Southwest region (AZ, NM, TX, OK, KS, AR, LA, MS, AL, GA, SC, NC and FL). Remaining states were assumed to either have negligible aflatoxin contamination (due to climates that are not conducive to A. flavus growth) or no appreciable corn production.





The TX, IL and IA data are shown in the following graph, with each year's data broken into three contaminant ranges. Corn with 20 ppb or less of aflatoxin is suitable for all uses (human and animal), and is generally not subject to any discount at the point of purchase. Corn with aflatoxin levels between 20 and 100 ppb, although deemed by the FDA to be not suitable for consumption by humans, pets or dairy animals, may be used for other livestock feeds, and is generally subject to some price discount. FDA further restricts corn with 101-200 ppb (suitable for finishing swine or feedlot beef cattle or non-animal uses), 201-300 ppb (restricted to feedlot beef cattle or non-animal uses only), and 301-500 ppb (non-aimal uses only) of aflatoxin. Corn with >500 ppb aflatoxin is unusable for any purpose (note however that some individual states place the unusable/non-salvageable cutoff at 200 ppb, rather than 500 ppb).


Aflatoxin levels in corn samples from Texas, Illinois and Iowa for 2011-2013.



Because aflatoxin is so variable within lots (see above), sampling plans can result in two types of statistical errors:

  • Samples can falsely test over a regulatory level and the lot can be rejected. This is Type 1 error or false positive.
  • Samples can falsely test under a regulatory level, allowing a lot that should have been rejected into general commerce. This is a Type 2 error or false negative.

These scenarios for regulatory and quality control programs are often evaluated, and costs and benefits weighed, through the use of operating characteristic curves (OC curves). OC curves represent the likelihood of acceptance as a function of the actual level of aflatoxin in a lot, based on the variability in the sampling protocol and the actual levels and distribution of the toxin in the lot. We used software from the Food and Agriculture Organization of the United Nations (FAO) to generate an OC curve that would determine the probabilities of obtaining false positives or false negatives if the rejection limit is set at 20 ng g-1 aflatoxin, the action level for corn acceptable into general commerce (i.e. safe for all human and animal uses). This curve, shown below, assumed laboratory sample size of 2.5 kg, from which a single 5-g aliquot is tested.




The first model was a straightforward economic model, which assumed that aflatoxin testing at elevators was conducted in all states in the three regions (which is not currently the case, especially in regions with historically low occurrence of A. flavus), that the measured aflatoxin concentrations in each accepted lot were accurate, and that no false positive measurements (Type 1 errors) or false negatives (Type 2 errors) occurred. Total value of corn produced was calculated for each state across different concentration ranges, and for each mycotoxin for which sufficient data was available. State-by-state baseline average prices were then applied, and the percentages of each state's crop in the various concentration ranges were subjected to either of two discount schedules (TX, KS):



As discussed above, 15 states were assumed to have no aflatoxin contamination within their corn. Running this model for all states under consideration, and for all years for which sufficient region-specific mycotoxin concentration data was available, gives the following results, which are limited to aflatoxin, as data for fumonisin is insufficient to allow meaningful estimates: 



Thus, this first model predicts that - even in a low-impact year like 2013 - aflatoxin contamination could be expected to cost U.S. farmers over $100 million in lost revenue. For climate/weather scenarions similar to the other two years for which data was available, the losses that might be incurred due to aflatoxin were determined by the straightforward economic model to be anywhere from nearly $200 million (2012 climate, Kansas discounts) to over $700 million in a bad year (2011 climate) coupled with more-conservative discount schedules (Texas) across all regions. In fact, these latter figures may be underestimates, as Upper Midwest (IA) data was not available for 2011 or 2013; thus, the estimates for these scenarios derive exclusively from the Ohio Valley and Southern regions. True national losses in a year comparable to 2011, according to this model, may be approach $1 billion.

The second model uses the same regional aflatoxin concentration-range frequencies, baseline prices, and OC curve as the first model, applies only the Texas discount schedule, and allows for Type 1 and Type 2 errors. Including errors in the model results in a range of estimated losses for each scenario; the low end of the range assumes chronic Type 2 errors (under-estimation of aflatoxin, with less-than-appropriate discounts), while the high end is what might be expected under chronic Type 1 errors (discount or rejection of lots due to over-estimates of aflatoxin). The results for each region from this model are as follows: 



The estimates derived using this model are generally in good agreement with the first, and imply that, if aflatoxin testing were required in all three regions, losses to the US corn industry would be in the range of $50 million to just over $1 billion. Our estimates, in both of the models, are focused on direct losses that would be incurred by the farmers and grain elevators, and do not include estimates of losses to the livestock industry or costs incurred for mycotoxin prevention. 

Using these baseline loss totals, information from the other three Themes of this project will inform how the effectiveness of our technologies in reducing aflatoxin and fumonisin, as well as the cost-benefit tradeoffs inherent in each technology, would benefit US corn growers.


Regulatory Environment for Novel Mycotoxin-Control Measures

We have attempted to assess the current regulatory framework for adoption of biocontrol and transgenic technologies to control mycotoxins, as well as likely future regulatory scenarios. As the tools developed in this project are cutting-edge, the regulations have not yet been finalized for some of the technologies, as described in greater detail below. The three mycotoxin control strategies outlined in Themes 1, 2, and 3 focus on, respectively: (1) designing and delivering best management practices to reduce mycotoxins; (2) next-generation biocontrol strains; and (3) transgenic corn lines harboring RNAi constructs that target fungal genes involved in mycotoxin production. Of these, it is expected that educational videos and smartphone apps on mycotoxin control (Theme 1) will not require special regulatory oversight or environmental assessment.

Because of the precedence of registering other biocontrol agents to control toxigenic Aspergillus flavus strains in crops in the US, the regulatory environment is well-established for introduction of biocontrol strains to control aflatoxin.  There have been multiple Federal Register notices since 1999 on tolerance exemptions for one biocontrol strain, AF36, which include safety and exposure assessment considerations (link).

Thus, the outcome of this project that is most likely to be subject to regulation is transgenic (RNAi) corn under development in Theme 3. For the purposes of regulating genetically modified organisms (GMOs), a Coordinated Framework for Regulation of Biotechnology was created within the US Office of Science and Technology Policy in 1986: 51 Fed. Reg. 23302. This Coordinated Framework determined that existing federal statutes were adequate to oversee modern biotechnology, and that risk assessment would be based on the rDNA product, not process, by which the GMO was developed.

Thus far, transgenic corn grown in the US with pest-protective and herbicide-tolerant traits has been subject to the Federal Food, Drug and Cosmetic Act (FFDCA), the Plant Protection Act (PPA), the National Environmental Policy Act (NEPA), the Federal Plant Quarantine Act (FQPA), and the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA). Hence, it is regulated by three governmental agencies: the Environmental Protection Agency (EPA), the US Department of Agriculture (USDA), and the Food and Drug Administration (FDA); as summarized in the table below.


Product class

Lead agency

Federal statutes







Food & additives




Together, these Acts require evaluation of multiple environmental and health impacts, including impacts to agriculture, impacts to non-target species, gene flow, insect resistance development (in the case of plant-incorporated protectants [PIPs]), and potential human allergenicity, toxicity, carcinogenicity, mutagenicity, and teratogenicity. This is done on a voluntary basis with FDA (link), and is required by EPA for plant-incorporated protectants (link). For the types of transgenic crops commercialized in the US today, the Coordinated Framework has proven adequate for regulatory oversight. RNAi-based transgenics, however, pose new challenges. RNAi-based transgenic crops have important differences from the types of transgenic crops thus developed in the US; hence, would face different regulatory requirements. The key difference is that nucleic acids, not proteins, are the novel substances produced in these crops. Hence, allergenicity testing would not be required; allergenicity is a concern specific to proteins.

Whether other types of food safety testing would be required is currently under discussion.  It has been argued that because nucleic acids are regularly consumed by humans, these should be exempt from PIP requirements for safety testing (link).Concerns were raised, however, based on a recent study (Zhang et al. 2012) suggesting that high consumption of plant microRNAs could cause unexpected and potentially adverse effects in mammals; specifically, altering the expression of low-density lipoprotein receptor-associated proteins in mice. Two studies have since challenged these findings (Witwer et al. 2013, Snow et al. 2013), but the question remains whether RNAi-based transgenic crops may pose food safety risks, and thus be subject to food safety testing.

EPA hosted a FIFRA Scientific Advisory Panel (SAP) in January of 2014, specifically dedicated to better understanding the nature and risks of RNAi-based transgenic crops (link). FIFRA SAPs are composed of experts in toxicology, biology, statistics, and other fields to provide independent scientific advice to EPA on health and safety issues related to pesticides.

This Panel generally agreed that there was not likely to be much risk to human health or the environment from RNAi-based transgenic crops. Indeed, allergenicity testing would likely not be required, as it has been in the past for other EPA-regulated transgenic PIPs. Thus, as far as human health, the regulatory process may be about the same or even less for RNAi-based transgenic crops than other more traditional transgenic crops.

However, the Panel was less comfortable with the uncertainty surrounding this technology and its potential impacts in the environment. It was suggested that, for every new RNAi-based PIP developed, different animals had to be chosen for toxicity testing depending on the function of the PIP and the environment in which the crop would be planted, to ensure that non-target organisms would not experience unreasonable harm. From a practical perspective, the challenges with implementing this suggestion would be twofold: the potential costs, and the lag time it may take to determine exactly which non-target organisms ought to be tested in each case

There is the question of whether RNAi-based transgenic corn for mycotoxin resistance would be subject to EPA regulation at all. Under FIFRA, EPA regulates transgenic crops that produce plant-incorporated protectants against pests. This includes, for example, the crystal (Cry) proteins in Bt corn, because these proteins protect the corn from insect damage. However, if the crop does not control the pest (in this case, the fungus), but instead prevents the pest from producing a toxin, does this trait fall under the regulatory oversight of EPA?  This is unclear at the moment.



Snow JW, Hale A, Isaacs SK, Baggish AL, Chan SY (2013). Ineffective delivery of diet-derived microRNAs to recipient animal organisms. RNA Biology 10:6, 1-10.

Witwer KW, McAlexander MA, Queen SE, Adams RJ (2013). Real-time quantitative PCR and droplet digital PCR for plant miRNAs in mammalian blood provide little evidence for general uptake of dietary miRNAs. RNA Biology 10:7, 1-7.

Zhang L, Hou D, Chen X, Li D, Zhu L, Zhang Y, et al. (2012). Exogenous plant MIR168a specifically targest mammalian LDLRAP1: evidence of cross-kingdom regulation by microRNA. Cell Res 22:107-26.