This website uses non-intrusive cookies to improve your user experience. You can visit our cookie privacy page for more information.

PN0907 - Potential Exposure of Birds to Treated Seed

PN0907 - FERA2001

Executive Summary

  1. Seed treatments are widely used for crop protection and present a particular risk to birds which eat seeds. Risk assessment for seed treatments has tended to focus on one or two highly granivorous species, especially woodpigeon and house sparrow. Recent evidence from FERA’s research (project PN0902) has shown that, at least under some conditions, small and non-granivorous birds will take wheat seeds. There is therefore a need for better data on which species take these and other types of seeds for which pesticide treatments are used.
  2. This project’s primary objective was to identify which bird species will take and eat a range of crop seeds in common agricultural usage. The project addressed this question by recording birds visiting bait stations at which about a kilogram of seed was presented. This quantity of seed is unlikely to be available to birds in one place under normal conditions, but this approach was necessary to ensure that as many as possible of the species potentially at risk would be detected. The secondary objective was to obtain information on how much seed is taken by individual birds.
  3. In autumn/winter 1997/98 wheat seed was placed on bait stations on mixed farmland, and video cameras used to record bird activity at them throughout daylight hours for up to four weeks. In spring 1998, autumn and winter 1998/99 and spring 1999, similar bait stations were set up using barley, maize, oilseed rape, grass, peas and sugar beet seed. They were pre-baited for three weeks before recording for two days in the fourth week.
  4. Table 1 (see the full document) provides the key results in a form intended for use in risk assessment. It shows the species seen taking seed at stations baited with various seed types. For each seed type, the species are arranged in order of exposure potential. Species at the top of the list would receive the highest doses of seed treatment under the assumptions normally used in initial risk assessment, due to their lower body weights.
  5. Calculations presented in the report show that, for many of the species, the amounts of seed consumed during single visits to the bait stations were sufficient to pose a potential risk (toxicity-exposure ratio < 10) if the seed had been treated with one of the more acutely toxic seed treatments. This does not necessarily imply that mortality is occurring (other factors such as avoidance may reduce risk), but it does emphasise the need to consider these species in risk assessment. As the potential worst-case risk is highest for the smallest species, one of these (or a generic ‘small bird’) should be used as the representative species for the initial, conservative stage of risk assessment.
  6. The species in Table 1 should be regarded as those most likely to be exposed in a typical farmland setting, and not a complete listing. Some additional species are known to have been involved in poisoning incidents with seed treatments in the past, but were not recorded at bait stations in this study. They should also be considered in risk assessment for all the tested seed types except grass and pelleted beet seed. Taken together, the number and 6 types of species listed (including those in the legend) are sufficient for the result of a risk assessment covering these species to be extrapolated to any other species which may not have been recorded by the method used.
  7. Table 1 lists the maximum amounts of seed taken during single visits to the bait stations for each species. The use of these data in risk assessment is considered in detail in the Discussion. It is important to remember that these data represent intake for single feeding bouts and should not be used to estimate total daily intake, which may be much higher. Therefore, theoretical estimates of total daily intake should continue to be used to estimate exposure in the initial stage of risk assessment, as it is intended to be conservative. For convenience, estimates of total daily intake calculated using methods developed in Project PN0908 are included in Table 1. If the initial assessment indicates a potential risk, the data on single visit intakes are appropriate as worst case estimates for acute, short-term exposures. These exposures merit special attention for seed treatments as they are a significant source of risk for this type of pesticide use. Exposure assessments over periods of a day or more will rarely require consideration of individual feeding bouts.
  8. Grass and pelleted beet seed are clearly less attractive than other types of seed. However, significant exposure to pesticides on these types of seeds might occur in cases where the pesticide loading per seed is high (eg approaching one LD50 per seed for sensitive species based on the extrapolation factors of Luttik and Aldenberg, 1997) and availability is high (ie seed broadcast, not drilled).
  9. Previous studies have shown that dehusking of seeds can substantially reduce avian exposure in at least some cases. The present project provides valuable information on which of the species recorded actually dehusk which seeds, in field conditions. These species are marked with asterisks in Table 1. It is important to note that dehusking is not all-or-nothing: not all small species dehusk, and some species dehusked some but not all of particular seed types, so dehusking should not be included in the initial, worst-case assessment. The treatment of dehusking in refined risk assessments is considered in the Discussion.
  10. The lists of species in Table 1 should also be helpful in identifying suitable species for use in special studies, which may be required in refining the assessment when the initial assessment indicates a potential risk. This includes the selection of species for captive studies such as avoidance testing, and the selection of suitable sites and focal species for field studies.
  11. Finally, the lists in Table 1 can be used again at the end of the risk assessment, to help the assessor check that the assessment result can be extrapolated with reasonable confidence to all the species which are likely to be exposed. This is necessary because risk assessments usually focus on a limited number of representative species, and is an important final step to ensure that the goals of the assessment are fully achieved.

Full Report

Potential Exposure of Birds to Treated Seed (pdf, 53 pages).

Further information

Updated 2016-12-21