Defra - Department for Environment, Food and Rural Affairs.

Science Search

Science and Research Projects

Return to Science Search homepage   Return to Project List

Robust tests for the acceptance of bait and treated seed by birds - PN0909

Annex III to Directive 91/414 states that seed-dressings, baits and granulars for which
TERa<10 must be tested for acceptance to birds. Methods for testing acceptance (or avoidance) were addressed by an OECD/SETAC workshop in December 1994. The
workshop considered that existing test guidelines could not be relied upon because there was no convincing example of avoidance being demonstrated in the wild, and in some cases substantial mortality has occurred in the wild despite strong evidence of avoidance in captive birds. The workshop concluded that until these uncertainties were resolved it would be prudent to ensure that tests of avoidance were robust by conducting them under realistic conditions.

Recent work in project PN0902 has succeeded in demonstrating avoidance of fonofos-treated seed in the wild, and has developed modified test designs which produce results more consistent with the levels of avoidance and mortality observed in the wild. This provides some confidence that appropriate tests can generate reliable results. However, experience shows that it is very difficult to decide how realistic tests should be, especially regarding the choice of species and test material; the time of year; the degree of energy stress; and the number of treatment levels. Realistic tests are very costly and technically difficult. A better basis for assessing the importance of realism is needed, to assist PSD specialists in setting data requirements and assessing data submitted by registrants.

The results obtained in project PN0902 provide a good basis for assessing the importance of realism in tests of seed treatments, and it is likely that birds will respond to baits in broadly similar ways. However, different mechanisms determine the ingestion of granules and these require more basic investigation before attempting to generalise test design for them. The project will therefore be restricted to seed treatments and baits.

The project will test the effect of varying the choice of species and test material. These factors are given priority because they offer more potential for simplifying and reducing the number of tests required, if it can be shown that appropriately-designed tests with a standard species and the technical active substance are robust. Time of year is potentially important and will be controlled for in this project, and could be considered for investigation at a later date. Other factors, notably group size, feeding rate and the choice of treatment levels, have already been dealt with in project PNO902, and it would be inappropriate to generalise test duration which should be determined using information about the mode of action and metabolism of each active ingredient.


It is not practical to test every combination of species and test material. Instead the project will test the effects of each using reasonable extremes (big species, small dehusking species, small non-dehusking species, laboratory species; formulation vs. technical active substance on standard laboratory feed). The endpoints will be those recommended by the SETAC/OECD workshop: mortality, sub-lethal effects, food intake and body weight. The effects of species and test material on these endpoints will be evaluated, and comparisons made with available data on poisoning incidents in the wild.

To provide confidence that the results apply generally it would be necessary to test several active ingredients and show the results were consistent. Active ingredients with a documented history of causing wildlife incidents are preferable, to guide the choice of conditions for the more realistic tests and to enable the reliability of the results to be assessed. This project will examine at least two active substances. In the first year we propose to test the effects of species and test material for fonofos, because it is well understood from results of PNO902 including data on the actual frequency of mortality in wild. One option for the second year would be to repeat the tests using carbophenothion, because it has caused higher mortality in wild involving a different family of birds. Another option would be to test fewer combinations of species and test material, but for more active substances. These issues will be decided in consultation with PSD in the light of the results from the first year.
Investigate the importance of realism in tests of avoidance for baits and seed treatments, and
its effect on the reliability of risk assessments.
Project Documents
• Final Report : Robust tests for the acceptance of bait and treated seed by birds   (117k)
Time-Scale and Cost
From: 1997

To: 1999

Cost: £78,008
Contractor / Funded Organisations
Central Science Laboratory
Fields of Study
Pesticide Safety