Testing of chemicals

Avian Toxicity Testing


This work describes the various endeavours of OECD countries to develop harmonised methodologies for birds toxicity testing, including the successful outcomes or the reasons for stopping activities in some cases. The work has required commitment from countries and industry in terms of manpower and laboratory resources to perform experimental testing and generate data. It was decided by the OECD Working Group of the National Coordinators of the Test Guidelines Programme (WNT) to create this information page for the public to provide useful background for those who are interested in avian chemical toxicity testing with the view to protect wildlife birds.



A number of avian toxicity studies projects were included in the Test Guidelines Programme over the last twenty years, following an OECD survey conducted in 1993 by the Pesticides Programme. The objective of the survey was to set priorities for the revision and development of Test Guidelines used to generate toxicity data prior to pesticides registration.

The OECD work on birds toxicity testing started in 1994 with a SETAC/OECD Workshop on Avian Toxicity Testing, held in Pensacola, Florida, in December 1994. The report of the workshop was published in 1996 in the series on Testing and Assessment, No. 5. The workshop identified four priority areas on which experts from OECD countries worked collaboratively between 1994 and 2015, as further explained below.

OECD work on acute toxicity testing and outcome

The Pensacola Workshop recommended that test designs for three acute oral toxicity studies be developed: (1) an initial (first tier) limit test to identify toxic chemicals; (2) a definitive LD50 test (usually for a single species), normally triggered by the limit test; and (3) an approximate lethal dose test in additional species, triggered if greater certainty is required for the risk assessment. An ad hoc Expert Group on Avian Acute Oral Toxicity was established in 1994. Face-to-face expert meetings were held in 2000, 2004, 2005, 2006, and 2007.



OECD work on avian dietary toxicity and outcome

The Pensacola Workshop recommended a fundamental revision of the OECD Test Guideline 205: Avian Dietary Toxicity Test (originally adopted in 1984) to increase the exposure period from five days to 21 days. The other recommended revisions to TG 205 included: (1) an initial range-finding test, triggered by the LD50 and data on bioaccumulation potential, and (2) a definitive LC50 test, triggered by the range-finding test.

An ad hoc Expert Group on Avian Dietary Toxicity was established in 1994. In 2004, the WNT Meeting agreed that the OECD TG 205 should also be revised to generate information useful for the assessment of bird avoidance behaviour. The Expert Group on Avian Avoidance and Dietary Toxicity Testing met in 2004, 2005 and 2006. But in 2006, the revision of TG 205 to integrate the workshop recommendations was abandoned. The Expert Group refocused its efforts on the development of methodologies on bird avoidance behaviour (see below).

OECD work on reproduction toxicity and outcome

The Pensacola Workshop recommended that the OECD Test Guideline 206: Avian Reproduction Test (originally adopted in 1984) be revised as follows: (1) the reproduction test would be a first tier test, i.e. conducted on all chemicals; (2) the test duration should be reduced from the current 20 weeks to eight weeks, i.e. two weeks’ pre-exposure and six weeks’ exposure; (3) the number and selection of test species (Japanese quail and/or Bobwhite vs. Mallard) should be clearly indicated; and (4) the breeding performance of individual breeding pairs pre-dose should be used as a covariate. The Workshop also identified the need for developing egg spraying, short-term exposure, and full breeding cycle tests for particular circumstances.

An ad hoc Expert Group on Avian Reproduction Toxicity was established in 1994. The group worked on the revision of the OECD TG 206 based on the Pensacola’s recommendations. Between 1995 and 2004 the Expert Group on Avian Reproduction Toxicity met several times at the occasion of the SETAC North-America and SETAC-Europe meetings: 1995, 1996, 1997 (April and September), 1999, 2000, and 2004.

In 2000, the WNT assigned an additional task to the group, based on a proposal from the United States: the development of a new Test Guideline for a two-generation study that should include endpoints to analyse possible endocrine disrupting effects in birds.

From 2004, the task of the Expert Group was four-fold: (i) revision of OECD TG 206 (one-generation reproduction test); (ii) development of a new Test Guideline for a two-generation study using the Japanese quail for assessing endocrine disrupter effects; (iii) drafting a Detailed Review Paper (DRP) to support the test design of an avian two-generation toxicity test; and (iv) development of an overarching framework for the avian reproduction testing strategy.

Revision of the avian one-generation test (TG 206)

Efforts to update TG 206 ended on a draft document dated 2001 and a comparison of existing testing data from Japanese and Bobwhite quail conducted in 2001-2002. The main obstacle to an agreement was the number and selection of test species. Considering that the revision of the one-generation test did not have enough support from member countries, and given the competing interest to develop a two-generation test, it was agreed to keep the TG 206 as is, and to focus efforts on the development of a two-generation test with the Japanese quail. In 2005 the WNT agreed to cease the project to revise TG 206.

Development of an avian two-generation test

A first draft Test Guideline for a two-generation study with endpoints for endocrine disrupter assessment was provided by the United States in 2001. A pre-validation exercise was conducted in 2002-2003, and a detailed literature review on avian endocrinology completed, including key issues like test species, exposure scenario and selection of endocrine-related endpoints, in support of the future Test Guideline.

The project remained on the work plan until 2013. Despite efforts to demonstrate the robustness and reproducibility of the test, given the logistical complexity, the numerous sources of possible failure of the test, and the large animal number used in the test to achieve statistical power, countries decided to stop the development of a harmonised OECD Test Guideline in 2014.

DRP on avian two-generation toxicity testing

The first draft Detailed Review Paper for Avian Two-generation Toxicity Test was available in 2003 for review and comment. The final document, consolidated using input provided during the commenting rounds, was published in 2007 (DRP on Avian two-generation testing, Series on Testing and Assessment No. 74).

Avian reproduction testing strategy

In 2004 the Expert Group on Avian Reproduction Toxicity started to develop an overarching framework for the avian reproduction testing strategy. However, there was no support from the WNT and no lead country. The activity was stopped in 2005.

OECD work on bird avoidance testing and outcome

The Pensacola Workshop recommended to develop principles of bird avoidance testing. Two types of tests were identified for chemicals classified as high-risk: (1) a critical test of avoidance under realistic conditions, and (2) an optional one, more standardised screening test which could precede the critical test.

An ad hoc Expert Group on Avian Avoidance Behaviour was established in 1994. The group started to work on the development of two types of tests: avian repellency and avian avoidance behaviour. In 2004 the avoidance group merged with the ad hoc Expert Group on Avian Dietary Toxicity. The Expert Group on Avian Avoidance and Dietary Toxicity Testing met three times in 2004, 2005 and 2006.

In 2010 a first draft Guidance Document for Avoidance Testing in Birds was available for review; many comments and critics were raised. This document was then discussed at an expert meeting held in October 2014 in York, United Kingdom. The Expert Group did not reach agreement on the draft guidance document; some experts considered the guidance too prescriptive with respect to the selection and conduct of avoidance tests for birds, test design and sequence of testing.

The Expert Group recognised however that risk assessors need aid for interpreting existing data generated in avian avoidance studies with captive birds, suggested developing a new document – Guidance for design and interpretation of studies on avoidance of pesticides by birds – that will: (i) provide guidance on the conduct of studies, interpretation of data, and how to use avoidance studies for birds; and (ii) identify the related key issues. However, given the time and efforts required and the difficulty to reach consensus, the WNT agreed in 2015 that the activity on the development of avian avoidance tests should be stopped. In order not to lose the knowledge developed during these twenty years on avian avoidance behaviour studies and make it available to risk assessors who evaluate existing data, it was agreed to make the 2010 draft avoidance document publically available, together with a list of unresolved issues.

The 2010 version of the avoidance document, now entitled: Report on activities of the Expert Group on Avian Avoidance Tests, 1994-2014, is made publicly available, but does not represent a consensus guidance document.

Unresolved issues relating to avian avoidance tests

The following key points for the design and interpretation of studies on avoidance of pesticides by birds are unresolved and demonstrate that the need for further OECD work on this topic is not warranted:

  • The purpose of a consensus document will be to provide help in deciding whether to do an avian avoidance study (for notifiers and regulators) and how to design studies and help risk assessors in evaluating the contribution of available data to an overall weight-of-evidence assessment. It will not be a prescriptive document because studies need to be designed case by case.
  • The role of avian avoidance testing in the risk assessment of pesticides: although testing for avoidance by birds is no longer a regulatory requirement for pesticides registration in OECD countries, notifiers often submit avian avoidance studies to the competent authorities.
  • Is there a need for guidance on the conduct, use and interpretation of avoidance studies in birds? It is generally recognised that due to the complexity of factors affecting avoidance, interpreting data on avoidance from studies with captive birds, and assessing its implication for risk in the field, is difficult and uncertain.
  • Use of avoidance tests: In general avoidance studies are considered as part of a wider package of information. In some cases, avoidance studies may be considered stand-alone information, supported by feeding rates.
  • Acute effects vs. chronic effects: explanation of reasons why avian avoidance behaviour data will not address reproduction/ chronic effects.
  • Available tests: description of available avoidance tests.
  • Delivery: seed types, seed size, treated food types, plants, etc. The mode of delivery should be acceptable to species used in the study. It should be appropriate to the intended crop formulation (seed of interest, coated/uncoated, moisture content, etc.). A matrix/ flowchart might be useful to provide answers.
  • Species: species include: (i) ecological relevance; (ii) sensitivity; (iii) feeding rate; and (iv) husking.
  • Loading: loading involves (i) relation to field levels including nominal and after decay; and (ii) actual level and potential change during experiment (measured, expected).
  • Feeding rate: feeding rate in laboratory involves: (i) presentation of food (amount, spread, in pots, etc.); (ii) preparation of the animals (hunger, feeding time, training); (iii) group size, flocking effects; (iv) body condition relevant for time of year; and (v) prior experience of untreated food of same type.
  • Number of choices available: the selection of choice (for food) involves: (i) no choice; (ii) choice in time; and (iii) concurrent choice.
  • Endpoints: the endpoints to be measured are: (i) mortality; (ii) sublethal effects, types, severity, how and when recorded; (iii) food consumption (g/day or g/bird for defined periods of time) and feeding rate (peck rate, etc.); (iv) other endpoints e.g. physiological, pathological; and (v) early termination and how to interpret this.
  • Statistical issues: the statistics-related issues are: (i) design; (ii) controls (sequential, concurrent, positive control); and (iii) analysis: confidence intervals, significance, validity of assumptions of statistical analysis.
  • Regurgitation: is regurgitation recorded, and what are the implications for interpretation and extrapolation of results?
  • Confounding factors: confounding factors influencing response can be: (i) vapour pressure (pots, etc.), indoor, ventilation, outdoors; and (ii) other routes of exposure in laboratory or field.
  • Test animals: there are a number of elements associated to the test animals: (i) source, wild versus captive, acclimatisation period; (ii) strain and homogeneity of laboratory-bred animals; (iii) age, sex, young birds (more sensitive?), birds metabolically mature; and (iv) health status, infection, etc.
  • Test environment: test environment involves: (i) indoors, outdoors; (ii) temperature, other conditions; (iii) quality of testing conditions including precautions for disturbance by observers, experimental/technical activities etc., noise levels and type.

Further Reading


Related Documents