VIII SIMPOSIO INTERNAZIONALE ICPBR

HAZARDS OF PESTICIDES TO BEES

Bologna, 4-6 Settembre 2002

Metodologie dei test e studi sugli effetti dei pesticidi sulle api

Introduction of indices for the evaluation of tent tests and field tests with honeybees

Hans-Werner Schmidt, Dietrich Brasse, Christoph Künast, Werner Mühlen, Werner von der Ohe, Ingo Tornier and Klaus Wallner

German working group "Bee protection". E-mail: hans-werner.schmidt@bayercropscience.com

According to the EPPO-guideline 170 different evaluations are requested in tent tests and field tests with honeybees. Many data are generated as counted or measured values, e.g. for mortality, foraging activity and brood development. Although they can be presented in tables or in graphs, these absolute figures are open for any interpretation.

It is proposed to calculate averages of the respective parameters for data before application and for data after application for all bee colonies in one test unit. This concentrates the data and allows the comparison of the state after application with the state before application. A simple mathematical division converts the absolute data into relative data, e.g. for mortality:

QM = dead bees per day after application / dead bees per day before application

The same calculation can be done for the foraging activity:

QF = bees per square and per evaluation after application / bees per square and per evaluation before application

The calculation for the brood is:

QB = size of the broodnest on day "y" after application / size of the broodnest before application

The use of such indices offers the possibility to adopt thresholds. When QM is >2, the mortality after application is more than twofold of the normal mortality. The reason for this should be enquired. When QF is <0,5, the foraging activity after application has decreased below half of the normal intensity. And again it should be enquired, whether the applied product had caused this. For a normal brood development QB should be >1, that means the broodnest expands during the test. If QB deviates very much from KB (which is the same quotient for the untreated unit), the reason for this should be enquired. We consider indices between 0,5 and 2 as a tolerable range. Only indices outside this limit require the investigation of the reason.

Furthermore the index Q of the treated plot can be divided by the same index K of the untreated plot. If this clearing quotient deviates very much from 1, the reason for the different behaviour of the honeybees should be enquired.

The use of the proposed indices, representing relative data instead of absolute ones, could be a helpful tool for the interpretation of the obtained data in tent tests and field tests with honeybees.