Table 2 In-house studies used to derive input data for modelling by Anthe et al. [1]

Stroking tests for spot-ons

16–20 dogs treated according to the registered label. Four samples taken between 10 min and 24 h post-application by stroking the dogs with highly absorbent cotton gloves

 < 10% of the active substance applied was measured on day 1 (worst case)

Considering an adequate margin of safety abrasion to resting places was estimated 20%

F_abr = 0.2 is used in Scenario 1 “Washing of pet bedding”

Stroking tests for collars

8 dogs and 8 cats were treated with a collar. The animals were stroked with absorbent cotton gloves on 11 days within a period of 1 to 238 days after applying the collars. The calculated amount of each interval was summed up to cover the 8-month period. Assuming stroking 4 times per day the results were multiplied by 4

 < 0.6% of the total active substance content was available for abrasion during the 8-month treatment period

An amount of 1% is used for the calculations (F_abr = 0.01) in Scenario 1 “Washing of pet bedding”

Washing and vacuuming of cat bedding

2 groups of six cats, 6 h post treatment, received in their cage a piece of cloth-covered plywood which was left in the cage for 4 consecutive days. The cloth was then cut in half. One piece was analysed immediately, remaining pieces of cloth from group 1 were machine washed and those of group 2 were vacuumed. Then extraction and analysis of the imidacloprid present was performed

Imidacloprid concentration measured in the cloths were:

- < LOQ after washing

- Decreased by 50% after vacuuming

A fraction released due to washing was set to F_washing of 0.5 taken also into account the pet owners survey that showed vacuuming is done normally in between each washing (relevant for Scenario 1 “Washing of pet bedding”). Therefore, by the time of washing a proportion of the active substance is no longer available for release [1]

Immersion test with collars

Study 1:

Eight collars for large dogs were immersed in plastic tubes filled with 180 L water for a period of 5 min

Study 2:

Eight dogs were treated with a collar. 83 and 90 days after treatment each dog was immersed in tepid water for a period of 5 min. The dog’s head was wetted three times during the immersion procedure

Imidacloprid concentration measured in water samples were below the LOQ (5 µg/L)

Less than 0.9% of the collar inventory was detected in water whether or not dogs wore the collar during immersion. The fraction of Imidacloprid released in these tests was estimated at 1.64% (90th percentile)

The fraction of imidacloprid released from treated pelt due to rain was set to 2% for collars and the input parameters F_rain 0.02 is used in the scenario 3 “Walking dogs in rain”

Based on the immersion tests with the collars and a safety factor > 10 taking into consideration the frequency of shampooing (Table S6 in Additional File 1), an average fraction F_washing of 0.2 was calculated for the scenario 2 “Washing/bathing of dogs” (see Eq. 12 and Table 1 [1])