Table 1 Criteria for associating insecticide resistance in malaria vectors with agricultural insecticide use

Timeline: resistance in mosquitoes to a particular insecticide pre-dates public health or personal use of that insecticide at that site

Mosquito populations from Khartoum express 60–80 % mortality when exposed to the carbamate beniocarb, but carbamates are not approved for public health use in Khartoum, and are the most commonly used class of pesticides for agriculture [28]

Correlation in time: vector resistance levels rise and fall with the agricultural spraying schedule at a sampling site over time

Mosquitoes collected from cotton farms in Cameroon during the cotton spraying season showed a 1.6-fold increase in median knockdown time relative to mosquitoes collected from the same sites just prior to the spraying [18]

Correlation in space: vector populations from agricultural sites are more resistant to insecticides than their counterparts at non-agricultural sites in the same region

In Mali, the frequency of the kdr resistance gene was highest in mosquitoes sampled from a cotton-growing site, compared to sites with just personal or public health insecticide use history [30]

Correlation in quantity: vector resistance levels in a population increase with increasing quantity of agricultural insecticides in that region

Mosquito populations collected from two vegetable farms in Benin showed mortality to insecticides that varied inversely with concentrations of insecticides in the soil and water from these sites [25]

Chemical correlation: the insecticides to which a mosquito population is resistant correspond to the insecticides used for agricultural purposes in that area

At a site in Cameroon where pyrethroids make up a higher proportion of pesticides used in agriculture, mosquitoes showed higher pyrethroid resistance compared to mosquitoes from another agricultural site with less reliance on pyrethroids [26]