The TFI was developed in Denmark 2008 and replaced the simple measurements of the applied pesticide volume as indicator. It has been in use since in several countries worldwide as national or regional indicator or as part of projects. The TFI is calculated by the theoretical number of pesticide treatments per hectare, based on standard dose rates of active ingredients, and the amount of pesticides sold yearly. An advantage of the TFI is that the indicator can be aggregated into a single value, e.g. a TFI of 1 is equivalent to one full dose applied on a certain agricultural area (Gravesen, 2003).
The TFI was e .g. used to calculate differences in the efficiency of a range of cropping systems concerning fungicide or herbicide use reduction (Andert et al., 2016; Bürger and Gerowitt, 2009), to compare organic and conventional cropping systems in terms of earthworm abundance in the topsoil (Pelosi et al., 2013) or to estimate the intensity of pesticide and fertiliser application on tomato plants in Benin (Perrin et al., 2015). Some working groups have used the TFI specifically to track herbicide usage in different cropping systems (Herbicide treatment frequency index). In general, the TFI can be used for any type of pesticide (fungicide, insecticide, herbicide etc.).
The TFI is also applied to fulfil the targets layed down in the German national action plan (NAP) to meet the requirements of Regulation (EC) No 1185/2009 concerning statistics on plant protection products (Freier et al., 2015).
One constraint of the TFI is that progress towards products with lower toxicity cannot be covered by the indicator: the TFI does not account for the chemical or toxic properties of some specific substances of the pesticide. Additionally, ecological effects or damages cannot directly be assigned to pesticide applications, since interactions and intermediate steps often have a major influence on pesticide environmental behaviour (Ongley, 1996). Consequently, a reduction in treatment frequency is not sufficient to reach conclusions regarding trends in environmental and health risks, even though a correlation is commonly assumed (Barzman and Dachbrodt‐Saaydeh, 2011).
As the TFI is not related to the active substances used, no relation can be established to elevated concentrations of single substances in raw water.