4.13. Control mortality
If in vitro larval rearing methods are supposed to become a commonly accepted method to test plant protection products, the maximum tolerated control mortality must be considered for validation. Techniques have improved over the years, and control mortality has progressively reached low levels. Nonetheless, control mortality has to be reported regularly, to demonstrate successful trials. Fukuda and Sakagami (1968) found in a normal colony setting, that 85 % of adult workers emerge from the eggs laid by a queen, with mortality occurring in all premature stages. This data suggest a natural mortality of about 15 % during larval and pupal development. Therefore, in control samples of in vitro-reared larvae, total mortality until late pupal development (~ day 14 after grafting of first instar larvae) should be lower or equal to 15 %. This is especially important for the assessment of the median lethal dose (LD50) or the median lethal concentration (LC50), while 20 % control mortality can be accepted for the assessment of a NOAEL (no observable adverse effect level) or a NOAEC (no observable adverse effect concentration). In case of higher mortality in control samples, the replicate is invalidated or convincing reasons, like adverse weather conditions before grafting of larvae or late season, must be given. However, when in vitro rearing of larvae is just one method among others to answer a scientific question, we assume a control mortality of 5 % from grafting to day 7 to be very good, and 10 % tolerable; from grafting to day 22, 20 % is very good, while 25 % can be tolerated. Care should be taken in presenting these parameters unambiguously and the way they were calculated. These levels should be achieved without exclusion of the grafting effect (see 5.6. Exclusion of grafting effect). Several research institutes have already accomplished control mortalities below these suggested ones (Aupinel et al., 2005; Aupinel et al., 2009; Brodschneider et al., 2009; Hendriksma et al., 2011a; Kaftanoglu et al., 2010; 2011).