Does size really matter? Effects of experimental unit size and relative humidity on the life-history parameters of...

Given the relative novelty of the research in insects as food and feed, experiment protocols often differ among each other, rendering comparison difficult. This experiment aims at assessing the effects of three rearing scales, or the size of the experimental unit, the trays in which larvae are reared. Experimental unit sizes tested were: small (S, 13*13*5 cm), medium (M, 30*20*11 cm), and large (L, 57*36*10 cm), at two relative humidity (RH) conditions (60 and 70%), for a total of 6 treatments. The parameters observed were the growth, bioconversion capacity and nutritional profile of Tenebrio molitor larvae. Each treatment was replicated 5 times. All other parameters were kept constant: genetic origin of the larvae, feed (wheat bran), feeding rate (0.26 g/larva), areal density (4 larvae/cm²), volumetric density (1 larva/cm³) and water source provision (agar, 20 g/l). Trial started with 3-week-old larvae and sampling occurred every 7 days until 5% of pupae were observed in a replicate. Data were analysed using SPSS software (P ≤ 0.05). The different scales did not have a significant effect on the survival of the larvae and the development time at either of the RH tested (P > 0.05). The weight of the larvae was lower at the S scale at both RH (P < 0.001) while the M and L scales were comparable (P > 0.05) under conditions up until larvae were 49 days old. At the end of the trial, differences could only be observed in 60% RH between S and M scales (P < 0.05). Bioconversion efficiency was higher at S scale, with a lower feed conversion ratio (P < 0.001) and a higher efficiency of conversion of ingested feed (P < 0.001). The nutritional composition of the larvae was affected by the scale of the experiment, with the M scale having higher CP and fat content than the S scale (P < 0.001). The temperature of the substrate was affected by the scale of the experiment. Notably, at 60% RH conditions, the temperature of the substrate was lower at S scale than M and L at all times except at the first and last measurements (P < 0.05). In conclusion, the size of the experimental unit has effects on the outcome of experiment. Bioconversion was more efficient in small scale. However, M scale allowed for higher larval growth. Based on these findings, the size of the experimental unit is an important parameter that should be taken into consideration when planning experimental protocols and comparing experiments.