3. Results
3.1. Starch digestibility
Tilapia showed higher starch digestibility than jundiá for all plant sources evaluated, except for wheat bran, which was similar to that found for jundiá (Fig. 1). Starch digestibility of the different plant sources was similar for tilapia and averaged 99.74 ± 0.35% for wheat bran, 94.67 ± 3.94% for cassava residue, 92.02 ± 5.86% for ground corn, and 96.74 ± 4.58% for broken rice. However, for jundiá, the highest digestibility was registered for wheat bran (90.61 ± 2.65%) and the lowest for broken rice (55.87 ± 4.05%). The starch digestibilities of cassava residue (78.26 ± 7.05%) and ground corn (68.48 ± 14.4%) were intermediate. The significant interaction between plant sources and fish species for starch ADC indicates that starch digestibility is influ- enced by the plant sources for jundiá, but not for tilapia (Fig. 1). For jundiá, starch or amylose content presented significant negative corre- lation with starch ADC, whereas amylopectin or dietary fiber showed positive significant correlation with starch ADC (Table 3). For tilapia, no significant correlation was found between any variable from test ingredients and starch ADC. Amylose and amylopectin data are comple- mentary. As such, their correlation coefficients (absolute values) and P-values are equal (Table 3).