Fig. 2 Average proportion of seed species consumed in N. minimusand T. striatus feeding trials. There is a significant difference in proportion of consumed A. saccharum seeds (t23 = -4.11, P < 0.001) andA. rubrum seeds (t23 = 6.82, P< 0.001). There was not a significant difference in proportion of consumed A. balsamea seeds (t23 = 0.37,P = 0.71) or P. pensylvanica seeds (t23 = -0.75, P = 0.46). The difference in proportion consumed of C. cornuta seeds approach significance (t23 = -2.41, P = 0.02).
T-tests compared the proportion consumed of each seed species betweenT. striatus and N. minimus (Figure 2). There was a significant difference in proportion consumed of A. rubrum seeds between T. striatus (0.04 ± 0.01) and N. minimus (0.65 ± 0.06; One-sample t test: t 23 = 6.82,P < 0.001). There was also a significant difference in proportion of consumed A. saccharum seeds between T. striatus (0.34 ± 0.02) and N. minimus (0.06 ± 0.01; One-samplet test: t 23= -4.11, P < 0.001). There was not a significant difference in proportion of consumedA. balsamea seeds between T. striatus (0.09 ± 0.01) andN. minimus (0.12 ± 0.01; One-sample t test:t 23= 0.37, P = 0.71). There was also not a significant difference in proportion of P. pensylvanica seeds consumed between T. striatus (0.18 ± 0.01) and N. minimus(0.11 ± 0.01; One-sample t test: t 23 = -0.75, P = 0.46). The difference in proportion consumed ofC. cornuta seeds between T. striatus (0.34 ± 0.02) andN. minimus (0.06 ± 0.01) approach significance (One-samplet test: t 23 = -2.41, P = 0.02) after a Bonferroni Sequential Adjustment (Rice, 1989).