Classification of samples to species and hybrid categories
Across both transects, 4,026 moths were classified as winter moth (n = 4,026 with strong support, and n = 1 with moderate support), 5,434 as Bruce spanworm (n = 5,432 with strong support,n = 1 with moderate support, and n = 1 with weak support), and 523 as one of the four hybrid-categories with the majority (n= 281) being classified as F1 hybrids (n = 254 with strong support, n = 23 with moderate support, and n = 4 with weak support), 155 individuals as F2 hybrids (n = 125 with strong support, n = 23 with moderate support, and n =7 with weak support), 86 individuals as winter moth-backcrosses (n = 60 with strong support, n = 18 with moderate support, and n = 8 with weak support) and 1 individual as a Bruce spanworm-backcross (with moderate support). The number of moths classified to each hybrid category along the Massachusetts transect is presented in Table 1 and along the Connecticut transect in Table 2. The proportions of individuals classified as winter moth, Bruce spanworm, or to one of the four hybrid classes are presented in Figures 2 and 3a for the Massachusetts and Connecticut transects, respectively. The numbers of individuals assigned to each hybrid classification are presented in Figures 4 and 3b for the Massachusetts and Connecticut transects, respectively.
On average, we observed a hybridization rate of 5.67% ± 0.62% across years along the Massachusetts transect. We found no significant difference in hybridization rates among years along the Massachusetts transect (P=0.332, F=1.139, df=11), however, we did observe a significant difference in mean hybridization rates among traps for all years combined (P=0.0024, F=1.98, df=32), though none of the pairwise comparisons among traps were significant based on Tukey’s HSD test (Figure 5). Along the Connecticut transect, we observed a hybridization rate of 5.97% ± 0.91% across years, with significant differences in hybridization rates between years (P=0.0285, F=3.817, df=2), with the hybridization rates between the 2017 and 2018 transect surveys being significantly different based on Tukey’s HSD test (adjusted P= 0.027). As per the Massachusetts transect, significant differences were also observed among traps for all years combined (P=0.0082, F=2.569, df=18), though none of the pairwise comparisons were significant based on Tukey’s HSD test (Figure 6).