Table 3. Generalized linear models with a Poisson distribution predicting the number of (a) elk and (b) deer on a snow transect. Variable     β     SE     t-value     P (a) Elk Snow depth -0.038 0.007 -5.86 < 0.001 Number of sample days 0.362 0.023 16.09 < 0.001 Node, reference = Signal Corridor -0.104 0.121 -0.86 0.390 Year, reference = Year 1 Year 2 0.267 0.110 2.43 0.015 Year 3 1.074 0.123 8.73 < 0.001 Node × Year Corridor × Year 2 0.113 0.151 0.75 0.455 Corridor × Year 3 -0.601 0.163 -3.69 < 0.001 (b) Deer Number of sample days 0.329 0.023 14.62 < 0.001 Node, reference = Signal Mountain -4.748 0.580 -8.19 < 0.001 Corridor -5.846 1.001 -5.84 < 0.001 Year, Reference = Year 1 Year 2 -0.460 0.084 -5.46 < 0.001 Year 3 0.105 0.097 1.09 0.278 Node × Year Mountain × Year 2 2.859 0.607 4.71 < 0.001 Corridor × Year 2 4.344 1.013 4.29 < 0.001 Mountain × Year 3 3.721 0.595 6.26 < 0.001 Corridor × Year 3 4.570 1.012 4.52 < 0.001