Appendix A. Diagnostic analyses of the multilevel confirmatory path analysis in Fig. 2.
TABLE A1. D-separation tests (d-sep) of conditional independence implied in Fig. 2. Parameter numbers are sequential (from left to right) in Fig. 2 as follows: (1) bay laurel infected prevalence, (2) redwood infected prevalence, (3) tanoak infected prevalence, (4) tanoak per capita mortality rate, (5) bay laurel per capita mortality rate, and (6) redwood per capita mortality rate.
Basis set |
t value |
Null probability |
5_||_2{4} |
1.718 |
0.0879 |
5_||_3{4} |
-1.532 |
0.128 |
6_||_1{4} |
1.518 |
0.131 |
6_||_2{4} |
-0.422 |
0.674 |
6_||_3{4} |
-0.787 |
0.432 |
C = 15.51, P = 0.21, df = 12
TABLE A2. Moran’s I values for tests of residual spatial autocorrelation for response variables in Fig. 2. Moran’s I values were calculated with calculated distance-weighted residuals according to Venables and Ripley (2002).
Response variable |
z value |
P value |
Moran’s I |
Tanoak mortality |
-2.031 |
0.042 |
-0.053 |
Bay laurel mortality |
-0.163 |
0.87 |
-0.001 |
Redwood mortality |
-0.184 |
0.854 |
-0.009 |
LITERATURE CITED
Venables, W. N., and B. D. Ripley. 2002. Modern Applied Statistics with S. Fourth Edition. Springer, New York, New York, USA. ISBN 0-387-95457-0