Appendix D. Results of hypothesis testing: tables of F statistics, P values, and aditional analyses described in the text.
Included here are the following sections: (1) Table D1. Comparisons of levels of endophyte infection for native vs. nonnative species, and invasive vs. noninvasive introduced species, (2) Table D2. Comparisons of the prevalence of disease symptoms for native vs. nonnative species, and invasive vs. noninvasive introduced species, (3) Table D3. Comparisons of the prevalence of damage by herbivores for native vs. nonnative species, and invasive vs. noninvasive introduced species, (4) Table D4. Means and Standard Deviations for severity of leaf damage from disease and herbivory over multiple censuses in Years 1 and 2, (5) Table D5. Statistics for disease and herbivory severity, (6) Table D6. Results of logistic regression models for each species, predicting mortality after 2 weeks from prevalence of herbivory, prevalence of disease, and number of leaves, and (7) Why we did not use severity instead of prevalence for logistic regressions.
TABLE D1. Comparisons of levels of endophyte infection for native vs. nonnative species, and invasive vs. noninvasive introduced species. Nonparametric tests and tests done on transformed data produced results similar to those presented here (usually with less significant P values). There was one exception: analysis of arcsine-square root transformed data in Year 1 suggested that invasive, introduced clovers experienced significantly higher infection than noninvasive, introduced clovers (t = 2.7, P = 0.043). Years were not combined into a single analysis because species are not independent among years. Repeated-measures analysis was not used because the identity of species used varied slightly from year to year. No adjustments were made for multiple comparisons because differences were already nonsignificant.
Native vs. introduced |
Invasive vs. noninvasive |
|||||
t |
df |
P |
t |
df |
P |
|
Year 1 |
0.84 |
14 |
0.42 |
2.22 |
5 |
0.08 |
Year 2 |
1.89 |
15 |
0.08 |
1.53 |
6 |
0.18 |
Year 3 |
1.11 |
13 |
0.29 |
1.25 |
5 |
0.27 |
Year 4 |
0.80 |
15 |
0.43 |
1.40 |
6 |
0.21 |
TABLE D2. Comparisons of disease prevalence for native vs. nonnative species, and invasive vs. noninvasive introduced species. In Years 3 and 4, prevalence (proportion of population showing symptomatic damage) was assessed only at harvest, but in Years 1 and 2, we used an integrated measure of cumulative incidence (area under the disease progress curve, AUDPC). We performed t tests assuming unequal variances; because of this and because different groups of species were used in different years, degrees of freedom differ among tests.
Native vs. introduced |
Invasive vs. noninvasive |
|||||
t |
df |
P |
t |
df |
P |
|
Year 1 |
1.16 |
8 |
0.28 |
6.85 |
4 |
0.002 |
Year 2 |
0.29 |
9 |
0.78 |
22.42 |
5 |
0.00001 |
Year 3 |
0.16 |
9 |
0.88 |
9.21 |
5 |
0.0002 |
Year 4 |
0.66 |
14 |
0.52 |
2.06 |
5 |
0.094 |
TABLE D3. Comparisons of the prevalence of damage by herbivores for native vs. nonnative species, and invasive vs. noninvasive introduced species. In Years 3 and 4, prevalence (proportion of population showing herbivore damage) was assessed only at harvest, but in Years 1 and 2, we used an integrated measure of cumulative incidence (area under the damage progress curve, AUDPC). We performed t tests assuming unequal variances; because of this and because different groups of species were used in different years, degrees of freedom differ among tests.
Native vs. introduced |
Invasive vs. noninvasive |
|||||
t |
df |
P |
t |
df |
P |
|
Year 1 |
3.14 |
7 |
0.016 |
1.15 |
4 |
0.31 |
Year 2 |
6.27 |
10 |
0.00009 |
2.16 |
5 |
0.083 |
Year 3 |
0.75 |
13 |
0.47 |
2.83 |
5 |
0.037 |
Year 4 |
1.57 |
11 |
0.14 |
0.19 |
1 |
0.88 |
TABLE D4. Means and standard deviations for severity of leaf damage from disease and herbivory in Years 1 and 2. Means were calculated by taking the mean of three leaves per plant, and then the means of all plants within each species. Reported are the grand means and standard deviations of those species means. Sample sizes per species and number of species per category are given in Table A2 (Appendix A).
Disease Year 1 |
||||
Census |
Native |
Introduced |
NonInvasive |
Invasive |
8-Mar |
3.09±2.4 |
5.68±2.16 |
5.01±2.2 |
7.34±0.91 |
21-Mar |
2.41±2.16 |
4.76±4.44 |
4.81±5.44 |
4.63±0.12 |
3-Apr |
1.43±1.22 |
4.63±2.53 |
4.82±2.95 |
4.16±1.73 |
17-Apr |
3.83±2.09 |
5.32±6.12 |
4.67±7.13 |
6.94±3.74 |
16-May |
2.24±1.76 |
3.88±3.24 |
2.07±1.07 |
8.4±1.16 |
Herbivory Year 1 |
||||
Census |
Native |
Introduced |
NonInvasive |
Invasive |
8-Mar |
13.8±4.34 |
8.39±6.27 |
10.58±6.18 |
2.94±0.16 |
21-Mar |
13.55±4.15 |
7.44±5.74 |
8.97±6.26 |
3.63±0.1 |
3-Apr |
11.86±2.83 |
6.54±4.29 |
7.61±4.75 |
3.85±0.54 |
17-Apr |
11.81±3.64 |
3.86±2.74 |
4.3±3.22 |
2.76±0.22 |
16-May |
12.68±4.91 |
3.63±2.68 |
4.15±3.06 |
2.31±0.97 |
Disease Year 2 |
||||
Census |
Native |
Introduced |
NonInvasive |
Invasive |
21-Mar |
0.36±0.69 |
1.05±1.32 |
0.6±0.58 |
2.41±2.38 |
24-Apr |
0.68±0.67 |
1.28±1.68 |
0.66±1.01 |
3.13±2.36 |
21-May |
2.42±2.38 |
1.77±1.8 |
1.72±1.93 |
1.92±2.01 |
Herbivory Year 2 |
||||
Census |
Native |
Introduced |
NonInvasive |
Invasive |
21-Mar |
3.28±2.52 |
2.84±3.6 |
3.44±4.03 |
1.02±0.86 |
24-Apr |
4.57±2.92 |
2.48±2.23 |
2.24±2.32 |
3.19±2.6 |
21-May |
2.79±1.87 |
2.19±3.38 |
2.62±3.86 |
0.91±0.86 |
TABLE D5. Statistics for disease and herbivory severity. Repeated-measures ANOVA was calculated for Years 1 and 2, using the asin√(proportion leaf area) affected by any disease or any herbivore, using only those leaves that were alive at the time of the census. There were five severity censuses in Year 1 and three censuses in Year 2, with seven blocks each year. The analysis was a repeated-measures analysis across the censuses, treating the mean disease per plant (from the mean of all leaves) as the experimental unit, and then nesting species within origin or invasiveness, and including a block term [Block Type Species(Type)]. For Years 3 and 4, two-factor factorial ANOVA was calculated using asin√(proportion leaf area). Separate analyses were done for Native vs. Introduced (“Type” = Origin) and Invasive vs. Noninvasive within the introduced species (“Type” = Invasiveness).
Year1 |
Native vs. introduced |
Invasive vs. noninvasive |
|||||
Test |
Damage |
Exact F |
df |
P |
Exact F |
df |
P |
Type |
Disease |
16.8736 |
1,157 |
0.0001 |
28.3536 |
1,66 |
0.0001 |
Species(Type) |
Disease |
5.4307 |
14,157 |
0.0001 |
3.8028 |
5,66 |
0.0044 |
Time |
Disease |
0.8747 |
4,154 |
0.4806 |
1.0249 |
4,63 |
0.4014 |
Block |
Disease |
1.4297 |
6,157 |
0.2065 |
0.4653 |
6,66 |
0.8315 |
Time × Type |
Disease |
1.0305 |
4,154 |
0.3934 |
2.4217 |
4,63 |
0.0574 |
Type |
Herbivory |
66.0271 |
1,157 |
0.0001 |
1.6596 |
1,66 |
0.2022 |
Species(Type) |
Herbivory |
3.9106 |
14,157 |
0.0001 |
6.5925 |
5,66 |
0.0001 |
Time |
Herbivory |
1.0570 |
4,154 |
0.3799 |
1.6131 |
4,63 |
0.182 |
Block |
Herbivory |
1.8109 |
6,157 |
0.1002 |
2.1959 |
6,66 |
0.0543 |
Time × Type |
Herbivory |
1.9098 |
4,154 |
0.1115 |
1.8340 |
4,63 |
0.1334 |
Year2 |
Native vs. introduced |
Invasive vs. noninvasive |
|||||
Test |
Damage |
Exact F |
df |
P |
Exact F |
df |
P |
Type |
Disease |
1.3611 |
1,49 |
0.249 |
5.3180 |
1,15 |
0.0358 |
Species(Type) |
Disease |
2.4633 |
16,49 |
0.008 |
1.9423 |
6,15 |
0.1391 |
Time |
Disease |
4.4290 |
2,48 |
0.0172 |
0.4842 |
2,14 |
0.6262 |
Block |
Disease |
0.9401 |
6,49 |
0.4752 |
0.4882 |
6,15 |
0.8072 |
Time × Type |
Disease |
1.6413 |
2,48 |
0.2044 |
3.0991 |
2,14 |
0.0769 |
Type |
Herbivory |
13.3410 |
1,49 |
0.0006 |
1.4344 |
1,15 |
0.2496 |
Species(Type) |
Herbivory |
1.2469 |
16,49 |
0.2689 |
0.9851 |
6,15 |
0.4691 |
Time |
Herbivory |
1.4272 |
2,48 |
0.25 |
0.6258 |
2,14 |
0.5492 |
Block |
Herbivory |
1.3115 |
6,49 |
0.2699 |
1.5328 |
6,15 |
0.2342 |
Time × Type |
Hebivory |
2.3849 |
2,48 |
0.1029 |
3.6962 |
2,14 |
0.0514 |
Table D5 continued.
Year3 |
Native vs. introduced |
Invasive vs. noninvasive |
|||||
Test |
Damage |
Exact F |
df |
P |
Exact F |
df |
P |
Type |
Disease |
1.2070 |
1,14 |
0.2751 |
20.8161 |
1,6 |
0.0001 |
Species(Type) |
Disease |
0.7565 |
14,84 |
0.7118 |
2.0284 |
6,43 |
0.0825 |
Type |
Herbivory |
1.2985 |
1,14 |
0.2577 |
3.0549 |
1,6 |
0.0876 |
Species(Type) |
Herbivory |
2.6006 |
14,84 |
0.0036 |
3.4711 |
6,43 |
0.0069 |
Year4 |
Native vs. introduced |
Invasive vs. noninvasive |
|||||
Test |
Damage |
Exact F |
df |
P |
Exact F |
df |
P |
Type |
Disease |
0.0111 |
1,15 |
0.9165 |
6.6009 |
1,6 |
0.0144 |
Species(Type) |
Disease |
1.6793 |
15,79 |
0.0724 |
1.3346 |
6,37 |
0.2667 |
Type |
Herbivory |
1.6451 |
1,15 |
0.2034 |
1.3039 |
1,6 |
0.2608 |
Species(Type) |
Herbivory |
1.0782 |
15,79 |
0.3897 |
0.9265 |
6,37 |
0.4874 |
TABLE D6. Results of logistic regression models for each species, predicting mortality after two weeks from prevalence of herbivory, prevalence of disease, and number of leaves. These model coefficients were then used to test for differences in the effects of disease or herbivory on mortality for plants of different origin (Table 3). Models fits were unstable for the following: Year 1: M. arabica, T. subterraneum; Year 2: T. glomeratum, T. willdenovii, T. wormskjoldii. Bold indicates species that are invasive in the Bodega Marine Reserve.
(P value) |
Model coefficients |
|||||||||
Origin |
Species |
Whole Model |
Herbivory |
Disease |
Leaves |
Intercept |
Herbivory |
Disease |
Leaves |
N |
Year 1 |
||||||||||
Intro |
M. polymorpha |
2.6 (0.455) |
0.1 (0.794) |
1.2 (0.277) |
0.2 (0.627) |
-3.729 |
0.105 |
-0.490 |
-0.031 |
341 |
Intro |
T. campestre |
12.9 (0.005) |
0.7 (0.41) |
0.1 (0.8) |
5.0 (0.026) |
-1.775 |
-0.338 |
-0.137 |
-0.455 |
370 |
Intro |
T. dubium |
16.5 (0.0009) |
3.8 (0.05) |
1.7 (0.19) |
6.5 (0.01) |
-1.232 |
0.572 |
0.472 |
-0.320 |
362 |
Intro |
T. glomeratum |
27.9 (0.0001) |
2.2 (0.137) |
0.6 (0.439) |
17.6 (0.0001) |
-0.423 |
-0.371 |
0.242 |
-0.920 |
299 |
Intro |
T. repens |
15.5 (0.001) |
0.2 (0.681) |
3.3 (0.0694) |
10.2 (0.001) |
-0.619 |
0.103 |
0.509 |
-0.851 |
342 |
Native |
T. barbigerum |
19.4 (0.0002) |
0.6 (0.43) |
2.6 (0.11) |
9.4 (0.002) |
0.257 |
-0.367 |
0.841 |
-1.124 |
352 |
Native |
T. bifidum |
10.7 (0.0136) |
0.1 (0.75) |
0.3 (0.6) |
5.0 (0.026) |
-1.863 |
-0.095 |
0.181 |
-0.212 |
350 |
Native |
T. fucatum |
1.1 (0.787) |
0.4 (0.55) |
0.2 (0.69) |
0.2 (0.69) |
-3.624 |
-0.255 |
-0.221 |
-0.024 |
365 |
Native |
T. gracilentum |
12.8 (0.0051) |
0.7 (0.41) |
0.0 (0.95) |
5.2 (0.02) |
-1.462 |
-0.236 |
-0.018 |
-0.315 |
290 |
Native |
T. macraei |
36.1 (0.0001) |
0.2 (0.66) |
0.4 (0.52) |
19.2 (0.0001) |
-0.150 |
-0.105 |
0.207 |
-0.619 |
256 |
Native |
T. microdon |
9.6 (0.0226) |
0.0 (0.84) |
1.0 (0.31) |
5.2 (0.02) |
-1.633 |
0.052 |
0.293 |
-0.178 |
313 |
Native |
T. microcephalum |
30.9 (0.0001) |
0.0 (0.889) |
0.3 (0.556) |
10.9 (0.001) |
-0.599 |
-0.047 |
0.261 |
-0.661 |
360 |
Native |
T. willdenovii |
10.1 (0.018) |
1.8 (0.17) |
0.3 (0.61) |
4.3 (0.038) |
-2.442 |
0.526 |
-0.175 |
-0.108 |
310 |
Native |
T. wormskjoldii |
7.4 (0.062) |
0.1 (0.81) |
0.0 (0.96) |
4.4 (0.037) |
-2.241 |
0.099 |
-0.022 |
-0.461 |
360 |
Table D6 continued.
Year 2 |
||||||||||
Intro |
M. arabica |
7.9 (0.047) |
4.8 (0.028) |
1 (0.314) |
1.7 (0.19) |
-2.100 |
0.577 |
-0.307 |
-0.079 |
232 |
Intro |
M. lupulina |
10.2 (0.017) |
5.2 (0.023) |
2.8 (0.0942) |
4.4 (0.035) |
-0.861 |
0.646 |
0.616 |
-0.459 |
238 |
Intro |
M. polymorpha |
5.9 (0.12) |
0.4 (0.521) |
0.3 (0.5929) |
2.6 (0.1063) |
-1.997 |
-0.220 |
0.175 |
-0.147 |
248 |
Intro |
T. campestre |
0.7 (0.866) |
0.0 (0.95) |
0.7 (0.41) |
0.0 (0.895) |
-1.830 |
-0.016 |
0.295 |
0.008 |
160 |
Intro |
T. dubium |
1.4 (0.702) |
0.6 (0.45) |
0.1 (0.73) |
0.6 (0.42) |
-1.939 |
0.193 |
-0.183 |
-0.052 |
215 |
Intro |
T. repens |
7.6 (0.055) |
0.0 (0.835) |
0.7 (0.39) |
6.3 (0.012) |
-0.405 |
0.051 |
0.357 |
-0.667 |
217 |
Intro |
T. subterraneum |
4.6 (0.2004) |
1.0 (0.31) |
1.4 (0.23) |
2.1 (0.145) |
-1.525 |
-0.257 |
0.413 |
-0.169 |
236 |
Native |
T. barbigerum |
0.7 (0.885) |
0.5 (0.47) |
0.1 (0.78) |
0.1 (0.8) |
-2.433 |
0.195 |
-0.151 |
-0.008 |
191 |
Native |
T. bifidum |
2.4 (0.494) |
1.3 (0.256) |
1.1 (0.286) |
1.1 (0.302) |
-1.611 |
0.283 |
0.374 |
-0.048 |
207 |
Native |
T. fucatum |
7.5 (0.058) |
3.1 (0.077) |
1.7 (0.19) |
2.3 (0.13) |
-2.227 |
0.937 |
0.584 |
-0.097 |
217 |
Native |
T. gracilentum |
9.6 (0.023) |
0.3 (0.611) |
3.3 (0.071) |
5.3 (0.021) |
-1.010 |
0.130 |
0.613 |
-0.126 |
205 |
Native |
T. macraei |
2.4 (0.498) |
0.4 (0.55) |
0.2 (0.64) |
2.1 (0.15) |
-1.695 |
0.151 |
0.195 |
-0.055 |
225 |
Native |
T. microdon |
4.6 (0.2) |
0.0 (0.89) |
1.6 (0.199) |
2.9 (0.088) |
-1.552 |
0.039 |
0.473 |
-0.066 |
224 |
Native |
T. microcephalum |
1.3 (0.73) |
0.0 (0.97) |
0.2 (0.64) |
1.0 (0.312) |
-1.765 |
-0.009 |
0.166 |
-0.064 |
176 |
Native |
T. variegatum |
2.1 (0.5559) |
0.3 (0.589) |
0.0 (0.943) |
1.5 (0.2163) |
-2.109 |
0.158 |
0.029 |
-0.043 |
211 |
Why we did not use severity instead of prevalence for logistic regressions: Data from the severity censuses in Year 1 and Year 2 could not be used for logistic regressions of mortality on percent leaf damage because there were not enough individuals and mortality events (especially for analyses split by species) in the severity census subset. In Year 1, 32 out of 222 severity census plants died over seven census dates, and in Year 2, only three plants out of the 126 severity census plants died over three census dates.