Appendix B. A table showing results for leaf and consumption attributes in paired feeding choice tests with Chrysopsyche imparilis caterpillars.
TABLE B1. Mean values ( ± 1 SE) of leaf and consumption attributes determined for 10 C. fragrans plants used in paired (P1a,b–P5a,b) feeding-choice tests with C. imparilis caterpillars. For inter-plant variation of leaf attributes, results of Kruskal-Wallis (H) test are shown. For variation within plant-pairs, results of paired t tests or Wilcoxon’s signed-ranks test (indicated by wsr), or of paired-sample randomization tests (for PLD and LWC) are shown. Number of caterpillars tested: P1 = 15, P2 = 15, P3 = 20, P4 = 29, P5 = 18. ns, nonsignificant; * P < 0.05; ** P < 0.01; *** P < 0.001.
Sample |
Pre-experimental leaf damage |
Area complete leaf |
Leaf water content |
Specific leaf area |
Consumed leaf fresh mass |
||||
P1a |
0.6 (0.2) |
30.7 (1.5) |
69.7 (0.9) |
11.7 (0.5) |
28 (13) |
||||
P1b |
7.2 (1.5) |
40.1 (1.3) |
68.1 (0.6) |
10.3 (0.3) |
72 (19) |
||||
P (1a vs. 1b) |
*** |
*** |
ns |
ns |
ns |
||||
P2a |
- |
- |
66.1 (0.7) |
9.7 (0.2) |
85 (16) |
||||
P2b |
- |
- |
67.3 (0.9) |
9.6 (0.4) |
97 (22) |
||||
P (2a vs. 2b) |
- |
- |
ns |
ns |
ns |
||||
P3a |
2.8 (0.7) |
39.6 (1.3) |
60.3 (0.5) |
8.9 (0.3) |
291 (57) |
||||
P3b |
8.8 (1.6) |
57.2 (2.1) |
61.1 (0.7) |
8.6 (0.2) |
317 (48) |
||||
P (3a vs. 3b) |
** |
*** |
ns |
ns |
ns |
||||
P4a |
5.2 (0.7) |
60.3 (2.4) |
54.8 (0.6) |
8.5 (0.1) |
269 (50) |
||||
P4b |
11.4 (1.6) |
59.4 (2.8) |
54 (0.5) |
8.1 (0.1) |
349 (69) |
||||
P (4a vs. 4b) |
** |
ns |
* |
* |
ns |
||||
P5a |
2.7 (0.7) |
51.1 (2.4) |
57.1 (0.2) |
6.4 (0.2) |
469 (71) |
||||
P5b |
14.2 (1.8) |
62.5 (4.1) |
57.2 (0.2) |
7.1 (0.4) |
324 (86) |
||||
P (5a vs. 5b) |
*** |
* |
ns |
ns ( wsr) |
ns |
||||
Inter-plant comparison: |
|
|
|
|
|||||
H ( df): P |
72.5 (7) *** |
87.3 (7) *** |
156.1 (9) *** |
117.1 (9) *** |
- - |
||||
Explanatory information Table B1: Comparisons of leaf characteristics of experimental plant pairs were conducted using paired t tests when parametric assumptions were met and Wilcoxon’s signed ranks tests when these assumptions were not fulfilled. For variables expressed as percentage, paired-sample randomization tests were performed with 10,000 iterations (Windows version of SPSS 11.0). Between-plant comparisons of single leaf characters were conducted using the Kruskal-Wallis test, since parametric assumptions were not met by these data even after application of transformation procedures ( Sokal and Rohlf 1995, Zar 1999).
Leaves were randomly (by blindly pointing to the plant with a stick and picking the leaf first encountered) collected from the five pairs of experimental plants. The paired plants were chosen to represent, judging by appearances, one pair member with high and one with low pre-experimental herbivory. One leaf of each pair member was offered to single C. imparilis caterpillars for 16 hours (from 1600 till 0800, to ensure day and night-time feeding and ad libitum fresh leaf material to feed). The caterpillars were kept in plastic terraria (18 × 11 × 14 cm; provided with moistened filter paper) and represented all larval stadia except first instars. Size of paired leaves was either standardized to square leaf pieces (9 cm²) or was aligned by cutting the leaves and providing comparably sized apical parts to the caterpillars. If leaf area was not standardized, it was determined before the experiment following the procedure described above for leaf area determination. Leaf a rea was also determined for leaf- remainings after the feeding experiment. From differences in leaf area before and after the experiment, consumed leaf area was computed. Leaf specific fresh and dry mass (mg fresh or dry mass per area) was determined to derive consumed leaf mass (fresh and dry) from consumed leaf area.
LITERATURE CITED
Sokal, R. R., and F. Rohlf. 1995. Biometry. Freeman and Company, San Francisco, USA.
Zar, J. H. 1999. Biostatistical analysis. Prentice-Hall, Upper Saddle River, New Jersey, USA.