Table 2. Associations between the most frequent species (relative frequency >5%) from the three transects sorted by decreasing rank of occurrence and tested using Chi-square tests (1 d.f., Yates' correction).
Relative frequency | Species | 1 | 2 | 3 | 4 | 5 | ||
Coastal | 1 | 70% | Technomyrmex albipes | |||||
transect | 2 | 15% | Brachymyrmex cordemoyi | X | ||||
3 | 9% | Tetraponera longula | X | 0 | ||||
4 | 6% | Camponotus sp.4 | X | + | 0 | |||
Inland 1 | 1 | 21% | Crematogaster ranavalonae | |||||
transect | 2 | 19% | Crematogaster kelleri | 0 | ||||
3 | 12% | Cremato. madagascariensis | 0 | + | ||||
4 | 6% | Crematogaster rasoherinae | 0 | 0 | 0 | |||
5 | 6% | Camponotus sp.1 | 0 | 0 | 0 | 0 | ||
6 | 5% | Crematogaster sp.1 | 0 | 0 | 0 | 0 | 0 | |
Inland 2 | 1 | 20% | Crematogaster ranavalonae | |||||
transect | 2 | 16% | Technomyrmex albipes | [−] | ||||
3 | 15% | Brachymyrmex cordemoyi | [−] | + | ||||
4 | 10% | Cremato. madagascariensis | 0 | 0 | 0 | |||
5 | 5% | Tapinoma subtile | 0 | 0 | 0 | 0 |
Symbols indicate the nature of the association.: +: positive, [−] negative, 0: not significant. X indicates that the test is meaningless since Technomyrmex albipes is present in every sample.