Table 5.
Behavior group vs. elephant (directed toward guides only).
Sally | Nandi | Thandi | Keisha | Mashudu | Thato | Shungu | Totalc | |
---|---|---|---|---|---|---|---|---|
Trunk out | 37 (32a) | 44 (39) | 43 (27) | 54 (33) | 32 (33) | 18 (24) | 67 (36) | 295 (33) |
Trunk to human | 60 (52) | 35 (31) | 54 (34) | 65 (40) | 32 (33) | 23 (30) | 45 (24) | 314 (35) |
Trunk to object | 9 (8) | 9 (8) | 14 (9) | 2 (1) | 3 (3) | 9 (12) | 8 (4) | 54 (6) |
Seeking | 3 (3) | 18 (16) | 34 (21) | 26 (16) | 27 (28) | 17 (22) | 57 (31) | 182 (20) |
Prolonged contact | 0 (0) | 0 (0) | 5 (3) | 3 (2) | 1 (1) | 6 (8) | 3 (2) | 18 (2) |
Other | 6 (5) | 6 (5) | 9 (6) | 12 (7) | 1 (1) | 3 (4) | 6 (3) | 43 (5) |
Totalb | 115 | 112 | 159 | 162 | 96 | 76 | 186 | 906 |
aNumbers in bolded parentheses are percentages of each behavior group out of total interactions that the elephant exhibited toward guides, not always 100% due to rounding.
bGeometric mean across behavior groups directed toward guides differs significantly among elephants (p = 0.0008). However, the main effect of elephant is hard to interpret given that it involves non-trivial averaging across behavior groups.
cTotals directed toward guides differ significantly across all behaviors (p < 0.0001) or across four most common behaviors (trunk out through seeking, p < 0.0001).