Figure 1.
Experimental setups for ethograms and choice experiments.
(A) Setup for observations on individual workers while excavating in soil and depositing pellets. Observations of individually marked workers were done both inside the clay arena, where soil pellets were excavated, and in the depositing box, where the most pellets were deposited. (B) Setup for choice experiments to test the influence of deposited pellets on the workers' decisions where to excavate. The motivation segment was connected to the subcolony to stimulate ants to excavate at that location, while soil pellets from the excavation box were placed at one side in the observation segment for the subsequent choice situation. After replacing the motivation segment with the observation segment (inset), it was observed at which side the ants started to dig first.
Figure 2.
Relationship between pellet mass (mP) and worker body mass (mW).
In each graph, the solid line represents the regression line and the dotted line a hypothetical 1∶1 ratio of load mass to carrier mass. Data points above that line correspond to carriers less heavy than their loads. (A) A. vollenweideri worker carrying a soil pellet in the laboratory (Photo: Andrew I. Bruce). (B) Pellets and carriers collected at the openings of field nests ( mP = 2.1 mW+6.5, r2 = 0.22, N = 60, P<0.001). (C) Pellets and carriers collected from laboratory colonies at 2 m distance from the excavation site ( mP = 0.8 mW+1.7, r2 = 0.24, N = 80, P<0.001). (d) Pellets and excavators collected directly at the excavation site ( mP = 0.3· mW+4.3, r2 = 0.02, N = 80, P = 0.1859).
Figure 3.
Characteristics of pellet transport as a function of time.
Transport duration (d), number of carriers involved (n) and the waiting times (w) between transport events are plotted as a function of time (t). (A) Time needed for a pellet to be sequentially carried over a distance of 2 m from the excavation site. Solid lines show the results of regression with ln-transformed d-values on both days (first day: d = 0.6·e−0.3·t, r2 = 0.45, N = 60, P<0.001; second day: d = 0.4·e−0.04·t, r2 = 0.05, N = 73, P = 0.047). (B) The number of workers involved in the transport of single pellets. Solid lines show the results of regression with ln-transformed Y-values on both days (first day: n = 6.6·e−0.1·t, r2 = 0.25, N = 60, P<0.001; second day: n = 7.9·e−0.03·t, r2 = 0.06, N = 73, P = 0.039). (C) Time a pellet remained on the floor between two transport events. Solid lines show the results of regression with ln-transformed Y-values on both days (first day: w = 3.8·e−0.2·t, r2 = 0.10, N = 225, P<0.001; second day: w = 5.5·e−0.1·t, r2 = 0.02, N = 177, NS).
Figure 4.
Distances walked (l) and speeds (v) measured in the different carriers involved in pellet transport as a function of time (t).
Black dots and solid regression lines represent excavators, grey dots and dashed regression lines short-distance carriers, and white dots and dotted regression lines long-distance carriers. (A) Distances covered (in cm) during one transport event by excavators (first day: l = 0.53·t+7.7, NS; second day: l = −0.07·t+11.4, NS), short-distance carriers (first day: l = 0.71·t+14.0, NS; second day: l = −0.58·t+32.8, NS) and long-distance carriers (first day: l = 1.98·t+142.0, NS; second day: l = 2.13·t+108.8, NS; for 1.6% confidence intervals see Table 1). (B) Walking speed in each transport event for excavators (first day: v = 0.18·t+7.9, r2 = 0.02, N = 60, NS; second day: v = −0.003·t+9.2, r2 = 0.0005, N = 73, NS), short-distance carriers (first day: v = 0.17·t+4.4, r2 = 0.01, N = 168, NS; second day: v = 0.11·t+2.3, r2 = 0.008, N = 113, NS) and long-distance carriers (first day: v = 0.04·t+9.9, r2 = 0.0008, N = 60, NS; second day: v = −0.008·t+11.9, r2 = 0.007, N = 73, NS). (C) Proportion of long-distance carriers among individuals that picked up pellets. The number of observed pick-up events is shown in parentheses for each time interval.
Table 1.
Statistics for carrying distances and temporal dynamics in the three types of workers: excavators, short-distance carriers and long-distance carriers.
Table 2.
Long-term observations of individual excavators.
Figure 5.
Example of the distribution of pellets deposited by a single worker in the depositing box.
Shown are the clay arena, where individually marked workers excavated soil pellets, and the depositing box, where most pellets were dropped by the observed individuals. The thick arrow indicates the location where the animals entered the set-up. To reconstruct the individual pattern of pellet deposition for each observed worker, the ground of the depositing box was divided into 16 equal squares. The number of pellets dropped in each square by one individual was counted. Each bar in the graph represents one of the possible 16 areas distinguished inside the box, with the height of the bar indicating the number of pellets dropped inside this field by the observed individual. The worker observed for this example (worker nr. 7) placed 23 pellets in the box within a period of 44 min. The resulting index of dispersion was 7.2. The three-dimensional barplot is based on a script by Michal J. Figurski (http://addictedtor.free.fr/graphiques/graphcode.php?graph=161).
Figure 6.
Choices between two alternative locations to excavate.
To different locations were offered for excavation. At one side, pellets were deposited on the ground (see Fig. 1). The sample size was 80 for all conditions. Significance levels for binomial tests are indicated above the bars (‘NS’ not significant, ‘*’ P<0.05, ‘**’ P<0.01, ‘***’ P<0.001).