Age-based spatial distribution of workers is resilient to worker loss in a subterranean termite

Abstract

Elaborate task allocation is key to the ecological success of eusocial insects. Termite colonies are known for exhibiting age polyethism, with older instars more likely to depart the reproductive center to access food. However, it remains unknown how termites retain this spatial structure against external disturbances. Here we show that a subterranean termite Coptotermes formosanus Shiraki combines age polyethism and behavioral flexibility to maintain a constant worker proportion at the food area. Since this termite inhabits multiple wood pieces by connecting them through underground tunnels, disastrous colony splitting events can result in the loss of colony members. We simulated this via weekly removal of all individuals at the food area. Our results showed that termites maintained a worker proportion of ~ 20% at the food area regardless of changes in total colony size and demographic composition, where younger workers replaced food acquisition functions to maintain a constant worker proportion at the food area. Food consumption analysis revealed that the per-capita food consumption rate decreased with younger workers, but the colony did not compensate for the deficiency by increasing the proportion of workers at the feeding site. These results suggest that termite colonies prioritize risk management of colony fragmentation while maintaining suitable food acquisition efficiency with the next available workers in the colony, highlighting the importance of task allocation for colony resiliency under fluctuating environments.

Figure 1

Composition of worker instars (≈ age) (circles) at the food (left) and at the reproductive area (right) after weekly removal events in Coptotermes formosanus colonies. Instar composition was investigated in randomly selected 40 workers for four different colonies (results were pooled) in each removal event and at the end of experiment for the composition of instars at the reproductive area. Worker instars were determined by counting the number of antennal articles followed by Chouvenc and Su (2014). The proportion of instars in each removal was determined by number of individuals in each instar divided by number of workers (n = 160) and the size of the circles varied depending on the percentage of worker instars. Double asterisks denote significant differences in Chi-square test (α = 0.01).

Figure 2

Per-capita food consumption rate (mg of wood consumption/number of workers in the colony/week) over average worker instar at the food area in Coptotermes formosanus colonies. The average worker instar was determined from each removal event by calculating the average from 40 individuals per colony. Square, circle, triangle and diamond shapes represent colony 1, 2, 3 and 4 respectively. Gray area and a solid line indicate 95% confidence intervals and regression line respectively.