Subterranean termiteslive in the soil and wood that is in contact with soil. They have to discover food by constructing underground tunnel networks. Once the food is discovered and connected to the galleries, one important aspect of the foraging behavior is the food transfer by individual termites moving within the existing tunnels that lead to multiple existing food sources. In order to reveal how much the tunnel network is reliable to the food transfer efficiency, we used a lattice model suggested by Lee et al. (2006), which is capable of simulating the tunnel networks of Coptotermes formosanus and Riparius flavipes. After constructing the simulated tunnel networks, we randomly distributed food particles on the tunnel networks and then computed path entropy for the networksby selecting and evaluating the shortest paths from encountered food particles to the nest. The path entropy measured the degree of reliability of the networks for the food transfer entropy. Simulation results showed that path entropy between the simulated networks of C. formosanus and R. flavipes was significantly different due to the combinational effects of the network components such as the number of primary tunnels and the branching probability. We discussed the meaning of the results in relation to termite foraging efficiency.

Subterranean termites excavate tunnels for searching and transporting food below the ground, which in turn causes complicated tunnel networks. In the present study, we explored the connectedness of the networks by using spectral graph theory. In the theory, tunnel network pattern can be constitutively expressed by the Laplacian matrix, and among the set of all eigenvalues of the matrix, the second eigenvalue directly reflects the degree of the connectedness. We constructed the simulated tunnel networks for Coptotermes formosanus and Riparius flavipes by the use of a lattice model suggested by Lee et al., (2006) and computed their connectedness. The results showed that the values of the connectedness between the two species were statistically-significantly different. We briefly discussed the results in relation to foraging efficiency.

Subterranean termites construct complicated tunnel network for foraging below the ground. Thus, they often encounter a number of tunnel intersections during their moving from place to place in the network. In order to understand how termites respond to the intersections, we artificially excavated two tunnels intersected with 90° degree in soil-filled arenas. The two tunnels had the width of W1 and W2 (=2, 3, and 4mm), respectively. We systematically observed the response behavior of advancing termites to the intersection with the combination of W1 and W2, (W1, W2). For (W1, W2)=(2, 2) and (3, 3), the advancing termites passed the intersection without directional changes because it was difficult for termites to bend their body to change their moving direction due to the small-sized width. For (W1, W2)=(4, 4), the termites statistically-equally chose the three directions, left, right, and straight, which was due to the fact that the intersection provided enough space for termites to bend. For (W1, W2)=(2, 3), (2, 4), and (3, 4), termites, advancing in narrower tunnels, tended considerably to turn right or left, while termites, advancing in wider tunnels, were favorably inclined to go straight. These results can be understood by considering the relationship between termite body length and tunnel width as explained for the cases of W1=W2. In addition, we briefly discussed our findings in relation to termite foraging efficiency.