Feeding signs give us information about the feeding ecology of animals, and can also provide information on animal distribution or abundance, as well as habitat selection. However, feeding signs do not necessarily directly reflect the distribution or density of animals. In addition, it is difficult to detect the presence of endangered species only from the presence of their feeding signs. Therefore, we need to understand the factors and the environmental conditions that influence the formation of feeding signs.
Most of the habitat of the Asiatic black bear (Ursus thibetanus) in Japan is close to human settlements, so monitoring of bear densities will help predict bear intrusions and decrease conflict between humans and bears. Because bears live in forests in low densities, it is difficult to directly find them. Therefore, feeding signs may be efficient as one of the estimating methods for bear densities.
Arboreal feeding sign (AFS) is a noticeable field sign made by Asiatic black bears. When the bears feed on trees, they break branches in the tree crown to acquire the fruits or leaves that are attached to the branch tips, and thus the broken branches get piled at the tree tops and these branches (i.e., AFS) look like birds’ nests. Some local governments in Japan have used observations of AFS as an indicator of bear density. However, it is unknown whether AFS are related to bear density and what kinds of factors may influence AFS formation. For example, bears may climb and eat hard mast (i.e., create AFS) before the fruits are ripened and the quantity of hard mast does not reach maximum value.
Here, we researched the influence of fruit production at individual tree and regional scales on the detection of arboreal feeding signs, to clarify in what cases bears climb to eat hard mast. For seven years, we counted the number of fruits of 374 to 481 trees each year of three dominant hard mast species (Quercus crispula, Quercus serrata, and Castanea crenata) in the Ashio-Nikko Mountains of central Japan, and estimated individual tree and regional mast energy values, calculated by averaging energy values of all three species annually. We also checked for the presence or absence of AFS in these trees, and analyzed the influence of factors such as individual tree energy and regional mast energy on the detection probability of AFS.
As a result, we found that AFS were more likely in individual trees with larger mast energy values. This indicates that climbing may be costly for bears, and so they only climb the trees bearing a lot of fruit to maximize their feeding efficiency. Because bears cannot digest fiber efficiently, quantity may be particularly important in the feeding plant. Moreover, AFS were created more during poor mast years. In those years, the energy that bears can obtain when climbing a tree is lower and there is much greater competition for the hard mast on the forest floor with other animals, such as mice. Whereas in good mast years, bears can eat a lot of mature hard mast without climbing, because a lot of the fruit will fall on the forest floor. There was little difference in these factors among the three hard mast species despite the several differences in characteristics among them (e.g., fruit size or masting interval).
Additionally, we found interaction between the factors (individual tree energy and regional mast energy). In poorer mast years, AFS were likely to be created in trees that were not selected in greater mast years. This is probably because the number of trees bearing a lot of fruit decreased. In greater mast years, bears might not climb trees to eat the fruit regardless of individual tree energy.
In conclusion, bears may maximize their feeding efficiency in various ways depending on fruit quantity of individual trees and across the region. Therefore, AFS cannot be used as an indicator of bear density unless both individual tree and regional mast energy values are monitored. To make AFS a reliable indicator of bear populations, comparative studies are needed at sites with similar tree species composition to our study area and obviously with different bear densities.