4 DISCUSSION
By monitoring a total of 44 sample lines, we found that the scent-marking behavior of giant pandas is indeed consistent with signal detection theory (Alberts, 1992), and that giant pandas do select and modify their marking sites to maximize their signaling function at both micro- and macro-habitat scales. Giant pandas are an energetically marginal species, their diet is bamboo and their nutritional phase is low, which affects aspects of movement, communication and reproduction (Hu et al., 1985; Nie et al., 2012b; Nie et al., 2015). The selection of marking sites will minimize energy expenditure while increasing the effectiveness of inter-individual communication.
Our findings show that pandas prefer to scent mark on ridges with animal trails, and we know by the density of faeces on the sample line that this is indeed the area with the highest intensity of use by pandas as well (Table S1). The presence of many animal trails on the ridges also makes them important paths for the movement of giant pandas (Bai et al., 2020). However, we likewise noted that slopes without animal trails as well as ridges had high fecal densities, demonstrating that pandas spend a significant amount of time in these areas, yet the preference index for these areas was significantly lower than the corresponding terrain with animal trails. From this we learn that pandas choose these locations for scent marking not because they happen to be there, but because they really need to choose such terrain to maximize signal propagation efficiency.
The tree species most frequently scent-tagged by giant pandas were Badung oak (Quercus engleriana ), Yaupon pine (Pinus tabulaeformis ), and Huashan pine (Pinus armandii ), which are dominant species with a wide local distribution (Ming et al., 1999). We speculate that they choose dominant species for marking because they take less time and energy to be found. In addition to their wide distribution, their high marking frequency may also be due to the fact that marked trees are well located for signal transmission and over time become scent stations (Hu et al., 1985), and they are often located along animal trails on mountain ridges (Nie et al., 2012a). Marking on these scent stations commonly used by giant pandas would also increase the likelihood of their information being detected by signal receivers. Zhou et al. (2019) have demonstrated that both volatile and non-volatile compounds in AGS marks were not significantly degraded within two weeks, suggesting that pandas visit and update marks frequently (Zhou et al., 2019). Thus depositing marks in better located scent stations would also reduce the energetic cost of returning and updating markers. Future work could increase research on scent stations and explore their microhabitats and conditions such as light that may affect marker persistence and dispersal to increase conservation of habitats where potential scent stations exist.
The roughness and SDT of the marked tree of giant pandas differed significantly from the control tree, similar to the results of previous study (Nie et al., 2012a), where giant pandas generally do not scent mark on smooth trunks, which would reduce the persistence of marking; and the SDT of the marked tree was significantly shorter than that of the control tree because the closer the distance to the animal trail, the more likely it is to be detected by the signal recipient, and because when marking here, giant pandas can travel shorter distances and save energy.
For both marked and control samples, bamboo density, bamboo cover and slope were lower and tree density and cover were higher in the marked sample. Higher bamboo density and cover make it more difficult for giant pandas to walk among bamboo forests, while relatively low bamboo density will allow giant pandas to obtain sufficient nutrition while reducing energy expenditure to traverse bamboo forests (Wei et al., 2015). As for the lower slope of the marked samples, the reason is that the gentle slope is a suitable habitat preferred by pandas (Hu et al., 1985), and the lower slope ensures that pandas complete scent marking in a relatively stable position (Hou et al., 2021), and walking in a gentle slope reduces energy expenditure. The higher tree density and tree cover in the marked sample is mainly due to the fact that the more trees there are, the greater the chance of trees with larger DBH appearing around them, and trees with larger DBH are more likely to form tree dens for giant pandas to breed and nurse their cubs (Wei et al., 2018; Zhang et al., 2011). In addition, the higher the density of arborvitae, the more concealed the habitat is, and thus the less likely to be detected by natural predators (Hu et al., 1985).
We found that the number of marked trees that were reused was much higher than those that were used only once, but did not detect significant differences between the variables of either. We hypothesize that the location of the repeated marked trees may be more prominent and have a wider field of view, where pandas often pass by, which may also increase the chance of other similar species finding the signal. The most used mark type in giant pandas is the AGS marks, which often remain in the natural environment for more than 3 months because of their low volatility (Hagey & Macdonald, 2003; Swaisgood et al., 2004), in order to better convey individual information. In contrast, urine markers contain more volatile substances than AGS marks and have a shorter retention time in the environment, and are often used to convey the estrus status of individuals (Zhou et al., 2019) and the competitive ability of males (White et al., 2002), among others.
The appearance of scratch mark is always accompanied by AGS mark, so we believe that scratch mark itself does not convey information, but is a by-product left behind when AGS marking is performed. An interesting finding is that some marked trees have only bite mark and are more numerous than those with only urinary mark, which was not found in previous studies, and we predict that its role is also correlated with mark height, and if the mark is higher, it indicates that the individual is larger and more competitive, similar to AGS mark and urinary mark (White et al., 2002), but specific role needs to be further investigated. McGuire et al. found that body size affects the frequency of marking in a study of domestic dogs (Canis lupus familiaris ) (Mcguire & Bemis, 2017). However, in the study of giant pandas, we only know that body size reflects their competitive ability (Nie et al., 2012b), but whether it affects the frequency of marking deserves further investigation.
Statistics on mark orientation revealed that the most marks were toward the animal path, which is because facing the animal trail also increases the chance of the mark being detected by its own species; for those marks that were backward toward the trail, we predicted possible scent counter-marking behavior, which is one of the main responses taken by animals when they encounter the scent of other competitors of the same species (Johnson, 1973). But before counter-marking they evaluate individuals that are already marked, probably because they are not as competitive as the former and want to mark in a superior position on the marker tree, so they mark on the back (Müller & Manser, 2008). However, the exact reason needs to be confirmed by a large number of future studies.
Footpad scent communication has been found in fellow bear species brown bear (Ursus arctos ) (Sergiel et al., 2017) and polar bear (Ursus maritimus ) (Owen et al., 2015), and most bear species have large home ranges and are non-dominant, and releasing scent while walking is a very effective form of intraspecific communication. An interesting question is whether giant pandas use footpad scent similarly for communication? We monitored some sites in this study where giant panda tracks were pitted (Figure S4). Unfortunately, we did not find other pandas observing and sniffing the tracks in our follow-up observations. This may be due to our small sample size or the fact that we did not specifically focus on this direction, and we hope to continue this study in the future to confirm our suspicions.
It has been shown that the frequency of marking in giant pandas (Nie et al., 2012a) and the compound content of AGS marks (Zhou et al., 2019) differ significantly between the sexes. Zhou et al. detected large differences in the composition and content of compounds in AGS marks between captive and wild giant pandas (Zhou et al., 2019), which may also be one of the reasons for the low reproductive ability of captive giant pandas, and future research on scent marks of giant pandas in the field environment should be strengthened to clarify which compounds affect the estrus and reproduction of giant pandas, and applied to captive giant pandas to improve their reproductive success. This is a tremendous step forward for the survival and reproduction of the panda populations.
However, current research on chemical communication has still not kept up with acoustic communication (Wiley, 2006). Due to the construction of Giant Panda National Park, some human facilities will inevitably be built in the living environment of giant pandas in the future. It has been shown that human facilities can affect the scent marking behaviors of domesticated felines (Krofel et al., 2017), and scent marking of pandas should be continuously studied to detect whether the marking behaviors of pandas will be changed due to the influence of human facilities, so that more precise conservation of panda habitats can be carried out. For example, assessing the impact range size of human facilities prior to their construction and then locating them as far away from core giant panda habitat as possible.