# The role of den quality in giant panda conservation

> Giant panda cubs are born almost naked, so where their mother raises them matters. In a Chinese reserve with old-growth forest, mothers chose roomy, well-hidden tree hollows with relatively narrow entrances, and these dens stayed warmer, drier and steadier than the air outside, and than rock caves used in a logged reserve. The authors suggest that protecting large old trees could help pandas raise more cubs.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2019-wei-swaisgood-role-den-quality-giant/

## Citation and access

- **Authors:** Wei Wei, Ronald R. Swaisgood, Megan A. Owen, Nicholas W. Pilfold, Han Han, Mingsheng Hong, Hong Zhou, Fuwen Wei, Yonggang Nie and Zejun Zhang
- **Year published:** 2019
- **Journal or source:** Biological Conservation
- **Full citation:** Wei, W., Swaisgood, R. R., Owen, M. A., Pilfold, N. W., Han, H., Hong, M., Zhou, H., Wei, F., Nie, Y. and Zhang, Z. 2019. The role of den quality in giant panda conservation. Biological Conservation 231: 189–196.
- **DOI:** 10.1016/j.biocon.2018.12.031
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1016/j.biocon.2018.12.031

## Tags

- **Species:** Mammals > Giant panda
- **Region:** Asia > China
- **Tracking:** Beds; Claw marks; Hair; Tracking; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Cub-rearing; Den site selection; Denning
- **Methods and technology:** Chi-square test; Kolmogorov–Smirnov test; Logistic regression; Mann–Whitney U test
- **Data type:** Existing data; New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2019

## Detailed keywords

- **Species:** Studied: giant panda (Ailuropoda melanoleuca), maternal females and their cubs. Mentioned for comparison only: black bear (Ursus americanus), other bears, birds, marsupials and bats that use tree cavities.
- **Location and habitat:** Fengtongzhai Nature Reserve (102°48′E to 103°00′E, 30°19′N to 30°47′N), Baoxing County, Sichuan Province, China: 390 km², 1000 to 4896 m above sea level; mixed deciduous broadleaf forest changing to coniferous forest with elevation, with many stands of old-growth forest left. Fieldwork was in the reserve’s core, where panda density is highest. Foping Nature Reserve, Qinling Mountains, China: almost completely clearcut in the past, now mature secondary forest with few trees big enough for dens, so pandas there den in rock caves. The paper spells the first reserve three ways (Fengtongzhai, Fengtongzai, Fengtonzai).
- **Topics:** maternal denning ecology; den site selection; tree cavities (hollows) versus rock caves; large old trees as “small natural features” and keystone structures; microclimate buffering (temperature and humidity inside versus outside dens); highly altricial cubs and thermoregulation; carrying capacity; old-growth versus secondary forest; logging; artificial dens and “ecological traps”; climate change refuges; forest monitoring and management.
- **Statistics or models:** Kolmogorov–Smirnov test for normality; independent-samples t-tests; Mann–Whitney U tests; principal component analysis (PCA), with components of eigenvalue over 1 entered into logistic regression; stepwise discriminant function analysis (chi-square test of the discriminant function); paired t-test on logit-transformed microclimate data. SPSS version 21; significance level 0.05; data given as mean ± standard error. Inconsistencies: the methods name a paired t-test for the inside versus outside microclimate comparison, but the results for used versus unused and tree versus rock dens are all reported as Mann–Whitney U values; in Table 1, maximum interior height has t = 6.84 with P = 0.01, which does not match (a t value that large would give a far smaller P); the text says “concealment degree < 50 m” while Table 1 gives concealment as a score (1.09 versus 2.28); the loggers are described both as “HOBO-H8 hygrothermograph data loggers” and as “i-buttons”, and “twenty” loggers are given for twenty cavities each measured inside and outside. The results say tree dens in Fengtongzhai were on steeper slopes than rock dens in Foping, but the discussion says Foping pandas chose steep slopes and Fengtongzhai pandas gentle ones (used tree dens averaged about 16°). Supplementary Tables S1 to S6 and Figures S1 to S3 are not in the PDF.
- **Data type:** New field data: monthly monitoring of dens in the cub-rearing season (July to December), 2005 to 2010; tree cavity surveys on foot in Fengtongzhai, April to December 2010; microclimate recordings from 1 July to 31 December (year not stated). Existing data: rock den characteristics in Foping from Zhang et al. (2007), used for a qualitative comparison.

## Synopsis

### 1. What they did

Giant panda cubs are born almost naked and helpless; the authors note they are the most altricial (least developed at birth) of all placental mammals. Pandas do not hibernate, but mothers raise their cubs in dens for at least three to four months after birth, in tree hollows or rock caves. Earlier work in Foping Nature Reserve, where old-growth trees have almost all been logged, looked at how pandas choose rock dens. This study looked at tree dens in old-growth forest, and at how well dens shelter cubs from the weather.

- Tree den survey. In Fengtongzhai Nature Reserve, Sichuan, the team searched on foot for tree cavities big enough for an adult panda to enter and stand on all fours. A cavity counted as a used den if a mother and cub were found there during monthly monitoring in the cub-rearing season (2005 to 2010), or if it held panda signs: fur, scratch marks or bedding material.

- Den features. At each cavity they measured 19 features of the den and its surroundings, such as slope, tree size, entrance and chamber size, and how well the den was concealed. They compared used dens with unused cavities, and compared tree dens in a general way with the rock dens described earlier in Foping.

- Den climate. In ten tree cavities in Fengtongzhai and ten rock cavities in Foping, they placed temperature and humidity loggers inside each cavity and 3 to 5 m outside it. These recorded every 1.5 hours from July to December, while the cavities were empty. A bigger difference between inside and outside meant better “buffering” against the weather.

### 2. What they found

Which tree cavities pandas used

- Of 54 tree cavities surveyed, 22 had been used as dens.

- Nine of the 19 features differed between used dens and unused cavities. Used dens were better concealed, on gentler slopes (about 16° against 28°) and in larger trees (average diameter at breast height 1.24 m against 1.10 m).

- Used dens had wider entrances, closer to the ground (about 71 cm against 113 cm), and larger chambers. Compared with the chamber, though, the entrance of a used den was relatively narrow.

- Slope, entrance width and concealment together correctly sorted 86.4% of cavities into used and unused.

Tree dens compared with rock dens (qualitative)

Tree dens in Fengtongzhai were at lower elevation, on steeper slopes, farther from water and showed more signs of use by other species than rock dens in Foping. Rock den entrances and chambers were wider and shorter, and tree dens were roomier inside relative to their entrances.

Climate inside the dens

- All dens, tree and rock, buffered against outside conditions. Inside, the average and lowest temperatures were warmer, the highest temperatures cooler, the air drier and conditions much steadier through the day and night.

- The difference was largest at night: at midday and early afternoon dens were only somewhat cooler than outside. Humidity followed the same daily pattern.

- Used dens buffered better than unused cavities. Used tree dens were on average 4.51 °C warmer than outside, unused tree cavities only 1.47 °C. Used tree dens were also drier, with the biggest difference in the colder winter months. Used rock dens also buffered better than unused rock cavities, though less strongly than tree dens.

- Used tree dens buffered better than used rock dens: by 1.44 °C more for temperature and 4.24% more for humidity, and they were also better against the coldest and most humid days.

### 3. What they concluded

- Den quality, including structure, surroundings and microclimate, decides the conditions in which panda cubs are raised.

- Mothers choose dens with a small entrance and a roomy chamber, which likely protects cubs from predators and weather, and concealed dens, which likely protect against predators and human disturbance.

- In Foping pandas chose steep slopes for rock dens, but in Fengtongzhai they chose gentle slopes for tree dens. The authors suggest this is because large old trees grow better on gentle slopes, while caves are more common in steep terrain.

- Tree dens give more stable and favourable conditions for rearing cubs than rock dens. The authors suggest that forests with old-growth trees may support more pandas than logged forests, which may help explain why pandas sometimes favour old-growth forest. Whether den microclimate affects cub survival still needs study.

### 4. How tracking was or was not used

Used dens were identified by finding a mother and cub during monitoring surveys and/or by panda signs at the cavity: fur, scratch marks and bedding material. The authors consider these signs a reliable sign of a maternal den, because pandas do not hibernate and are only known to use dens for raising cubs. How the monitoring surveys found the dens is not described.

### 5. Tracker or observer certification

Not reported. Who identified the signs at the cavities is not described; four people are thanked for help with fieldwork. This matters because the split between used dens and unused cavities rests partly on sign identification.

### 6. Other technologies, skills or innovations

- Temperature and humidity data loggers placed 1 m above the ground in the centre of each cavity and 3 to 5 m outside it, recording every 1.5 hours, day and night.

- Comparing inside with outside readings for each den let the authors compare tree and rock dens in two reserves with different climates.

### 7. Implications for conservation (from the paper)

- Large old trees may be a limiting resource for denning mothers, possibly helping cubs survive. Preserving them may raise the number of pandas a reserve can hold, and managers and policy makers are encouraged to prioritise them in panda habitat.

- Current panda monitoring does not include systematic surveys of large old trees; future monitoring and management plans should include them. Declines in cavity-dependent species can lag behind the loss of tree hollows.

- Where large old trees are scarce, artificial dens could be trialled, following the structure and microclimate features of preferred dens and tested for how well they buffer the weather. Otherwise they risk becoming “ecological traps” that attract pandas without giving cubs what they need. Trials are planned in Liziping Nature Reserve.

- Tree cavities and old-growth forest may become even more important as refuges under climate change.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- Fur, scratch marks and bedding material at a cavity were enough for the researchers to count it as a used maternal den. Den selection studies like this one depend on reliable sign identification at the den, yet the paper does not say who read the signs or how well.

## Cite this summary

Lawrence, K. 2026. Summary of: The role of den quality in giant panda conservation (Wei and Swaisgood et al., 2019). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2019-wei-swaisgood-role-den-quality-giant/