# Effects of water level on sleeping-site selection and inter-group association in proboscis monkeys: why do they sleep alone inland on flooded days?

> Proboscis monkeys are known for sleeping in trees on the riverbank, where a clouded leopard can only reach them from the land side. But when floods put the forest floor of a Bornean river forest under a metre of water, one group moved inland to sleep, and slept alone. The authors suggest that deep water keeps predators out, freeing the monkeys from their usual nightly retreat to the river.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2010-matsuda-tuuga-effects-water-level-sleeping/

## Citation and access

- **Authors:** Ikki Matsuda, Augustine Tuuga and Seigo Higashi
- **Year published:** 2010
- **Journal or source:** Ecological Research
- **Full citation:** Matsuda, I., Tuuga, A. and Higashi, S. 2010. Effects of water level on sleeping-site selection and inter-group association in proboscis monkeys: why do they sleep alone inland on flooded days? Ecological Research 25(2): 475–482.
- **DOI:** 10.1007/s11284-009-0677-3
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1007/s11284-009-0677-3

## Tags

- **Species:** Mammals > Proboscis monkey
- **Region:** Asia > Malaysia
- **Behaviour and ecology:** Behaviour; Daily travel distance; Feeding; Habitat selection; Movement; Predation risk; Predator avoidance
- **Methods and technology:** Direct observation; Global Positioning System (GPS); Mann–Whitney U test
- **Data type:** New field data
- **Observer qualification:** Not needed
- **Discipline:** Behaviour and ecology
- **Year published:** 2010

## Detailed keywords

- **Species:** Studied: proboscis monkey (Nasalis larvatus), one habituated one-male group (the BE-Group, 16 animals at the end of the study: one adult male, the alpha male Bejita, six adult females, five juveniles and four infants), plus counts of other one-male and all-male groups along the river. Potential predators at the site, not studied: clouded leopard (Neofelis diardi) and estuarine crocodile (Crocodylus porosus).
- **Location and habitat:** Riverine forest along the Menanggul River, a tributary of the Kinabatangan River, Lower Kinabatangan Wildlife Sanctuary, Sabah, Malaysia (Borneo). Natural forest to the south of the river; to the north, oil palm plantation with a protected strip along the river. Seasonal flooding of the forest floor; monthly mean river level from 66 cm (August 2005) to 345 cm (February 2006).
- **Topics:** sleeping-site selection; riverbank versus inland sleeping sites; seasonal flooding and water level; predation threat and predator avoidance; clouded leopards as terrestrial predators; multi-level society; inter-group association between one-male groups; ranging behaviour; daily path length; distance from the riverbank; food availability and fruit phenology; diet and dietary overlap; tree falls along the river; a flood as a “natural experiment”.
- **Statistics or models:** Spearman’s rank correlation; Mann–Whitney U test; Kruskal–Wallis test; Holmes-Pitelka index of dietary overlap; focal animal sampling; means and standard deviations. Possible inconsistencies: the Spearman result for water level and riverbank sleeping is given as rs = 0.31 with p = 0.01 and n = 13, although the text says riverbank sleeping was less frequent at higher water (which would be a negative correlation), and a correlation of 0.31 over 13 months would not normally be significant; the Kruskal–Wallis results for feeding time and for fruit eating across the three January periods are both given as H = 1.87, p = 0.39.
- **Data type:** New field data, May 2005 to May 2006: full-day focal follows (11 to 17 days a month, except February 2006), late-afternoon boat surveys (6 to 13 days a month, 138 in total; 22 evenings in February 2006), GPS locations every 10 minutes, river level measured each evening, and monthly phenology surveys of 2,142 labelled trees and vines on 16 trails.

## Synopsis

### 1. What they did

Proboscis monkeys live only on Borneo. They are thought to return every evening to sleep in trees on the riverbank, where predators such as clouded leopards can only reach them from the land side, and neighbouring one-male groups often gather there for the night. Because the inland forest is swampy and hard to work in, most earlier studies watched the monkeys from boats.

The authors followed one habituated one-male group of proboscis monkeys, the BE-Group, in riverine forest along the Menanggul River in Sabah, Malaysia, from May 2005 to May 2006. During the floods of early 2006, the group stopped coming to the riverbank to sleep. The authors asked why.

- Following the group. Observers (the first author and two research assistants) followed the alpha male, Bejita, and one female from dawn to dusk, recording what they did and what they ate. Bejita’s position was logged with a Global Positioning System (GPS) unit every 10 minutes; the group rarely moved more than 50 m from him.

- Boat surveys. On other evenings they searched the river by boat for the BE-Group’s sleeping site, and counted other groups sleeping within 500 m of them.

- Water level. They measured the river level every evening.

- Food. Each month they checked 2,142 labelled trees and vines along 16 forest trails for young leaves, fruits and flowers.

- The January flood. They compared Bejita’s feeding before, during and after the flood (three 3-day periods), and compared fruit availability near the river and further inland.

### 2. What they found

Where the group slept

- The group’s sleeping site was found or inferred on 295 nights: 261 on the riverbank and 34 in the inland forest. Nine of the inland sites were seen directly; the others were inferred because the group could not be found on the riverbank by boat.

- Riverbank sleeping sites were used at river levels of 8 to 433 cm (median 108 cm); inland sites at 220 to 455 cm (median 350 cm). The group slept at the riverbank less often when the water was higher.

- On 11 January 2006 the river reached 238 cm and the forest floor was under about 1 m of water. The group then wandered in the forest and did not return to the riverbank until the water dropped on 18 January. Inland, they slept in small open gaps or near old tractor roads.

- The group could not be found at all in February. It was found again on 3 March, while the forest was still flooded.

How far they moved

- Daily path length did not differ between inland nights (mean 830 m) and riverbank nights (798 m).

- On inland days the group went much further from the river: on average 564 m (356 to 800 m), against 159 m on riverbank days.

Other groups

- On high-water days, boat surveys found far fewer groups of proboscis monkeys along the river (mean 2.7 per survey) than on low-water days (7.0), suggesting that other groups also slept away from the river.

- When the BE-Group slept inland, no other one-male group was seen within 200 m (one was seen by chance about 300 m away).

- When they slept on the riverbank, another one-male group slept within 200 m on 101 occasions (40.1% of riverbank nights), 37 times within 50 m, sometimes in the same tree, and usually on the same side of the river.

Food and feeding

- Fruit was more available 150 to 400 m from the river than within 150 m, in both flooded and non-flooded months, so fruit inland did not increase during the floods.

- In January, the time Bejita spent feeding (23 to 29%) and the share of fruit in his diet did not differ significantly between the periods before, during and after the flood. His diet overlapped by 46 to 59% between the periods.

Tree falls

- Only three of 21 tree falls along the river happened during the flood months (I. Matsuda, unpublished data). After five trees fell on a windy day in September, the group still slept on the riverbank.

### 3. What they concluded

- The water level, not food or falling trees, best explains why the group slept inland. Deep water over the forest floor stops terrestrial predators such as clouded leopards from hunting, so the monkeys no longer need the safety of the riverbank.

- With less threat from predators, one-male groups may also have less need to sleep near each other. Water level and predation threat may shape both where proboscis monkeys sleep and how their multi-level society is organised.

- The results come from a single group in riverine forest and may not apply to proboscis monkeys in mangrove or peat swamp forest. More studies of ranging during floods in other forest types are needed.

### 6. Other technologies, skills or innovations

- A handheld GPS unit recording the alpha male’s position every 10 minutes during full-day follows.

- A simple way to estimate how often the group slept on the riverbank: before the floods, boat surveys always found the group when it slept there, so nights when it could not be found were treated as nights spent inland.

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

- The paper does not discuss conservation directly. The authors note that the floods acted as a kind of natural experiment, and that the study is an example for thinking about how animals behave to avoid predators, in primates and other mammals.

## Cite this summary

Lawrence, K. 2026. Summary of: Effects of water level on sleeping-site selection and inter-group association in proboscis monkeys: why do they sleep alone inland on flooded days? (Matsuda and Tuuga et al., 2010). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2010-matsuda-tuuga-effects-water-level-sleeping/