Building processes and plant species used in construction of boat-shaped houses

The construction of traditional boat-shaped houses follows a series of coordinated steps that integrate practical craftsmanship with deep-rooted ethnobotanical knowledge. The process typically begins with site selection and land clearing, after which a simple foundation is laid using local materials.

The primary frame and structure are erected using durable hardwoods for columns and beams. Species commonly used include Liquidambar formosana (Altingiaceae), Erythrophleum fordii (Fabaceae), and Homalium ceylanicum (Salicaceae), which are favored for their strength and resistance to decay. Other species such as Litchi chinensis (Sapindaceae), Melia azedarach (Meliaceae), Madhuca hainanensis (Sapotaceae), and Dalbergia hainanensis (Fabaceae) are also selected depending on local availability and builder preference.

Framing members, especially for the sidewalls and roof skeleton, are typically constructed using various bamboo species for their flexibility, light weight, and rapid growth. Commonly used species include Bambusa bambos (Poaceae), Bambusa textilis (Poaceae), Lingnania intermedia (Poaceae), and Bambusa xiashanensis. These are split, shaped, and tied into curved profiles that resemble an inverted boat hull.

For binding and joining structural elements, natural fiber materials are preferred over nails or modern fasteners. Rattan species such as Calamus tetradactylus (Arecaceae), Calamus simplicifolius (Arecaceae), and flexible vines like Urceola huaitingii (Apocynaceae), as well as Dinochloa multiramora (Poaceae), are used extensively due to their tensile strength and pliability.

Once the frame is completed, roof covering is applied. The most frequently used material is Imperata cylindrica (Poaceae), a hardy grass with excellent thatching properties. In some cases, leaves of Livistona chinensis (Arecaceae) are also used as supplementary roofing material.

The final step involves internal layout, flooring, and weatherproofing, often with adjustments based on seasonality and household size. Throughout the construction process, plant selection is not only based on availability but also on functional properties such as water resistance, weight, and ease of renewal—reflecting generations of empirical knowledge embedded in Li architectural traditions.

After these steps, the beautiful boat-shaped house is completed, and the whole community of local Li people will celebrate it in a big housewarming party for several days. They drink with the owner of the new boat-shaped house, sing traditional folk songs, and dance together. According to the traditional custom, special wizards are invited to hold religious rites to exorcize bad ghosts and to pray for happiness for those living in the new boat-shaped house. This kind of custom has lasted for thousands of years. Some folk housewarming songs were created and sung by local communities using their language long ago. One sentence in the lyrics of a song: “building a boat-house on a hillside. You can’t get enough of looking at it because of deep love. You drink spring water when you’re thirsty, and you eat betel nuts when your mouth is sour.”

A total of 26 plant species were documented as used in construction of the houses, including 13 tree species, 9 bamboo species, 3 liana species, and 1 herbaceous species, belonging to 13 botanical families. Each species or plant part was selected for a specific architectural function: Leaves were used mainly for the thatching of the roof, stems for beams, columns and frames, strips of bamboos and rattans for the binding or affixing of components (Table 3; Figs. 4, 5).

Table 3 Plant species used in construction of traditional boat-shaped houses by the Li people in HainanFig. 4

Ten main tree species used in the construction of boat-shaped houses in Li villages

Fig. 5

Main species of lianas (including rattans) and bamboos used in construction of boat-shaped houses in Li villages

Several species served as key materials in multiple functional domains. For example, Imperata cylindrica (Poaceae) and Livistona chinensis (Arecaceae) were the plants whose leaves were most commonly used for thatch to cover the roof. Binding was achieved using stems and strips of Urceola huaitingii (Apocynaceae), Calamus tetradactylus (Arecaceae), Dinochloa multiramora (Poaceae), and Calamus simplicifolius (Arecaceae). Hardwoods such as Erythrophleum fordii (Fabaceae), Dalbergia hainanensis (Fabaceae), and Litchi chinensis (Sapindaceae) were frequently used for structural support (e.g., columns and beams).

This study identified both unique and shared patterns of plant use among indigenous communities, consistent with prior ethnobotanical research. For example, Calamus species are commonly used for construction and handicrafts in tropical Asia, as confirmed in recent studies [36,37,38]. Similarly, the use of Imperata cylindrica leaves for roofing aligns with traditional practices documented in West Africa rural communities [39]. However, this study highlights the novel utilization of Madhuca hainanensis as a structural timber, a use less commonly reported in the literature, suggesting localized adaptation to available flora. Such findings underscore ecological and cultural drivers behind regional variation in ethnobotanical knowledge [40].

Ethnobotanical investigation shows that the construction of boat-shaped houses by the Li people relies heavily on specific plant materials that are primarily gathered from nearby mountainous areas, particularly in forested zones surrounding the villages. These resources include hardwoods for beams and columns, flexible vines and bamboo for structural frames and binding, and leaves for thatch roofing—each selected for its functional properties such as strength, flexibility, resistance to termites and rotting, and water resistance. Harvesting typically follows seasonal patterns aligned with the lunar calendar, with most materials collected during the dry season to ensure optimal durability and ease of processing due to dryness. Local knowledge plays a key role in determining the appropriate harvesting time and in identifying superior species based on criteria such as longevity, resistance to pests, and ease of manipulation. The Li people’s preference for boat-shaped houses is rooted not only in their climatic adaptability—providing ventilation, insulation, and structural efficiency—but also in their cultural and spiritual significance, as these structures symbolize ancestral heritage and identity. However, the continuity of this tradition faces mounting challenges, including the depletion of key plant species due to environmental change, declining transmission of craftsmanship, and reduced interest among younger generations, many of whom now favor modern brick structures over traditional forms.

According to the National List of Protected Wild Plants and the IUCN Red List, five of the tree species recorded: Erythrophleum fordii, Dalbergia hainanensis, Tectona grandis, Litchi chinensis (wild), and Madhuca hainanensis, are listed as nationally protected in China (Table 4). Homalium ceylanicum is listed as a provincially protected species. In particular, Madhuca hainanensis, Diospyros strigosa, and Dalbergia hainanensis are endemic to Hainan Island. Due to increasing habitat degradation and anthropogenic pressures, many of these species have become rare in tropical rainforest ecosystems [41,42,43].

Table 4 Five nationally protected plant species used in construction of boat-shaped housesCollection of performance data for plant materials in boat-shaped houses

The mechanical and ecological properties of plant-based materials used in the construction of Li people’s boat-shaped houses were evaluated on the basis of literature previously published and field data [44,45,46,47]. Different construction functions, including roof, structural support, framing, and binding, require different material characteristics (Table 5, Table 6).

Table 5 Comparison of mechanical properties of materials (leaves) used for roof covering of boat-shaped housesTable 6 Comparison of mechanical properties of woods, bamboos, and lianas used in construction of boat-shaped housesAnalysis of the thermal environmentIndoor and outdoor air temperature trends

Indoor temperature significantly affects human comfort by influencing thermal sensation, productivity, and overall well-being. Field measurements showed clear differences in temperature profiles between traditional boat-shaped houses and modern brick structures [50, 51]. In Chubao Village, during the dry season, the outdoor daytime temperatures ranged from 21 to 27 °C, while the brick houses recorded indoor air temperatures (23.5 °C) approximately 1.3 °C lower than those of the traditional dwellings (24.8 °C). In the rainy season, outdoor temperatures ranged from 27 to 31 °C, with brick houses indoor averaging 0.8 °C hotter than traditional buildings. Among traditional house types, Long Boudoirs exhibited slightly better thermal insulation, being on average 0.9 °C cooler than other traditional boat-shaped houses.

In Baicha Village, daytime temperatures during the dry season ranged from 25 to 33 °C. No significant differences were observed between the indoor temperatures of traditional houses and brick houses. However, in the rainy season (31–35 °C outdoors), the brick houses were consistently 0.8 °C warmer than the traditional houses (Fig. 6).

Fig. 6

Comparison of air temperature trends of boat-shaped houses between the two traditional Li villages. a Chubao Village; b Baicha Village

Relative humidity patterns

Relative humidity (RH) was generally higher in the rainy season than in the dry season, as expected. In Chubao Village during the dry season, brick houses had an RH approximately 4.4% higher than traditional boat-shaped houses. The granary recorded RH 4.3% higher than long boudoirs. In the rainy season, the indoor RH of brick houses was 5.5% higher than in traditional buildings, with granaries showing the highest RH among traditional types.

In Baicha Village, the outdoor dry season RH ranged from 44 to 62%. Interestingly, in this season, brick houses showed slightly lower indoor RH compared to traditional structures (4.4%). However, in the rainy season (outdoor RH 53–64%), indoor RH in brick houses was 7.3% higher than in traditional houses. Differences between traditional house types were minimal (Fig. 7).

Fig. 7

Trends in relative humidity of boat-shaped houses in two traditional villages. a Chubao Village; b Baicha Village

Comparisons of WBGT (wet bulb globe temperature)

The WBGT values reflect the risk of thermal stress. In Chubao Village during the dry season, all buildings maintained WBGT values below 24 °C, indicating low heat stress for their human occupants. Values for brick houses were 0.8 °C higher than those for traditional buildings. In the rainy season, brick houses recorded WBGT values that averaged 1.1 °C higher than those for traditional structures.

In Baicha Village, the WBGT during the dry season ranged from 24 to 27 °C outside, with brick houses showing WBGT 4 °C higher than traditional buildings. In the rainy season, outdoor WBGT values ranged from 27 to 31 °C, suggesting that human occupants would suffer a moderate to high risk of heat stress. Traditional boat-shaped houses, particularly residential types, performed significantly better than brick houses under natural ventilation, maintaining lower WBGT levels (Fig. 8). These results show that the Li people have cleverly planned and designed the production and living space according to the local natural and geographical conditions in the long-term practice of production and living and made the space as comfortable and livable as possible.

Fig. 8

Trends for WBGT values of boat houses in two traditional villages. a Chubao Village; b Baicha Village

Differences in microclimate by altitude in Chubao

In Chubao Village, boat-shaped houses at lower elevations (foothills) showed slightly better thermal performance than those at higher elevations (hillside). During the dry season, the indoor temperatures on the foothill were 0.3° C lower than on the hillside. The WBGT values were 0.6 °C lower in the foothills, while the humidity was slightly higher. However, in the rainy season, foothill houses recorded slightly warmer temperatures (+ 1 °C) and higher humidity (+ 3.6%)(Fig. 9).

Fig. 9

Temperature, humidity, and WBGT values for boat-shaped houses at different altitudes in Chubao Village

Differences in microclimates by orientation in Baicha

In Baicha Village, the orientation of the building influenced indoor environmental conditions. During the dry season, north–south oriented boat-shaped houses had indoor temperatures 2° C higher and WBGT values 0.8° C higher than those oriented east–west. In the rainy season, the differences in temperature and RH were minimal, although WBGT remained slightly higher in the north–south structures (Fig. 10).

Fig. 10

Temperature, humidity, and WBGT trends of boat-shaped houses in different orientations in Baicha Village

Independent samples t test analysis of indoor thermal-humidity environment differences among various building types

Variation for measurements in the same type of house, in the same village, and at the same time was relatively small. To ensure clarity and readability of the figures, error bars are not shown. However, standard errors for each measurement are provided in Tables 7 and 8. We compared the daytime indoor temperature, humidity, and wet bulb globe temperature (WBGT) of three types of traditional buildings within two villages. Independent samples t tests were conducted to assess the significance of differences among the three traditional building types, as well as between each traditional type and single-story modern brick houses. Comparisons with non-significant differences were filtered out, and the remaining results are presented in Table 7

Table 7 Significance of differences in temperature, relative humidity, and WBGT between two traditional Li villagesTable 8 Significance of differences in temperature and WBGT according to elevation and orientation

Similarly, we conducted independent samples t test on the differences in building elevation in Chubao Village and building orientation in Baicha Village. After filtering out non-significant comparisons, the results are presented in Table 8.