Ancient waterlogged seeds trace the melon’s global spread

Waterlogged melon seeds recovered from a medieval Chinese port are rewriting the story of one of the world’s most culturally significant fruits, revealing that the melons cherished during China’s Song dynasty arrived from South Asia rather than being domesticated from local wild plants — and that they likely tasted nothing like the sweet dessert melons we know today.

The discovery, published in Proceedings of the National Academy of Sciences by researchers at Washington University in St. Louis along with collaborators in Germany and the United Kingdom, offers a striking example of how ancient DNA can answer historical questions that conventional archaeology alone cannot resolve. It also paints an unexpectedly vivid picture of medieval Chinese life, one in which the color of a melon’s flesh may have resonated deeply with the aesthetic ideals of an entire era.

The seeds in question were unearthed in 2024 during an archaeological excavation of Shuomen Gugang, an ancient port located in present-day Wenzhou in southeastern China. During the Song imperial dynasty, which ruled from roughly 960 to 1279, this bustling harbor served as a critical node connecting continental East Asia with the Pacific and Indian Ocean worlds. The excavation revealed sunken shipwrecks laden with porcelain and the remains of imported fruits including grapes, olives, lychees, kiwifruits — and melons.

For ancient DNA researchers, such a find represents an extraordinary stroke of luck. Genetic material degrades rapidly in most environments, and successful ancient DNA extraction depends on exceptionally favorable preservation conditions. The perpetually waterlogged deposits at Gugang — anaerobic, cool and stable — provided exactly the kind of environment where plant DNA can survive for centuries. Amid the submerged remains of medieval maritime commerce, biologist Xinyi Liu, a professor of archaeology in Arts & Sciences at Washington University, saw an opportunity that few archaeological sites could offer.

The genomic analysis of these seeds yielded two major revelations. First, the ancient melons of East Asia most likely arrived from South Asia, challenging the long-standing assumption that they had been developed from wild Chinese plants. Second, these melons lacked the defining sweetness of modern dessert varieties. Instead, the analysis of color-related traits suggests they had light-green flesh and yellow peel — characteristics associated with melons consumed as vegetables, mild fruits, or for their seeds rather than as high-sugar treats.

The question of where melons truly came from has intrigued evolutionary biologists for decades. Susanne Renner, a biologist in WashU Arts & Sciences and co-author of the new study, has been investigating melon origins for more than 25 years. When she began her career, the prevailing view held that because the genus Cucumis, to which melons belong, includes numerous wild African species, melons must have originated in Africa.

Renner was never convinced by that logic. In 2010, she published a study based on approximately 100 Cucumis samples drawn from Africa, Australia and Asia, which revealed numerous previously overlooked species-level relatives in Australia and around the Indian Ocean basin. Her biogeographic analysis pointed to a different conclusion: melons most likely originated in Asia, not Africa. Subsequent research has since established that melon domestication occurred independently once in Northeast Africa and twice in India. Yet archaeobotanical evidence had hinted at a possible additional domestication event in southeastern China — a possibility the new ancient DNA analysis now complicates considerably.

By demonstrating that Song-era melons in China came from South Asia rather than from local domestication, the study warrants a reexamination of older archaeological evidence in the region. But how did these melons make their way to China? The researchers outline two historically plausible corridors. The first is the famous northern “Silk Road,” which ran along the foothills of the Inner Asian Mountains and the northern margin of the Tibetan Plateau. The second is the lesser-known southern “Tea Horse Road,” which traced the southern slopes of the Himalayas and passed through the valleys of southwest China.

“Both corridors are historically plausible, as demonstrated by the movement of other food resources,” Liu said. However, the archaeological record currently favors the southern route. The earliest melon remains in China have been recovered from Han and pre-Han royal tombs in southern China, and these finds predate comparable discoveries from northern Silk Road sites by approximately 800 to 1,000 years. The researchers therefore consider a southern route connecting southern China with Southeast and South Asia to be the leading candidate for the early introduction of melons into China.

Beyond the geography of melon dispersal, the study illuminates something far more intimate: the tastes and values of Song dynasty society. The light-green flesh of these ancient melons, the researchers suggest, may have carried profound cultural and symbolic weight. “The light-green color resonates with the fine-toned color of Chinese jade and with the glazes of celadon porcelain, both highly valued in East Asian aesthetic traditions,” Liu explained.

Indeed, during the Song dynasty, green celadon ceramics produced in the lower Yangtze region and beyond were frequently fashioned into melon-shaped forms, with glazes deliberately echoing the jade-green tones of melon flesh. This greenware ceramic tradition was popular across East Asia, spanning China, Korea and Japan, and many examples were recovered from the Gugang port excavation itself. The melon, with its abundance of seeds, has long served as a traditional symbol of fertility and prosperity in Chinese culture — a symbolism that may explain why its form and color were so eagerly adopted by artisan potters and prized by consumers throughout the region.

The researchers describe this as a rare and compelling case of “cultural-genetic symbiosis,” reflected in the parallel historical selection of melon traits and celadon aesthetics. It is a scenario in which human preferences and crop biology shaped one another: people may have favored melons whose flesh color matched the refined hues of their most treasured ceramics, while the popularity of those melons reinforced the aesthetic values encoded in the pottery. Culinary traditions echo this history as well. In the lower Yangtze region today, melons are still consumed as pickles or vegetables — a living remnant of medieval foodways preserved in modern kitchens.

The study also represents a significant technical achievement in the young field of plant ancient DNA. According to Liu and Renner, only a handful of crops have had ancient DNA obtained and contextualized with the evolution of specific traits. These include barley, certain types of wheat, cotton, maize and beans — all economically pivotal crops. The difficulty, Renner notes, lies in the generally very few genetic substitutions that can be attributed to selection on specific traits. Detecting the genomic signatures of domestication and trait selection in ancient samples requires both exceptional preservation and sophisticated analytical methods, making each successful study a milestone for the discipline.

What makes the melon study particularly remarkable is that it connects molecular genetics with material culture, trade history and culinary practice in a single analysis. The waterlogged seeds from a shipwreck in Wenzhou speak not only to the movement of a crop across the Himalayas but also to the sensibilities of potters, the symbolism of fertility, and the palates of medieval consumers. They demonstrate that the story of a fruit is inseparable from the story of the people who grew it, traded it, ate it and admired its colors.

The findings open new avenues for future research. If Song-era melons were introduced from South Asia, the precise source populations and timing of introduction remain to be determined through additional ancient DNA sampling across both regions. The possibility of an independent Chinese domestication event, suggested by earlier archaeobotanical evidence, likewise demands further investigation with modern genomic tools. As more waterlogged archaeological sites are excavated — and as sequencing technologies continue to improve — scientists expect that other crops with contested histories will receive similar treatment, transforming our understanding of how agriculture, trade and culture intertwined across ancient Eurasia.

For now, the melons of Shuomen Gugang stand as a testament to the power of ancient DNA to recover not just biological history, but human history — the tastes, aesthetics and aspirations of people who lived nearly a thousand years ago, preserved in the genetic code of seeds that sat quietly beneath the harbor waters.