1. Why Watermelons With Three Chromosome Sets Bloom but Cannot Make Seeds
Seedless watermelon plants produce normal-looking flowers. However, according to breeding resources from NC State University, triploid watermelons have three sets of chromosomes. During meiosis—the cell division that halves chromosome numbers to create sex cells—these three sets cannot divide evenly. As a result, the female and male reproductive cells do not function properly.[1]
If there are 2 sets, they can be divided into half each. With 3 sets, they cannot be divided into whole numbers. Studies on other triploid plants show that chromosome distribution becomes irregular during this process.[10] Without fertilization, no seeds can form.
Yet, the fruit still grows. The stimulation from pollination causes the ovary to develop into a fruit even without seeds. This process is called parthenocarpy (parthenogenesis). It shows that seedlessness and fruit production are separate events.[2]
2. How Seedless Watermelon Seeds Are Made by Crossing 4-Set and 2-Set Plants
Triploid seeds are made by crossing a diploid plant (with 2 sets of chromosomes) and a tetraploid plant (with 4 sets). NC State University resources explain that the tetraploid is used as the female parent, and the diploid provides the pollen. This direction matters; if the diploid is the female parent, no seeds are produced.[1] The egg from the tetraploid carries 2 sets, and the pollen from the diploid carries 1 set, totaling 3 sets.
The tetraploid parents are created by treating diploid seedlings with colchicine. This drug interferes with the spindle fibers that pull chromosomes apart during cell division, causing the chromosome number to double instead of splitting. Note that triploid varieties and their fruits do not come into contact with colchicine.[1]
Japanese geneticist Hitoshi Kihara (born 1893) is credited with establishing genome analysis methods, discovering ancestors of bread wheat, and successfully creating seedless watermelons.[4] The English introduction from the Kihara Institute for Biological Research states he received an award from the American Society for Horticultural Science in 1952 for his triploid watermelon research.[3] Since success dates vary by source, specific years are not emphasized here.
Making seeds every year is labor-intensive. But if you plant these triploid seeds in a field, will they fruit on their own? The next section explains why they need help.
3. Why Farmers Plant Seeded Watermelons in Seedless Watermelon Fields
Triploid watermelons cannot use their own pollen to set fruit because it is non-functional. They need pollen from diploid (seeded) plants. In fields, the typical ratio is one-third diploid and two-thirds triploid.[1] For example, if planting 30 plants, you might plant 10 seeded and 20 seedless. This ratio is a guideline.[1]
This means seedless watermelon fields are supported by seeded ones. The fact that triploids cannot fruit alone dictates how the field is planted.[1]
When you cut open a seedless watermelon, you may find white, soft structures inside. These are not always immature seeds; they are often empty seed coats (white or colored). Their presence depends on the environment and the size of the parent’s seeds.[1] Not all fruits are completely empty.
In watermelons, humans have engineered a way to get fruit without seeds. But does this happen in foods we do not engineer? Let’s look at bananas.
4. How Bananas Are Also Seedless Because of Three Chromosome Sets
Eating bananas have almost no visible seeds. The Japanese Encyclopedia (via Kotobank) explains that bananas are triploid, so normal meiosis does not occur, and they cannot produce viable pollen. Even if pollinated, germination is poor, and egg cells are not fertilized, so no seeds form, but the fruit still develops.[2]
However, sterility is not caused by meiosis alone. Research in Brazil suggests that infertility in edible bananas involves multiple stages, including defective spore formation due to irregular meiosis in triploids. The presence of a mature embryo sac does not prove fertilization can happen.[5] Saying "no seeds because there are 3 sets" is a bit too simple.
The three sets are also different from each other. A 2023 genome analysis found that the Cavendish banana (genome size about 1.48 Gb) contains three subgenomes: Ban, Dh, and Ze. These come from subspecies of Musa acuminata: banksii, malaccensis, and zebrina.[6] Triploids do not necessarily mean the same species’ chromosomes doubled; here, three different ancestral lines combine.
Finally, let’s look at a plant found in familiar places: the red spider lily.
5. Why Red Spider Lilies Rarely Make Seeds and Grow From Bulbs
The red spider lily (Higanbana), which blooms with red flowers in autumn, is also triploid. According to KSB News, Japanese red spider lilies are triploid, so they produce almost no seeds (and those that do are hard to germinate). They grow by bulbs. Chinese varieties are diploid and can produce seeds.[7] Japanese ones are thought to be "prehistoric naturalized plants" brought from China before recorded history, though the timing and route of arrival are debated.
Looking at the entire Lycoris genus, some species are diploid and make seeds, while others cannot make seeds and reproduce vegetatively. These non-seed-producing species are thought to be triploids or interspecific hybrids.[8]
The word "almost" is important. A report from the Kanagawa Prefectural Museum of Natural History found that some red spider lilies did flower and produce seeds, which germinated normally. They concluded it is not completely sterile but has extremely low fertility (Seto et al., 2011).[9] Therefore, "almost cannot" is more accurate than "absolutely cannot."
We see this triploid "seedlessness" in cultivation and observation.
6. How to Check for Seedlessness in Bananas and Watermelons Yourself
Cut a banana lengthwise and look closely at the center. Are there large seed-like clumps? If you eat a seedless watermelon, check for white skin-like structures inside and compare it with a seeded watermelon. When using a knife, do this with an adult.[1]
If you want to read more, the main sources are NC State University’s breeding resources [1], Kotobank [2], the Japanese KSB article [7], and the Kanagawa Prefectural Museum report [9]. Starting with materials in your own language can make it easier to begin.