Report of the Luffa Project


By Hugh M. Lewis


Copyright 2001, by Hugh M. Lewis



Fully laden Luffa plants on trellis in about late July, 2000 just ready for harvesting. Notice the overhang of the upper canopy and the angle of the sun by mid afternoon.



This brief paper is a report of an effort to cultivate a set of 30 1 Luffa plants, or dish scrubber or dishcloth gourds, from seeds brought back from China in late 1999 and planted during the growing season of the year 2000. These seeds were planted in March of 2000 and harvesting of melons began in August and extended through late October. Thus, the plants were allowed to extend in a growth period of just about 8 months. My interest in the Luffa was from the structure of the dried matrix of the melon that is used commonly in rural China as a dish and body scrubber. In China, the younger smaller melons are also eaten, usually cooked in a soup. The matrix of the melon has interesting properties when it is dried, and suggests a number of possible applications beyond their common use as scrubbers. The Luffa or Luffa cylindrica, is actually a member of the gourd family, or Cucubitaceae being a relative of the watermelon, muskmelon, pumpkin, cucumber, and the various kinds of squashes, including zucchini and bottle and other kinds of gourds. These are ornamental climbing vine plants with large leaves with pointed lobes and yellow flowers.

Back side of the Luffa trellise in about mid-August, at which time harvesting of dried melons had begun, notice the thick crowding of the vines that are matted and tangled.


I planted these plants in a single row 18 feet long, spaced about 6 to 7 inches apart, comprising a total area of about eighteen square feet. This was within the context of my last super garden that was also planted in two other rows of Chinese long beans and baby bak choi, as well as a set of sun flowers at one end of the garden, all of which were harvested earlier in the season by July. Timing between plants was almost perfect. The elevation achieved for drainage of the Luffa plants, being planted along the back wall that reflected the sunlight in the early part of Spring was about a foot off the ground compared to the local ground level. The row ran almost directly due West from East, thereby allowing sunlight from most of the day to reach the location.

The dried up Luffa vines by late October, just before Halloween, after the roots had been pulled out of the ground and the vines left to dry on the trellise. The trellis, made of furring strips, broke and bowed considerably under the weight of the vines and melons.

Based upon the growth of a single plant from the previous year in the same location without competition that resulted in ten usable melons plus one or two preformed and immature melons by the end of the growing season, I estimated that the maximum yield that might be coaxed from these 30 odd plants would be as much as 300 melons and more realistically around 250 gourds.

Final count of all melons grown, not including a number of preformed and immature melons that were mostly leftover at the end of the season was about 210 plus or minus 5. Of these 210 melons, about 150 were of sufficient size and many were much larger in size that were grown on the single plant in the previous season. Yield per plant was on average about 7 melons, 3 below the expected yield based upon the single plant from the year below.

It is my opinion that there occurred several important limiting factors that constrained the final output of this Luffa project. These limiting factors appeared to be the spacing between the plants and the competition between closely spaced vines for both sunshine and growing room. In the end, I believe that the competition for sunshine was the critical factor that determined the net productivity of this row of plants. Increased productivity would be demonstrated by more melons per plant, on average, and larger melons overall, on average.

Spacing for roots did not seem to be a critical factor, though it is a valid question to ask what an optimal spacing of plants might be. I have concluded that the optimal spacing for plants should be about 8 inches.

If plants were spaced 4 -5 inches apart, then a total of 48 to 50 plants might be accommodated in the same area, with a maximum yield possible of close to 4 to 500 melons. It is unreasonable to assume though that this many melons could be squeezed from this small of a space in one growing season. The expectation would be that the number of melons per plant would fall in proportion to the spacing and the number of plants, and that the average guord size would also be reduced.


Above, intermediate, straight Luffa of usable size. Below, larger, curved Luffa.


On the other hand, if a wider spacing were adopted, say 9 to 10 inches, in the number of plants that could be accommodated in the same area would be reduced from 30 to approximately 18 to 20 plants. The maximum yield of these 18 to 20 plants would be no greater than 200 melons. Therefore, it is concluded, that the plants were spaced at an optimal range, being an average of 6 to 7 inches between plants.


Long, large Luffa sorted and stacked at the time of the count


Short Luffa that were too small to be useful. Note the damaged odd size Luffa that are dried. These were short and fat, and unusual compared to the average.


I would suggest that in an intensive gardening venture, a spacing of 8 inches apart might be optimal for achieving maximum results.

I have therefore come to the conclusion that spacing in and of itself was not the critical determining factor in the final output of this particular garden, being close to the optimal level. In other words, the maximum number of plants that might be supported, at a spacing of 3-4 inches apart, might be greater than 30 but it would be expected that the number of melons per plant would fall off dramatically. On the other hand, too great a spacing between plants would dramatically reduce the number of plants and thus the total number achievable. It is expected that with an eight-inch spacing, a total of 27 plants could be acommodated, and the number of melons and size of melons per plant might be increased from 7 that was achieved to possibly 8 or 9, thereby achieving a greater total of between 220 to 250 melons. This would be the optimal output for such a garden in such a confined space, with the maximum possible being achieved at about 270 melons.

The crowding that occurred, and that I believe affected the net outcome in more decisive way, was the competition for sunlight and growing room for the long extending vines and the leaves. If this is true, then closer planting can be feasibly acommodated if a method can be designed that permitted a greater and more even spacing between the growing vines to achieve maximal distribution of sun-light. A trellis system larger and of a different design than the one constructed for this garden might accomplish this, especially if every other plant in the row can be trained in the opposite North/South direction in a herring bone manner.

There was considerable total loss in the number of melons harvested and subsequently handled during the drying period, which extended from the Fall months into November and early December. Damage to the skin of the melons, and breakage of the longer melons especially, was considerable due to poor hanging technique from the vines and poor handling and storage of the drying melons after being plucked. I estimate about 15% of the total melons were destroyed in this way. Once the skin of an undried melon is broken, it becomes very susceptible to subsequent bacterial and fungal infection and rot. It is possible to yet salvage the dish scrubber matrix after the fact of this rot, but the smell and sliminess of the melon makes it unattractive to handle and needs to be kept apart from the good melons.

Melons ranged in size from very small, found at the end of the growing season, to quite long and large. The majority of melons were intermediate in size and either straight or else twisted in a curving manner. Seeds obtained from the smaller, straighter melons tended to be black and smaller in color, while the seeds from the larger melons were white and larger. It is unknown if the difference between black and white seeded melons is a matter of different varieties of the plant or heterozygous traits of the same plant.

It is suspected that this difference is in fact a matter of heterozygous recessive traits of the black seeded melons, though not enough control was introduced at the beginning of this project to make a clear determination.


Total number of Luffa, sorted for counting, with the range of size and shape shown in front.

All the melons that survived were collected and saved and then sorted out. There were counted 50 large curved melons, or about 24% of the total, and 65 large straight melons, or about 31% of the total. Thus, of a total number of about 210 ( 5), about 55 percent of the melons were fairly large. The distribution of the Luffa appear to be by size (length) and curvature, being shown on a simple chi-square table below:










It is evident that as gourds grow longer, they tend either to curve or remain straight, though all appear to start of as relatively straight. Gravity may have a role in determining the shape of the resulting melon. It was noticed that those melons that tended to hang freely, suspended without obstruction, grew longer and straighter, while those that were supported by vines or the trellise, tended to curve. Heavy, long or severely curved melons tended, on average, to be more delicate and more easily broken due to handling and dropping. One or two long curved Luffa gourds broke merely under their own weight as a result of being lifted from their support during their growing phase.

The dimensions above conflate intermediate categories, and also variance in terms of the weight and diameter of the resulting mellons.


The larger, leftover Luffa during the colder months, drying in a protected area.



Of the remaining melons, about 33 were classified as short but of sufficient size to be usable, being about 16%, and the remaining, or about 47 melons, or about 22%, were too small to be used for drying. About 30 to 35 melons, or about 15% to 17% were destroyed either before, during the handling period, and afterward during the drying period.

Handling during the drying period was almost daily, having to move the melons periodically to the sun and then, in the colder months especially. We had to retire them in the late afternoon to a protected area so that they would not be affected by the drop in temperature and the dew at night.



Several tricks were used for drying, trying to allow air to circulate around the melons at the same time as being placed in the sun.




When the melons are sufficiently dry, a natural mechanism of seed dispersal is for the tip of the melon to break off, at the place where the melon connects to the vine, or the cap. A hole is produced from which the seeds separate from the dried core inside and fall out from the hole. Melons should not be stripped of their dried skin until after this dispersal stage has been reached by the plants, indicating that the melon is sufficiently dried.


Luffa showing naturally broken end, and seeds fallen out.


To summarize my report of the Luffa project, yields were close to the expected outcome for the super garden given the constraints of space at both the root level and in the resulting canopy, being about 75% of the optimal expected yield and about 50% of the maximum possible yield. Methods can be refined to improve production rates and efficiency of the overall arrangement and growing process, particularly in increasing total yield and average yield per plant, and in handling and drying techniques after harvesting. Furthermore, more careful experimentation will be undertaken in the third generation Luffa project to adapt and refine the methods developed during this last season, and to select melons of different sizes and shapes for separate test-area growing. For larger scale production of Luffa, special drying facilities need to be developed that reduce the amount of handling and increase the efficiency of solar drying and wind circulation and that protect the melons from damp and frost in the late fall and early winter months. Such a drying facility would also reduce the amount of labor involved in the drying and handling of the melons, which proved to be a considerable limiting factor in the overall process.

It is my intention to extend the Luffa project to subsequent generations, and to possibly grow other kinds of gourds as well.