English (en) | Change Language
By: Sophie Roberts, Yuwadee Danmalidoi, Rattakarn Arttawuttikun, Kitichai Sampunsinkor, and Abram Bicksler

This article is from ECHO Asia Note #26


ECHO Asia Impact Center staff members first heard about alternative herbicide recipes that use fermented papaya and pineapple from a retired technical school teacher and organic farmer, Kru Pratoom. As weeding is a big part of any farmer’s life, the Seed Bank staff wanted to try out a lower-risk herbicide to see if its effects on weeds would warrant its use. They also wanted to ensure that this herbicide would not pose a risk to soil pH, microbiology, structure, and plant uptake and health. This ECHO Asia Research Note describes the process used to create this herbicide, as well as a sampling technique to determine its efficacy on weeds. Look for a future note about the methodology used to help determine its effects on soil microorganisms and health.

Herbicide Use

Herbicides are one method used to control unwanted plants in crop fields, gardens, and orchards. Herbicides may be used alongside tillage, hand weeding, burning, crop rotation, and crop spacing to control weed pests. Weeds can cause agronomic and economic harm by reducing the quantity or quality of the desired crop plant by interfering at important stages in the growth and development of the crop (O’Donovan, 2009). Weeds compete with crops for space, sunlight, water, and nutrients; some species of weeds may also release toxins into surrounding soil that may damage crops (Swanton, 2009). The damage caused by weeds depends on the weed species, crop species, and stage of crop development. Weed control is most critical when crop

Alternative Herbicide Article EAN #26 H1

Figure 1. Weeds present in our sample plots. (1) Imperata cylindrica, (2) Euphorbia hirta, (3) Mimosa pudica, (4)Convolvulus arvensis.

plants are still young and are not tall enough to out-compete the weeds. A low quantity of weeds may not damage crops, and completely clearing weeds is associated with high economic and time costs for the farmer. Besides, such clearing may cause damage to the agroecosystem. Additionally, certain weeds can be used for food, medicine, and fodder, so some so-called “weeds” can be economically useful. When applied at the proper time, and in the right quantity, herbicides can be a useful tool as part of an Integrated Pest Management System to control weed species that are too numerous or that may cause undue agronomic or economic harm (FAO, 2015). In sustainable systems, herbicides are used when other control methods fail to effectively control weed pests. Unlike in conventional agriculture, they are not the primary method of controlling weeds.


The alternative herbicide recipe discussed in this ECHO Asia Note contains lye, which can be extremely caustic; this means any physical contact with the undiluted product can cause burns, and therefore caution must be used, even around the mixed product. Always add lye or solutions containing lye to water or other liquids; never add water to concentrated lye or lye solution. Adding them in the correct order can minimize the damage caused by backsplash, because adding the lye/herbicide to water means the backsplash is more diluted. Whenever you are working with lye or the herbicide, wear rubber gloves and boots. Long pants and sleeves are recommended for both mixing and spraying. To protect your eyes, wear safety goggles or sunglasses. Perform all mixing and pouring in an open and well-ventilated area, in case the lye reaction gives off fumes. Do not mix or store in a metal container, because the lye could react with the metal.

If you do come into contact with the herbicide or with concentrated lye, immediately wash the affected area with cool water for at least fifteen minutes. Test all containers beforehand with boiling water to ensure they won’t melt, because when water mixes with lye (when the herbicide is being prepared), the reaction may give off a large quantity of heat.


This herbicide recipe was developed by Kru Pratoom, an organic farm educator in Chiang Mai, Thailand. Faced with a limited number of non-chemical herbicide choices, she made her own recipe using ingredients that were familiar to her, including pineapple, papaya, salt, and lye. Pineapple and papaya are used because they contain the enzymes bromelain and papain, respectively. Bromelain breaks down protein (Dubey et al., 2007), and papain disrupts a plant’s ability to photosynthesize (Itoh et al., 2013). Salt can remove the water in the leaves by changing the ion balance outside of the cells; additionally, the leaves absorb too much of the chloride, which is present in salt, and this also damages the leaves (Romero-Aranda & Syvertsen, 1996). Lye is highly alkaline; it is used to help dissolve the salt and can cause chemical degradation of weed leaves. One concern with this alternative herbicide is its highly alkaline nature, which might negatively affect the soil biota and pH. As far as pH is concerned, we would not recommend this herbicide if your soils are highly alkaline, but it may actually be helpful for very acidic soils. The soil pH at the ECHO Asia Seed Bank is typically around 4.8-5.2, so this highly alkaline herbicide may actually help to bring our soil pH into a more neutral range.

The Basic Herbicide Recipe:

1. 20 kg very ripe pineapple or papaya (with peels, cut into 3cm X 3cm chunks)

2. 10 kg salt

3. 1 kg lye (Sodium hydroxide - NaOH)

4. 20 L water


Table 1. Cost of herbicide ingredients.


Amount (kg)

TB (฿) Cost per kg

USD ($) Cost per kg

Total Cost in USD ($)






















Table 2. Final cost for a single 20L batch of the pineapple or papaya herbicides using the ingredients listed in table 1.

Herbicide Type

Total cost ($)

Cost per L ($)








Mixing Directions:

Add the fruit, salt, and water to a watertight plastic barrel. Slowly add the lye, carefully stirring between additions. Cover the barrel with a lid and allow the mixture to rest for 45 days (Figure 2). On day 45, strain out the pieces of fruit. The herbicide should be ready to use after straining. Keep the herbicide in the barrel until you are ready to mix it for an application. Do not add the water needed to prepare the herbicide spray if you aren’t planning to use it immediately. With the lid screwed on tightly, and if kept in the shade, the herbicide can be stored approximately 6 months.

Alternative Herbicide Article EAN #26 H2

Figure 2. Fermenting pineapple (left) and papaya (right).


1. Ideally, target weeds should be young and 8-10cm tall. If your weeds are more mature than that, cut the weeds with a weed whacker to a height of 8-10cm tall prior to spraying, to make them more susceptible to the herbicide;

2. Add 3 L of water into a backpack sprayer;

3. Stir the herbicide in the barrel with a stick for 30


Figure 3. Spraying experimental plots with the alternative herbicide.

Alternative Herbicide Article EAN #26 H3

seconds, and then transfer 1 L of the herbicide into the sprayer;

4. Mix the herbicide and water in the backpack sprayer; and,

5. Hold the nozzle of the sprayer approximately 30cm from the weeds. Spray the weed-infested area at a rate of 500ml/m2 . Our test plots were 2 m2 and required about 1 minute to spray.


1. An unused fallow plot at the ECHO Asia Seed Bank was identified as the experiment location—it contained a mix of overgrown broadleaf and grassy weed species. We used water for the control treatment. The the pineapple treatment was 1 part pineapple herbicide mixed with 3 parts water, and the papaya herbicide was 1 part papaya herbicide mixed with 3 parts water.

Alternative Herbicide Article EAN #26 H4

Figure 4. Plot sampling locations and seedling flat location. Xs represent soil sample locations. Circles represent buried filter paper. Triangles represent seedling flat.

2. Three days before spraying, the different types of weeds were identified and the percent composition was visually estimated, with weeds categorized as broadleaf or narrow leaf weeds.

3. The experimental area was divided into 1m X 2m plots. Treatments (control, papaya herbicide, and pineapple herbicide) were randomized with 4 replications, for a total of 12 experimental units (3 treatments [control, papaya, or pineapple] X 4 replications).

4. Immediately before spraying, soil samples were collected (we sent them to a local university lab to test pH and microbial content).

5. Weeds were cut down to 10cm using a weed eater.

6. Filter papers were weighed, placed in mesh nylon bags, and buried to a depth of 8cm in order to observe the effects of the herbicide spray on microbial health. The filter papers would normally be decomposed by

Alternative Herbicide Article EAN #26 H5

Figure 5. Seed flats with potting soil were placed in each plot to test the effect of the herbicide on seeds.

microorganisms in the soil; if the herbicide has a negative effect on the microorganisms, filter paper decay will be stopped or slowed.

7. In order to test the effect of the herbicide on seeds, seedling flats containing potting soil (but no seeds) were embedded into the middle of each 1m X 2m plot. A hole roughly the dimensions of the flat was dug to a depth of around 10cm. The flats were placed in the holes and dirt was packed around the flats so they were flush to the rest of the ground. Later, after completion of spraying, the flats were dug up and planted with seeds, to test the time to germination and the germination rate (Figure 5).

8. Each plot was sprayed with its assigned treatment—papaya herbicide, pineapple herbicide, or water. The weeds in each plot were sprayed for 1 minute at 500ml/ m2 every other day for one week.

9. The morning after a spray, treatment plots were visually rated for herbicide injury using a scale from 0-100, with a score of 0 having no injury and a score of 100 being full death. Each weed type, and species in a plot received its own damage injury score.

10.Three days after the final spray, additional soil samples were taken for soil pH and microbial testing to compare to the pre-trial samples. On this day, the germination flats were also collected and black bean seeds were planted 1 per chamber.

11. Seed germination was observed over 10 days. Mean time to 50% germination and percent germination for each treatment were used to evaluate the impacts of possible herbicide persistence in the soil on seed germination.

12.On day 10 after the final spray, the filter paper was retrieved and then weighed to find the amount of mass lost (a measurement of decay). These numbers were compared to the weight of the individual papers at the beginning of the study.


Alternative Herbicide Article EAN #26 H6

Figure 6. Herbicide injury visual damage score for broadleaf weeds over the course of a weeklong herbicide treatment. White indicates herbicide injury score for papaya, black is the injury score for pineapple, and grey is the injury score for the control. Letters indicate no significant difference between herbicide types (A) but a significant difference in herbicide injury compared to the control (B).

We found that the herbicide treatments significantly damaged the weeds within treated plots. We observed that both the papaya and the pineapple herbicide worked better on broadleaf weeds than narrow leaf (grassy) weeds (Figures 6 & 7). We did not observe a significant difference between the damage scores of weeds treated with the pineapple herbicide and weeds treated with the papaya herbicide; they were equally effective against weeds (Figure 8). Herbicide injury increased after each treatment, and after 4 treatments, the broadleaf damage was around 90% while narrow leaf damage was around 50% for both the papaya and the pineapple herbicides.

Alternative Herbicide Article EAN #26 H7

Figure 8. Damage to the weeds after 4 applications of the alternative herbicide treatments, photo taken on July 17th, 2015. Front row, left to right: Plot treated with pineapple herbicide, water control, papaya herbicide. Middle row, left to right: Plot treated with water control, pineapple herbicide, papaya herbicide. Bottom row, left to right: Plot treated with papaya herbicide, pineapple herbicide, and water control.

There was no measurable difference in

Alternative Herbicide Article EAN #26 H7

Figure 7. Herbicide injury visual damage score for grassy weeds over the course of a weeklong herbicide treatment. White indicates herbicide injury score for papaya, black is the injury score for pineapple, and grey is the injury score for the control. Letters indicate no significant difference between herbicide types (A) but a significant difference in herbicide injury compared to the control (B).

the soil pH before and after the spraying. The pH may have changed during the treatments but any difference in pH, if it existed, had washed out of the soil by the end of the experiment.

We had mixed results for our measures of soil microbial activity. Our filter paper decomposition test suggested that there was no difference in microorganism activity between soil treated with the herbicide and soil left untreated. In lab quantification of microbial populations, some of the treated plots experienced increased microbial populations and others experienced reduced microbial populations. The experiment was run twice but no consistent pattern was found. The decay rate of the paper suggested that treated soils were not consistently different from the control, although the papaya herbicide had the lowest decomposition rate (Figure 9). We

Alternative Herbicide Article EAN #26 H8

Figure 9. Filter paper decomposition 10 days after the final spray.

performed a number of serial dilutions to approximate soil microbial health ourselves (see ECHO Asia Note 15, “Soil Quality Assessment: Why and How” by Marcia Croft) but the microbial populations varied widely between the different plots before treatment and after treatment so there wasn’t a pattern in the change of soil microorganism populations. Soil samples submitted to Maejo University for microbial populations also varied widely. Differences between the initial treatment plots were so great, it was difficult to distinguish between what effects the herbicides had and what effect were from random variation in the soil. Further testing needs to be done to more accurately assess if the herbicide decreases or otherwise alters soil microbe populations, or if it has no effect.

Seedlings planted in the soil mix in the seedling flats which received the pineapple herbicide spray had the longest mean time to 50% germination, and soil treated with the papaya herbicide also took longer to reach 50% germination than the control treatment. However, the seeds in the pineapple herbicide treated soil also had the highest germination rate. The experiment was run twice; more seeds germinated in the second run when it had rained after many of the herbicide applications, suggesting that increased rainfall may wash the herbicide out of the soil more quickly and have fewer negative consequences for future crops and or soil microbial health.


The papaya and the pineapple herbicides both significantly damaged broadleaf and grassy weeds. We observed no obvious damage to the soil microorganisms; decomposition and colony counts were similar for plots treated with the water, pineapple herbicide, and papaya herbicide. Seeds planted in the soil of plots that had been treated with herbicide took longer to reach 50% germination, but the herbicide did not impact the overall germination rate. Rainfall may influence the persistence of the herbicide, because more seeds germinated in soil receiving the herbicide during frequent rains than in soil receiving herbicide applications during the dry season. Future studies should examine the long term impacts of this herbicide on the soil and on future crops. We would like to do further testing to ensure that the salt doesn’t stay in the soil, change the soil pH with long-term use, or damage the microbial populations of the soil, but the herbicide shows promise. We would also like to conduct trials using each of the herbicidal components, to figure out the active ingredients; if the lye isn’t that necessary, we would like to reduce the amount to make the herbicide safer to make and to use. We might also plant seeds and then apply the herbicide in order to examine if the herbicide could reduce weed competition without harming seedlings. Overall, it appears that this reduced-risk herbicide holds promise to help smallholder farmers to lower costs, reduce reliance on purchased inputs, and control weed pressure in their fields or gardens.

Literature Cited

Dubey, V. K., Pande, M., Singh, B. K., & Jagannadham, M. V. (2007). Papain-like proteases: Applications of their inhibitors. African Journal of Biotechnology, 6(9).

[FAO] Food and Agriculture Organization. (2015). Integrated Weed Management. Available: http://www.fao.org/agriculture/ crops/thematic-sitemap/theme/spi/scpihome/managing-ecosystems/integrated-weed-management/en/

Itoh, S., Aoki, K., Nakazato, M., Iwamoto, K., Shiraiwa, Y., Miyashita, H, Okuda, M., & Kobayashi, M. (2013). Novel Conversion of Chl a into Chl d Catalyzed by Grated Vegetables. In Photosynthesis Research for Food, Fuel and the Future (pp. 804-807). Springer Berlin Heidelberg.

O’Donovan, J. T., Harker, K. N., Clayton, G. W., Newman, J. C., Robinson, D., & Hall, L. M. (2009). Barley seeding rate influences the effects of variable herbicide rates on wild oat. Weed Science, 49(6), 746-754.

Romero-Aranda, R. & Syvertsen, J. P. (1996). The influence of foliar-applied urea nitrogen and saline solutions on net gas exchange of Citrus leaves. Journal of the American Society for Horticultural Science, 121(3), 501-506.

Swanton, C. J., Shrestha, A., Clements, D. R., Booth, B. D., & Chandler, K. (2009). Evaluation of alternative weed management systems in a modified no-tillage corn–soybean–winter wheat rotation: weed densities, crop yield, and economics. Weed Science, 50 (4), 504-511.