Annual artemisia as a high-value crop and for weed control

Annual artemisia as a high-value crop and for weed control

Summary

Objectives/Performance Targets

Artemisia annua is a medicinal herb used extensively in the treatment of malaria[i], and there are indications that it may have wider benefits for other health concerns. The active ingredient, artemisinin, is derived from the leaves of A. annua and is reported to be allelopathic as well as providing a medicinal benefit. Artemisinin is already in short supply, and if additional medicinal uses are developed there will be increased demand. This project will provide information on the possibility of A. annua as a new high value crop that can double as a cover crop with some weed control efficacy. The project will evaluate the productivity of Artemisia annua in order to provide yield estimates. It will also evaluate the effectiveness of the crop in weed suppression.

Although we completed most of the work on this project in 2012, there were some unexpected results so we decided to repeat some of the work at a reduced level in 2013. There were reports in the literature that a dual harvest would increase productivity of artemisinin, the anti-malarial compound found in A. annua. However, our results showed no such effect, and yields in this study were not statistically different between the single harvest plots and the double harvest plots. Also, in the second year we expanded the scope of the study by using different planting densities to determine the effect on biomass productivity.

[i] Elfawal MA, Towler MJ, Reich NG, Golenbock D, Weathers PJ, et al. (2012) Dried Whole Plant Artemisia annua as an Antimalarial Therapy. PLoS ONE 7(12): e52746. doi:10.1371/journal.pone.0052746 http://dx.plos.org/10.1371/journal.pone.0052746

Accomplishments/Milestones

We had planned to use the other side of the field in which the A. annua was planted in 2012, but we were unable to produce a usable seedbed due to heavy rains just prior to planting time. We decided to use an adjacent field, with plots about the same distance from the 2012 planting as the area which we had planned to use. The replacement field was slightly higher and drier so we were able to produce a good seedbed for planting. We did not add any fertilizer or bone meal to this plot, nor did we install a drip line since the 2012 planting did not really require water beyond normal rainfall.

The SAM cultivar was again produced by Professor Weathers at WPI. The planting density in 2012 was 12 plants per square meter. In 2013 we tried 10/m², 8/m², and 4/m². The seedbed was prepared June 6^th, and flame weeded immediately prior to setting out the plants on June 12^th. The 1 square meter plots were separated by wooden stakes as in the 2012 planting, but we did not leave a buffer between the plots. We did leave a small buffer between different density zones, just for easier identification of the zones.

Half-season harvest was done July 17^th, but instead of harvesting 1 m² blocks, the harvest was done by pruning every other plant to half height. As before, the plants were labeled at harvest so that the different density planting zones could be compared. These plants were taken to WPI and dried in the laboratory, where the leaves were subsequently removed, sieved, and stored for later analysis.

Final harvest was done September 21st, when the plants were just starting to develop flower buds. These plants were bagged, dried in the greenhouse, then transported to the laboratory for processing as drying progressed.

In late June Professor Weathers decided it was time to do the mid-season harvest. On July 2 she brought some of her students to the field and cut the plants at about one foot height. The cuttings were bagged and labeled and taken to Worcester Polytechnic Institute (WPI) for drying and analysis. Also, at this time, a string was placed on the stakes to delineate the weed test plots and the weeds were pulled, separated into monocots and dicots, bagged, labeled, and taken to WPI for drying and weighing. Weeds were pulled if the root was within the area defined by the string. The monocots and dicots were counted separately, so there are two measures of weed growth for the test plots.

After the first harvest, the plots were allowed to proceed without intervention. In September, the plots were monitored for flower bud development, signaling the appropriate time for the final harvest. Around mid September the flower buds started to appear. By this time the plants in the single-harvest plots had grown to a height over seven feet. The perimeter of the test plots had grown up to weeds by this time so a rough mowing was done to remove the worst of the weeds and improve access to the test plots. On September 18-19 a wind storm came through which bent over some of the taller plants. There was no damage to the plants, but it mixed the branches of the test plants and the weeds. On September 20, the weeds immediately adjacent to the test plots were removed and baler twine was run around the plants to pull the outer branches into the plot and make it possible to separate the weeds and the crop. Having cleared the area next to the plants, the weed suppression efficacy of the Artemisia annua could be seen by the lack of weed growth under the canopy. On September 23 the plants were harvested. Since these plants are destined for laboratory analysis rather than a bulk sale, the harvest is different from the common cover crop harvest involving just cutting and packing. Individual plants are cut at the base and placed into a bag. This keeps the leaves from getting lost in transport since the leaves are the part of the plant that contains most of the artemisinin. Care is taken during harvest to ensure the plants are clean so that the extracted artemisinin can be used for clinical trials if necessary. The volume of bagged plants was much higher than the mid season harvest, so they were taken to a local greenhouse, where they were hung to dry. The plants need light during the drying process to fully develop the production of artemisinin. The bags are perforated to allow air exchange for drying.

Impacts and Contributions/Outcomes

Combining data from 2012 and 2013 work we found that planting densities of 8 plants/m² yielded the highest biomass per plant. Yield dropped off after 8/m², with 12 plants/m² giving the lowest yield per plant. The 4 plants/m² zones gave yields per plant similar to 8/m², but the total biomass per acre was significantly reduced because of the lower number of plants (Table 1). Funding was insufficient to complete analysis of artemisinin content for the 2013 harvest, but that was not within the expected scope of this project.

Part of the study was to determine the value of A. annua as a cover crop. As far as weed suppression is concerned, there was no evidence for any allelopathy in this study. However, weed suppression was considerable by shading, for all densities tried in this study.

The value of A. annua as a crop (other than pure weed suppression) is based on the quantity of artemisinin that can be derived from the crop. The global value of the dried leaves of A. annua for the extraction market varies considerably. At this point we should point out that the SAM cultivar used in this study was selected for consistently high artemisinin content. Moreover, this cultivar is currently vegetatively propagated. Work is ongoing to develop a cultivar that can be produced from seed and still maintain consistently high artemisinin content.

We did not see any significant allelopathic effect in the test, but since there were reports of allelopathy when growing artemesia we decided to check whether there was any residual allelopathic effect after the plants were removed but the roots remained in the field, a situation likely to occur after harvest of a cover crop. About 2 days after the plants were removed in the final harvest four of the plots were cleared by flame weeding (which killed the few emerging weeds and eliminated debris remaining from harvest) and radish seeds were planted. The selected plots were one from the untreated fallow group, one from the single harvest group, one from the two-harvest group and one from the leaf-amended group. To check for allelopathy, we counted out radish seeds so we could check the germination rate. Twenty-five seeds were placed on a strip of paper masking tape at roughly equal intervals of approximately 3/4 inch. The tape strips were folded so the seeds would not fall off and three tapes were planted in each plot. The plots were observed every couple of days and after two weeks the number of emergent plants were recorded in each plot. Germination rates for the plots were about 63% for the two harvested plots, 63% for the leaf amended plot, and 70% for the untreated soil. Germination rates for a tray seeded in the greenhouse was about 90%. Statistical tolerances are about 12% for the field test and 7% for the greenhouse test. That means that there is no statistical difference between the germination rates in the field, but there is a difference between the germination in the field and in the greenhouse. This could be caused by warmer temperatures in the greenhouse. It is interesting to note that emergence of radish plants was delayed by about a week in the leaf amended plot compared to the harvested and untreated plots. This is consistent with unpublished observations of delayed germination in tests by Professor Weathers at WPI.

• Leaf to stem biomass ratio of the 3 harvest groups. Different letters indicate statistical groups at significance P less than 0.01
• Comparative yield of leaves, stems, and total biomass from all harvests. Different letters show statistically different data for P = 0.01 for 2H1+2H2 vs 1H1 as follows: a,b? stems c,d? leaves e,f? leaves + stems

Collaborators: