Survival of Taro: Agronomic and Pathological Research For Sustainable Production

Objective 1: Characterization of the undescribed Phytophthora.

Efforts to characterize the new homothallic Phytophthora species isolated from pocket rots of taro (Colocasia esculenta) are continuing. Great difficulty has been encountered with this pathogen and simple procedures, such as the establishment of single hyphal tip cultures has been problematic. Single hyphal tip cultures are needed to insure the genetic purity of each isolate before characterization can begin. Single tips of other Phytophthora species, which are less than 1 mm grow well. However, for this new Phytophthora species, tips that were l mm long did not grow. Several types of agar media were also used to encourage tips to grow. Finally, a combination of several methods was employed to establish single hyphal tip cultures. These included growing the parent culture on thin agar that restricted mycelial growth to a single layer, transferring long tips/branches (2 mm+) to water agar with sterilized rye seeds and placing tips adjacent to the seed. The Phytophthora tips grew slowly but were reestablished by using these methods. Single tips of the three populations of the new Phytophthora were prepared and characterizations were begun. Preliminary measurements taken of the parental cultures demonstrated no difference among the isolates for length or diameter for sporangia produced on 10% vegetable juice agar (V-8).
The new Phytophthora isolated from pocket rots was characterized as follows: Colony growth at 24 C was optimal and temperatures below 20 and above 28 reduced the growth rate. No growth occurred below 16 C or at 31 C.
Hyphal diameter was variable and primarily ranged from 4.5 to 9 microns on vegetable juice (V8) agar culture at 24 C. All isolates were homothallic (self-fertile) and formed amphigyous antheridia that were 16.2 microns in width and 16.8 microns in length. Oogonia were 36.6 microns in width and oospores were 30.9 microns wide. Oospore walls were thickened, averaging about 2.5 microns wide. In general, aborted oospores had walls that were 5 microns wide. The frequency of aborted oospores varied among the isolates, ranging from about 25% to 70%. However, a change in cultural media may improve this. Chlamydospores were not observed in host tissue or in any of the cultures, in agar or in liquid medium. Small vesicles occasionally formed in liquid culture.
Morphology of sporangia has been variable and measurements have not been made. In liquid culture, long sporangia, semi to distinctly papillate are produced. On agar, sporangia are few and tend to be shorter and papillate. Different sequences of growth on agar followed by placement in liquid were tested. Potato Dextrose agar, 2 to 40% V8 agar, low rates of agar and host tissue were used, but sporangial morphology continued to be contorted, or were few to rare.
Characterization of this undescribed species will continue.

Objective 2: Field Tests

Ka'ala Program Site #1:
The plan for the second test at the Kaala Educational Center was placed on hold as the Center did not receive funds to continue beyond 2003. Most of the staff was released and only one person and one part-time person is at the Center to maintain it. Tours for youngsters are continuing. In the first experiment at Ka'ala, a plot at the Center was fallowed and cover cropped with Sunn Hemp, a legume that increases the nitrogen level in soil. The crop was cut and allowed to decompose. Blood meal, was added to increase the amount of nitrogen needed for the crop. A control plot, that was not fallowed, had no cover crop grown in the paddy, but received blood meal to equalize the nitrogen level. The field with the legume cover crop yielded very high survival rates (80%) of planted taro huli (young plants) and total biomass produced was very high (370 kg), but overall yield was poor after a year, possibly because of the frequent addition of blood meal to the paddy. Only about 32% plants survived in the control plot and biomass was very low (80 kg). Progress: Although a second experiment was not installed at the Educational Center, the plot of land where the cover crop was grown and tilled into the soil still contained a lot of nutrients after the original taro crop was harvested. The remaining nutrients are from the residual organic matter and the unused organic fertilizers that had been added for the first taro crop. This plot was replanted and after six months we visited the site. The taro crop was growing very well. After 12 months, the crop was harvested and a high yield was reported by the grower. Corms were large and corm rots were very few. Blood meal was only added at planting. The growers at Kaala are very pleased although we were not able to quantify this harvest. This planting was not part of our experiment and the second experiment was designed for an area adjacent to the paddy used in the first experiment.

After receiving news that the Kaala project did not receive funds for 2004, Drs. Janice Uchida and James Silva met with the leaders of the Kaala Educational Center, the field managers, growers and other personnel including volunteers. We reviewed the entire project and what we had learned. They were anxious for us to return in the future to build on what we had learned in the first test. Beyond the scientific information that we gathered at this site, it was even more important for this group that we made a tremendous impact demonstrating that faculty from the University was willing to travel on numerous occasions to Waianae and make the effort to share our knowledge with them. We built this project as a partnership, that each of us had things to test and learn. An important bond has been built with the members of the Waianae community, an area with high crime, low educational levels, high numbers of food-insecure residents and a needy social population with native Hawaiians in the ethnic mix. They will continue to use some of the results found (eg. using Blood meal only before planting; using green manure as opposed to dried grass; appropriate nitrogen levels).

Haleiwa site #2:
A taro paddy was drained, tilled, and planted with Sunn Hemp as a cover crop, and another adjacent paddy was planted with sections of Sunn Hemp, Clover, and Alfalfa. Overall, Sunn Hemp grew quite well but was not as tall as in other sites, averaging about 4-5 ft at flowering. A few Sunn Hemp plants wilted and isolations revealed a Pythium tentatively identified as P. carolinianum. This Pythium species has not been pathogenic to taro. Sections of the cover crop in each paddy grew poorly as the Haleiwa grower struggled without an effective method to water these crops. When the fields are constructed, drainage is deliberately made poor allowing the taro crop to be grown partially submerged in water. Normally, the taro paddy is flooded and the water slowly drains through the bottom of the paddy. This is different from other taro farms that drain the water from the paddy. For the cover crops, the water was left on overnight at a low flow rate but this was excessive and plants grew poorly. Eventually the water was reduced to only a few hours late in the day but growth was not greatly improved. Soil analysis did not reveal high salt levels, excessive nutrients or nutrient deficiencies. Seeds that accidentally landed on the banks and walkways grew very well and were more than 3 times the size of plants in the paddy. The level of nodulation was good with Sunn Hemp but poor for alfalfa and clover growing in the paddy and on the banks. After 6 weeks, the cover crops were cut and tilled into the soil. At least 4 weeks were allowed for decomposition of the organic matter. The paddies were then formed, flooded, and planted with taro cuttings.
To obtain better growth of clover and alfalfa, in the future we want to grow these on the top of furrows to improve the drainage. The banks were only about 12 inches or less high and growth on these raised sections were good.
The original experimental design had the control paddy planted with taro until it was time to plant the new crop. Growers generally allow only a month between crops. The paddy adjacent to the cover crop paddies was selected as the control paddy and planted with taro. The taro crop was 4 months old when the Sunn hemp in the fallow paddy was cut and worked into the soil. Taro growth was extremely good (and is frequently good for that time of year and that stage of growth) and these plants were to serve as the source of cuttings for plants needed to plant the fallowed paddy and to replant the control paddy. The entire design and the schedule of treatments had been discussed twice with the owner. Unfortunately, when the owner saw the great growth of the taro in the control plot, he did not want to sacrifice this crop that was growing and instead provided another field in a different area to serve as a “control”. The field manager and county agent tried to reason with him but the concept of the control and its value could not be impressed upon him. He is a successful entrepreneur and well known for his obstinacy. Thus, although the best location for the control would have been in the nearby paddy, this could not be obtained. We were informed of this development after the fields had been planted.
The crop is now almost 7 months old and will be harvested in 5 to 6 months. The weather in the past 6 weeks has been extremely wet and high levels of the leaf blight pathogen, Phytophthora colocasiae, are present. The dew that persists on the leaves until well past 11 am, maintains moisture on the leaves and the humidity in the canopy is high. Sporangial (spore) production on every lesion is extremely high and is easily distributed by air movement. All fields were severely blighted and this is not unusual for this time of year.
Four samples of plants were obtained randomly from the fallowed paddy, the control paddy #1, that was assigned to us, and a second control paddy #2, adjacent to the fallowed field. The second control paddy # 2 was 4 months older but was needed since control paddy # 1 assigned to us had a different soil type and different microenvironment. Each sample consisted of a large corm and several new cormels. Disease levels and biomass produced were as follows:
*about 32-35 cormels per treatment. The corms also had different levels of pocket rot but data is not significant because only 4 corms were harvested per treatment. Root development in all treatments was good.
Thus disease levels were lower in the fallowed paddy but yield was higher in Control paddy # 1.
According to the field manager, the paddy in which the Control # 1 is growing consistently has very high yields and is one of the best fields on the farm. He believes it is a poor comparison. Control # 2 is in the same soil series as the Fallow paddy but it is 4 months older, therefore biomass is larger. Thus, the strong growth in the paddy with Control # 1 is evident as the biomass is already the same or exceeding the biomass in a paddy that is four months older (Control # 2).
We find this change in the experimental design to be highly unfortunate but that is the disadvantage of working with growers. This grower however, began as an individual who had an extremely low opinion of the University and has now become a strong supporter of the College of Tropical Agriculture and Human Resources.
On March 5, the average biomass data show (Table 1 and Fallow Test Histogram) that there is little difference between the Control #1 and the Fallowed plot, although Control #1 is in a better environment. Both plots are the same age. This plot had the highest biomass in the first harvest. Control #2 is 4 months older and plants are showing increased corm weigh with the average biomass significantly heavier than the other plots.
Although we had problems with the Control plots assigned to us by the grower (discussed on the Dec 2003 report), this test will be harvested in 4 months.
In March, about 9.5 months after taro was planted, plots were sampled. Three plants were harvested from two sections of the plot, for a total of 6 plants per plot. The evaluations were for the fallowed plot, the control #1 that is in another area of the field and the control #2 which is adjacent to the fallowed plot but 4 months older.
Disease levels on roots and corm were none to very low and no data was collected. A few root rots were found and Pythium was isolated but in pathogenicity tests, these isolates were not pathogenic. Pythium cultures were placed directly on healthy root systems and no infection occurred after a month. Pythium pathogens will kill roots in less than a week.
* Average is for the 6 original plants and the cormels that formed from the mother corm. The number of cormels varied from 6 to 17, were mostly 9 to 11 and there were no significant difference between the treatments for number of cormels produced.

Waihole Site #3:
Most recently we secured an agreement with a family in the Waihole area, a traditional ecosystem for taro growing. It is located in a cooler area of Oahu and has adequate water supply. The family has been growing taro for two generations and also sells their taro and taro products. It is a small but very successful business.
This experiment was initiated in October as an on-farm trial to determine N availability from a fish bone meal waste product (9.33 %total N) and benefits of cover cropping with sunn hemp.
Sunn hemp was planted as the cover crop and 4 levels of blood meal (0, 50, 100 and 200 lbs per 1,000 sq ft of blood meal) were added. The Sunn hemp was grown, cut, in incorporated into the soil and allowed to decompose. The field was tilled and sectioned with plastic barriers to separate treatments. This method has worked well in several other experiments. The blood meal was added to each section and volunteers helped to work the material into the paddy. The water level was kept low to avoid losing the blood meal. After a month, the field was planted with young plants. Because the area is cool, there is little taro growth from November to March. At this time, almost all plants have survived but growth is very slow.
Dr. Jonathan Deenik, who was recently hired in the Tropical Plant and Soil Science Dept, has joined our program. He will replace Dr. Silva and is gradually learning about the tests we are conducting. At the Waihole site, he is measuring and tracking the level of ammonia and nitrates in the taro paddy. This intensive type of tracking has not been done for taro culture and is timely as the fate of nitrogen is so critical. Our Maui site, selected to monitor nitrogen levels is still on hold, as the county agent was injured recently. However, the resources for the Maui test are being used at Waihole and before the end of this SARE project, we hope to have some soil and water analyses completed to determine the movement of nitrogen in the taro paddy.
The nitrogen was added to the to paddy as complex organic nitrogen (primarily proteins) and forms ammonia in the soil. In an aerobic soil environment, microbial action converts the ammonia into nitrates. To prevent this conversion, the paddy is kept flooded and maintained in an anaerobic state. Soil samples were taken at weekly intervals for the first 3 weeks and analyzed for NH4 –N (nitrogen as ammonium) and NO3-N (nitrogen as nitrate). Six handfuls of soil was taken from across the plot and mixed in a bucket for a composite sample for each plot. Taro shoots were planted after 4 weeks. After 3 weeks, soil samples were collected every 2 weeks for 15 weeks. After 15 weeks, 2 taro plants from each plot were removed and fresh and dry weights were determined. Tissue was analyzed for nutrients.
Nitrogen mineralization and availability of ammonium:
There was rapid nitrogen mineralization in the first 2 weeks after incorporation of the organic material (Fig. 1). Ammonium production achieved a maximum for all rates after two weeks. There was significantly more ammonium production at the highest rate (10.4 tons/ha) whereas at the two intermediate rates, ammonium levels in the soil did not differ significantly. Ammonium levels in the 0 plot increased slightly from 17 to 33 mg/kg during the first 2 weeks and then fluctuated between 15 and 37 mg/kg through the 19th week. Ammonium levels decreased in all rates except the 0 rate from the 15th to the 19th week. This represents the use of the ammonium by the growing plant. Thus, soil ammonium levels decreased as the observable plant biomass increased.
Nitrate production was very low in all of the plots (Fig. 2) as the flooded condition inhibits the conversion of ammonium to nitrates. Nitrogen in the water leaving each plot was sampled immediately before planting and after 13 weeks. The highest level of nitrate, 3.7ppm, leaving the plot was for the 10.4 tons/ha treatment after 3 weeks (Table 3). The analysis shows that the organic amendments increase the nitrates slightly in the water in the first 3 weeks after it was incorporated. All plots should very slight increases in nitrate levels but all are very low. By the 13th week the nitrate levels in the water for all plots is not significant.
Fig. 1. Ammonium production from fish-bone meal added to a flooded taro soil. Data represent a mean from two plots and bars are the standard deviation. Treatments are tons/ha.
Figure 2. Nitrate production from fish-bone meal added to a flooded taro soil. Data represent a mean from two plots and bars are the standard deviation. Treatments are tons per ha.
* Surface water from the mountains that enters each plot separately.

This data is highly encouraging as it confirms that organic nitrogen is retained in the anaerobic soil and slowly released as ammonium that is used by the plant. This confirms the value of employing organic amendments in the tropics.

Each of the plots was sampled by removing two plants per treatment. The total biomass (fresh weight) was 152, 125, 232, and 284 grams for the 0, 50, 100, and 200 lbs fish-bone meal per 1000 sq ft respectively. Thus good growth is occurring for the highest level of fish-bone meal. There was no statistical difference between the 0 and 50 treatments, but the 100 and 200 treatments were statistically different from the 50 lbs treatment. For the dry weight data, the 200 lbs level was statistically higher in biomass than the 0 and 50 lbs.

This experiment will be continued for at least 6 to 7 more months. We will continue to monitor soil ammonia levels and will evaluate plant growth in April and July.

Laboratory test:
In the laboratory, a soil incubation experiment is in progress to determine the availability of nitrogen from different organic amendments. The amendments were: fish-bone meal, chicken manure, horse “manure” in straw bedding, horse “manure” in soil, Java plum leaves (Eugenia cuminii), Hau leaves (Hibiscus tiliaceus), fern leaves, sunn hemp (Crotolaria juncea), and soil only. Amendments were mixed with soil, covered with 3” of soil and then submerged in water. The level of ammonia has increased for the sunn hemp, chicken manure, fish-bone meal and the two horse manures. Thus, although the other plants are common in the Hawaiian environment, they were not useful as a source of ammonia or nitrogen. Other plants will be tested.

New Hanalei Experiment:
A major reason for the reluctance of growers to use the cover crop and fallow period is that it decreases the time for crops to grow. But, faced with continued pathogen problems, some are beginning to accept the idea that at the very least, organic matter or compost should be added to the paddy soil and allowed to decompose for a month or more. This is an investment of 2-3 months to improve soil health. We are reporting some early findings as this project made these compost tests possible.

The source of good organic matter or compost of predictable quality is often the biggest factor for small island communities. An economical, local compost source was identified. Before field application, compost quality was evaluated by testing for the amounts of available nutrients, pH, salts, etc. After several tests, a local source of compost was identified.

Objective 3: Plant Spacing
At this time most growers are using the 24” spacing between plants. Thus, over the past two years many more growers have increased the spacing in their fields. Yields are nearly the same as for the 18” spacing between plants yet cost of planting (labor), cost of new plants, and time (labor) for harvest are all reduced by converting to a 24”interplant distance. More growers are aware that planting young plants with any blemish (small rots) will decrease survival of these plants. At this time growers have reduced the amount of nitrogen used from nearly 700 labs per acre to about 350 lbs per acre. They also see the value of testing the soil before planting to determine nutrient status and how it can be corrected.