Comparison of the damage potential and yield loss of the rice root-knot nematode, Meloidogyne graminicola, on lowland and upland rice varieties from Myanmar Текст научной статьи по специальности «Сельское хозяйство, лесное хозяйство, рыбное хозяйство»

Russian Journal of Nematology, 2015, 23 (1), 53 - 72

Comparison of the damage potential and yield loss of the rice root-knot nematode, Meloidogyne graminicola, on lowland and upland rice varieties

from Myanmar

Pa Pa Win1, Pyone Pyone Kyi1, Zin Thu Zar Maung2, Yi Yi Myint3 and

Dirk De Waele4, 5' 6

1 Plant Protection Division, Department of Agriculture, Ministry of Agriculture and Irrigation, Bayint Naung Road,

West Gyogone, P.O. Box 1011, Insein, Yangon, Myanmar 2 Agricultural Nematology Laboratory, 256 Giltner Hall, Michigan State University, 48823, Michigan, USA 3 Department of Plant Pathology, Yezin Agricultural University, Yezin, Myanmar 4 Laboratory of Tropical Crop Improvement, Department of Biosystems, Faculty of Bioscience Engineering, University of Leuven (KU Leuven), Willem de Croylaan 42, Heverlee, B-3001, Leuven, Belgium 5 Crop and Environmental Sciences Division, International Rice Research Institute (IRRI), DAPO Box 7777,

Metro Manila, Philippines 6 Unit for Environmental Sciences and Management, North-West University, Private Bag X6001, 2520,

Potchefstroom, South Africa e-mail: dirkdewaele@pandora.be

Accepted for publication 16 May 2015

Summary. In Asia, the rice root-knot nematode, Meloidogyne graminicola, is an important pathogen of Asian rice (Oryza sativa) in most rice producing countries including Myanmar. In the first part of our study, the damage potential of M. graminicola on commonly cultivated newly released high-yielding lowland rice varieties and upland rice varieties (traditional, improved and aerobic), which are being grown in different regions in the summer-irrigated lowland and rainfed upland rice ecosystems in Myanmar, was investigated under screenhouse conditions. In the second part of our study, a field experiment was carried out in the Ayeyarwady River Delta, the major lowland rice producing area of Myanmar, to investigate the impact of M. graminicola on plant growth and yield of the same lowland rice varieties included in the screenhouse experiment in a naturally-infested farmer's field. None of the 15 lowland varieties and none of the nine upland varieties included in the screenhouse experiments was resistant to M. graminicola. Although differences in susceptibility were observed among the lowland and upland varieties, and between the two types (lowland and upland) of rice varieties, with an Mfeggs (multiplication factor of the second-stage juveniles without counting the eggs) ranging from 13.2 to 52.8 for the lowland varieties vs 39.8 to 108.4 for the upland varieties, all varieties included in our study can be considered as highly susceptible to M. graminicola when the nematode population densities are assessed at harvest. Also in the field experiment, all the 15 lowland varieties were susceptible to M. graminicola infection. This part of our study shows that upland rice varieties are more susceptible to M graminicola infection than lowland rice varieties. The percentage reduction in lowland and upland varieties was very similar in six out of the ten plant growth and yield-contributing traits measured. The highest differences in percentage reduction were observed for fresh root weight (41.0 vs 26.1% for the lowland and upland varieties, respectively), number of tillers per plant (29.1 vs 14.1%, respectively), percentage filled grains per plant (11.7 vs 0.8%, respectively) and filled grain weight per plant (34.7 vs 47.6 g, respectively). Within the two types of rice varieties significant differences in percentage reduction of plant growth traits between uninoculated and inoculated plants were observed among the rice varieties, so we suggest that the effect of M. graminicola infection on the different vegetative and reproductive plant growth stages of rice varieties is highly genotype-dependent and that no general conclusions can be made. In the screenhouse experiments, infection with M. graminicola caused on average a yield reduction with 31.1% in the lowland rice varieties vs 44.9% in the upland rice varieties, which indicates that in the upland varieties the higher nematode population densities per root unit (1 g) and per root system resulted in a higher yield loss compared with the lowland varieties. In the field experiment, carbofuran treatment resulted, on average, in a 16.5% increase in yield of the lowland rice varieties.

Key words: lowland rice, multiplication factor, plant growth traits, resistance, root galling, tolerance, upland rice, yield-contributing traits, yield loss.

Plant-parasitic nematodes are one of the limiting factors of the production of rice in the tropics (Bridge et al, 2005). Quantification of the production constraints of a rice crop caused by a specific nematode species is essential to develop efficient and, by preference, integrated management strategies (Ramakrishna & Sharma, 1998). Yield losses caused by nematodes are influenced by many factors including the susceptibility and sensitivity of the crop, the pathogenicity of the nematode species (including its reproductive fitness and virulence), the nematode population density at planting and environmental factors (Bridge et al., 2005). Following the terminology of Bos & Parlevliet (1995), resistance/susceptibility on the one hand and tolerance/sensitivity on the other hand are defined as independent, relative qualities of a host plant based on comparison between genotypes. A host plant may either suppress (resistance) or allow (susceptibility) nematode development and reproduction; it may suffer either little injury (tolerance), even when heavily infected with nematodes, or considerable injury (sensitivity), even when relatively lightly infected with nematodes. Resistance/susceptibility can be determined by measuring the nematode population densities in and on the roots, whereas tolerance/sensitivity can be determined by measuring the effect of the nematode population on plant growth, yield-contributing traits and/or yield.

In Asia, the rice root-knot nematode Meloidogyne graminicola Golden & Birchfield, 1965, is an important pathogen of Asian rice (Oryza sativa L.) in most rice producing countries (Jain et al., 2012). It is equally prevalent on lowland (irrigated), upland (rainfed) rice as on deepwater rice (Bridge et al., 2005). In shallow intermittently flooded lowland and rainfed upland rice ecosystems, M. graminicola is considered to be by far the most damaging Meloidogyne species on rice (De Waele & Elsen, 2007). However, this nematode species is also well adapted to flooded conditions (Bridge & Page, 1982) and is considered a threat to the production of tropical aerobic rice (De Waele & Elsen, 2007; Kreye et al, 2009a, b, c).

In a M. graminicola-infested rainfed lowland rice field in Bangladesh, nematicide application resulted in a yield increase of 16 to 20% or about 1 t ha-1 (Padgham et al., 2004). In M. graminicola-infested upland rice fields in Thailand and Indonesia, nematicide application resulted in a yield increase of 12 to 33% (Arayarungsarit, 1987) and 28 to 87% (Netscher & Erlan, 1993), respectively. In glasshouse studies, Soriano et al. (2000) reported that yield losses due to M. graminicola ranged from

11 to 73% in simulated intermittently flooded lowland conditions, while in simulated upland conditions yield losses ranged from 20 to 80% (Plowright & Bridge, 1990; Prot & Matias, 1995; Tandingan et al., 1996).

Based on the occurrence of M. graminicola in Myanmar (Win et al, 2011), this nematode was considered the most important nematode species attacking all ten lowland rice varieties examined in the summer-irrigated lowland rice ecosystem. Differences in host response (varying from less susceptible to highly susceptible) to M. graminicola infection at an early plant growth stage (stem elongation) were observed among 15 lowland and nine upland rice varieties (Win et al, 2014). These differences may translate in differences in yield. Hence, it is necessary to examine the effects of M. graminicola infection on yield at plant maturity (harvest). This host response evaluation at maturity may also lead to the identification of less sensitive or tolerant rice varieties and this may offer an interesting alternative to the use of less susceptible or resistant rice varieties to limit yield losses caused by M. graminicola. Therefore, in the first part of our study, the damage potential of M. graminicola infection on commonly cultivated, newly released, high-yielding lowland and upland rice varieties (traditional, improved and aerobic), which are being grown in different regions in the summer-irrigated lowland and rainfed upland rice ecosystems in Myanmar, was investigated under screenhouse conditions.

It is essential to demonstrate the damage and yield loss potential of the predominant nematode species of an agricultural crop under the most common production conditions (De Waele & Elsen, 2007). Validation of screenhouse yield loss experiments by field experiments is essential, also to understand how nematode population densities and yield losses may change under natural environmental conditions (Nombela & Romero, 1999; Taylor et al., 2000). Therefore, in the second part of our study, a field experiment was undertaken in the Ayeyarwady River Delta, the major lowland rice producing area of Myanmar, to investigate the impact of M. graminicola on plant growth and yield-contributing traits, and yield in a naturally-infested farmer's field of the same lowland rice varieties included in the screenhouse experiment.

MATERIALS AND METHODS

Three experiments were carried out during 2010 and 2011. From January until April 2010 (i.e., during the summer-irrigated rice growing season), a

first screenhouse experiment including 15 lowland rice varieties was conducted at the campus of the Plant Protection Division, Yangon. At the same time, a field experiment including the same 15 lowland rice varieties was carried out in the summer-irrigated lowland rice ecosystem. From June until October 2011 (i.e., during the rainfed (monsoon) rice growing season), a second screenhouse experiment including nine upland rice varieties was conducted also at the campus of the Plant Protection Division, Yangon. The lowland rice variety Thihtatyin was included in all three experiments as a reference variety.

Preparation of plants. Seeds of 15 high-yielding lowland varieties were obtained from the Myanmar Rice Research Centre, Hmawbi, and from the Rice Division, Department of Agricultural Research (DAR), Yezin, Nay Pyi Taw (Table 1). Seeds of nine upland varieties, including not only traditional but also one improved and one aerobic variety, were obtained from the Aung Ban Research Farm in Shan State (Table 1).

For the screenhouse experiments, the seeds were first soaked in water overnight and germinated on wet paper in Petri dishes at room temperature. For the field experiment, the seeds were placed in perforated plastic bags and soaked in water inside a drum overnight. Then, the seeds were germinated in these bags at room temperature for 3 days.

Nematode inoculation, treatments and experimental set-up of the screenhouse experiments. The nematode inoculum used in the screenhouse experiments consisted of the offspring of a single M. graminicola female isolated from a rice plant (variety unknown) in Pathein region, Ayeyarwady Delta (Lower Myanmar) and multiplied on the rice variety Thihtatyin under upland conditions in a so-called sick plot at the campus of the Plant Protection Division, Yangon. The second-stage juveniles (J2) were extracted from the rice roots using the tray method (Whitehead & Hemming, 1965). Only freshly extracted J2 (i.e., collected during a 24 h period) were used as inoculum.

Three-day-old pre-germinated seeds were planted singly into 17-cm-diam. pots containing 1,500 ml of sterilised soil. The soil in the pots was saturated (100% of soil pore volume filled with water) at planting. The lowland varieties were grown in a clay loam soil (42% clay, 25% loam, 32% sand) containing 0.18% nitrogen, 24.3 ppm P, 9.2 mg (100 g)-1 K2O at 5.8 pH. The upland varieties were grown in a sandy loam soil (10% clay, 13% loam, 75% sand) containing 0.16% nitrogen, 26.9 ppm P, 9.7 mg (100 g)-1 K2O at 5.6

pH. One week after planting the pre-germinated seeds in the saturated soil, six plants of each variety were inoculated with 3,000 M. graminicola J2 per plant by pipetting three aliquots of the same volume in three 5-cm-deep holes around the base of the seedlings. Six plants of each variety were not inoculated and acted as control plants. At the time of inoculation, the soil in the pots was still saturated. The pots were placed in a screenhouse at an air temperature ranging from 26 to 38°C and from 28 to 33°C (during the summer and monsoon season, respectively) according to a randomised complete block design (RCB). For the screenhouse experiment with the lowland varieties, flooding started 6 days after inoculation of the seedlings. Thereafter, the plants were intermittently flooded three times per week to simulate the same water regime as applied in the summer-irrigated lowland rice farmer's fields. After flooding, there was a small layer of standing water in the pots. For the screenhouse experiment with the upland varieties, the soil in the pots was maintained at field capacity (50% of soil pore volume filled with water) throughout the duration of the experiment to simulate the same water regime as applied in the rainfed (monsoon) upland rice farmer's field.

Site selection, field preparation, treatments and experimental set-up of the field experiment.

A farmer's rice field naturally infested with M. graminicola located in Wanatkone village, Hlegu region, Yangon Division, was selected to carry out the field experiment. This field was chosen because it was found to be infested with M. graminicola during the 2009 survey (Win et al, 2011). The field had been cultivated with summer-irrigated and rainfed (monsoon) rice continuously for more than 15 years. The soil was acid with a pH of 5.5 and had a clay soil texture with a composition of 59% clay, 23% loam and 16% sand, with 0.18% total N2, 2.04 ppm P and 13.46 mg (100 g)-2 K2O. To control the M. graminicola populations in part of the experiment, carbofuran (Furadan 3G, Myanmar Aventis Crop Science Ltd.) at 2 kg a.s. ha-1 was applied.

Light irrigation was applied until a thin layer of water was standing in the field for land preparation (ploughing twice, harrowing once and land levelling) and because carbofuran must be applied in a thin layer of water according to the manufacturer's instructions. The soil was saturated when the pre-germinated seeds were planted 2 days after carbofuran application. After land preparation, the experimental area was divided into 12 main plots (10 x 9 m), each of which had 15 sub-plots (3 x 2 m) resulting in 180 sub-plots. Six main plots were

Table 1. Characteristics of the lowland and upland rice varieties included in the experiments

№ Rice variety (abbreviation) Original name Crop cycle (days) Plant height (cm) Grain yield (t ha-1) Lowland or upland variety Variety type (HYV: high-yielding variety)

1 Thukhayin (TKY) MNTKM 4-10 105-110 85-95 3.5-4 lowland HYV

2 Manawthukha (MNTK) Mashuri-M 135 90-105 4.5-5 lowland HYV

3 Shwemanaw (SMN) 110 110-120 5.2-6.2 lowland HYV

4 Sinthukha (STK) Yn 2068-7-1 135-140 105-120 3.5-4 lowland HYV

5 Yatanartoe (YTNT) Thai 1-9-3E 120 120-135 4.6-6.2 lowland HYV

6 OM 4900 (OM) OM 4900 100-105 110-115 4.5-5.5 lowland HYV

7 MR 9 (MR 9) MR 9 110-120 110-115 3.5-4 lowland HYV

8 Sinnweyin (SNY) Yn 2883-12-2-1 110-115 90-105 3.5-4 lowland HYV

9 Hmawbi 5 (HB 5) 6201 R 105 115-120 4.5-5 lowland HYV

10 Sinthwelatt (STL) IR 53936 135-140 120-135 3.5-4.5 lowland HYV

11 Yezin Lonethwe (YZLT) LTHMY-14 125 110-115 3.5-5 lowland HYV

12 Hmawbisan (HBS) 120-125 115-120 2.5-3.5 lowland HYV

13 Saytanar 1 (STN 1) IR 13240-108-2-2-3 125-130 105-110 4.5-5 lowland HYV

14 Thihtatyin (THY) IR 50 115 85-95 5-6 lowland HYV

15 Shwethweyin (STY) IR 55423-01 106-110 80-90 4.5-5.5 lowland HYV

16 Yezin Yar 9 (YZY 9) local 120 105-110 2.5-3 upland aerobic

17 Bukyauk (BK) improved 130-135 100-105 1-1.5 upland traditional

18 Kone Myint 2 (KM 2) local 140-145 105-110 1.5-2 upland traditional

19 Kaukme (KM) local 130-135 105-110 1-1.5 upland traditional

20 Myeshel (MS) local 130-135 120-125 1-1.5 upland traditional

21 Paullar (PL) local 130-135 110-120 1-1.5 upland traditional

22 Motehsoema Kyaekyay (MSMKK) local 140-150 110-115 1-1.5 upland traditional

23 Khaukphephan (KPP) local 135-140 110-115 1-1.5 upland traditional

24 Khaukparmauk (KPM) 140-145 100-110 1-1.5 upland traditional

Fig. 1. Screenhouse experiment. Reduction in yield (%) and population densities of Meloidogyne graminicola J2 per root system recovered from 15 upland rice varieties at harvest after inoculation with 3,000 MI. graminicola J2 per plant. For the full names of the rice varieties see Table 1.

treated with carbofuran and the other six main plots were left untreated. The carbofuran was broadcast

evenly over the plots and then incorporated into the

wet soil to a depth of 5 cm using a hoe. Earthen barriers (30 cm high * 50 cm diam.) were

constructed around each main plot to prevent movement of the carbofuran among the main plots. Two days after the carbofuran application, 3-day-old pre-germinated seeds of the 15 lowland varieties were uniformly direct wet seeded at 20 g seeds m-2

in the 90 sub-plots treated with carbofuran and the 90 untreated sub-plots. The first flooding was applied 1 week after sowing and a fertiliser application (25 kg N ha-1 applied as urea) was done at the tillering growth stage and again at the panicle initiation growth stage of the plants in accordance with the common rice production practices in the summer-irrigated lowland rice ecosystem in Myanmar. At 45 days after sowing, carbofuran was again applied to the six main plots that had been treated with carbofuran before to prevent recovery of the nematode populations. Weeding was practised throughout the duration of the experiment. The field was intermittently flooded twice a week.

The field experiment was laid out in a split plot design with two main plots (carbofuran-treated and untreated plots) and 15 sub-plots with six replications. The main plots were sub-divided into 15 treatments (rice varieties) that were randomly assigned to the sub-plots (Fig. 2).

Assessment of plant growth, yield-contributing traits, yield, Meloidogyne graminicola population densities and severity of root galling. All three experiments were terminated when the panicles of each rice variety were mature and ready to harvest. Since the varieties matured at different times, they were harvested at different times. At harvest, the following four plant growth and six yield-contributing traits were measured for all three experiments: fresh root weight, root length, dry shoot weight, plant height, number of tillers per plant, number of panicles per plant, number of filled grains per panicle; percentage filled grains per plant, filled grain weight per plant and weight of 1,000 filled grains. For the screenhouse experiments, yield was estimated according to Yoshida (1981): grain yield (t ha-1) = number of panicles m-2 x number of spikelets per panicle x percentage filled spikelets x 1,000 grain weight (g) x 10-7. "Spikelets" include all filled, partially filled and unfertilised spikelets. Filled spikelets are called "grains". The number of panicles m-2 was calculated as the number of plants m-2 x number of panicles per plant. For the field experiment, five plants, including rhizosphere soil, were up-rooted from each sub-plot and the plant growth and yield-contributing traits measured. The grain yield from all plants harvested was measured for each sub-plot (6 m2) at 14% moisture content.

For the two screenhouse experiments, at harvest rhizosphere soil of each plant was collected, mixed and the J2 were extracted from a sub-sample of 100 ml soil. For the field experiment, the nematode soil population density in each sub-plot was determined at 2 days before sowing (Pi) and at harvest (Pf). In each sub-plot, the soil was collected up to a depth of

15 cm at five places: in the four corners and in the middle. Five soil samples from each sub-plot were pooled and mixed. Then, the J2 were extracted from a sub-sample of 100 ml soil. For all three experiments, the tray method was used for the extraction of the J2 from the soil (Whitehead & Hemming, 1965). For the two screenhouse experiments, the up-rooted roots from each plant were chopped into approximately 1-cm pieces, thoroughly mixed and 1 sub-sample of 3 g roots was taken. For the field experiment, five up-rooted roots from each sub-plot were pooled, chopped, thoroughly mixed and one sub-sample of 3 g roots taken as described above. For all three experiments, each root sub-sample was macerated twice in a kitchen blender for 10 s and the J2 extracted from the resulting homogenate using the tray method (Whitehead & Hemming, 1965). After 24 h, the J2 from the soil and the root sub-samples were counted using a stereomicroscope. For the two screenhouse experiments, the final nematode population density was calculated as the number of J2 in the soil in the pots (1,500 ml) + J2 per root system. The nematode multiplication factor (M^sgs) was calculated as the final nematode population density per 3,000 J2. Eggs were not extracted.

For all three experiments, the severity of root galling (root galling index) was assessed from each up-rooted plant by rating the percentage of roots with root tip galls on a 0 to 10 scale according to the rice root-knot rating chart of Bridge & Page (1980).

Analysis of data. The data were analysed using STATISTICA 11.0 software (StatSoft Inc., Tulsa, USA). Prior to analysis of variance, plant growth and yield data, and nematode population densities were log(x+1) transformed, while percentage filled grains per plant and root galling indices were arcsin(x/100) transformed to meet the assumptions of ANOVA (i.e., normality and homogeneity of variances). The Shapiro-Wilk test was used to examine whether the dependent variable was normally distributed within groups while the homogeneity of the variances of the groups was tested with the Levene's test. The outliers were determined by calculating the standardised residuals falling outside the range from -2 to +2. When the assumptions for ANOVA were met, the data were analysed by using ANOVA.

For the two screenhouse experiments, factorial analysis of ANOVA was used to examine the effect of nematode inoculation on plant growth and yield-contributing traits, and yield compared with the uninoculated control plants. In the case of interaction between the two factors (rice variety and nematode inoculation), individual comparisons were made

between inoculated and uninoculated plants with a t-test (P < 0.05) and presented for each rice variety separately. If there was no interaction between the two factors (rice variety and nematode inoculation) for a specific plant growth or yield-contributing trait, the

factor level means were compared by Tukey's HSD test and presented for all rice varieties together. Oneway ANOVA was performed for mean comparisons of the root galling indices and nematode population densities with Tukey's HSD test.

Fig. 2. Field experiment. Reduction in yield (%) and population densities of Meloidogyne graminicola. A: J2 per root system; B: J2 per g roots recovered from 15 lowland rice varieties at harvest grown in sub-plots not treated with carbofuran. For the full names of the rice varieties see Table 1.

For the field experiment, data for main effects and interactions of rice variety and carbofuran treatment were analysed using factorial ANOVA to examine the effect of carbofuran application on the nematode population densities and root galling indices compared with untreated control plants. When treatment effects were significant at P < 0.05, means were separated using Tukey's HSD test. For plant growth traits and yield data, factorial ANOVA was used to examine the effect of carbofuran-treated compared with untreated plants. In the case of interaction between the two factors (rice variety and carbofuran treatment), individual comparisons were made between carbofuran-treated and untreated plants with a t-test and presented for each rice variety separately. In the case of absence of interaction between the two factors (rice variety and carbofuran treatment), the factor level means were

compared by Tukey's HSD test and presented for all rice varieties together.

For all three experiments, linear relationships between yield, nematode population densities and root galling indices were analysed using Pearson correlation analysis.

Mean numbers are shown in the tables, after back transforming the data, to facilitate interpretation.

RESULTS

Screenhouse experiment with lowland rice varieties. In Table 2, the soil and root population densities of M. graminicola J2, the severity of root galling (root galling index) and the nematode multiplication factor (M^sgs) on 15 lowland rice varieties at harvest are presented. At harvest, the soil

population density was significantly (P < 0.05) higher in the rhizosphere of the variety Thihtatyin (THY), which had the highest soil population density (911 J2 (100 ml soil)-1), compared with the variety Hmawbisan (HBS), which had the lowest soil population density (52 J2 (100 ml soil)-1). Only the soil population densities of THY (highest) and HBS (lowest) were significantly different. The soil population densities of the other varieties were in between these of THY and HBS and not significantly different from either THY or HB5. Differences in nematode reproduction in the roots were observed among the 15 lowland varieties. The highest J2 population densities per g roots were recovered from the varieties Sinthwelatt (STL; 7,979), Thihtatyin (THY; 7,743), Hmawbisan (HBS; 7,436) and Hmawbi 5 (HB 5; 7,299), which were all significantly (P < 0.05) higher compared with the varieties Shwethweyin (STY) and Sinnweyin (SNY) which had the lowest J2 population densities per g roots (2,029 and 2,917, respectively). The J2 population density per root system was highest in the variety Hmawbisan (HBS; 157,705) and lowest in the varieties Sinnweyin (SNY) and Shwethweyin (STY) (37,989 and 39,247, respectively). The J2 population densities per root system of the other varieties were not significantly different nor from the variety Hmawbisan (HBS) or from the varieties Sinnweyin (SNY) and Shwethweyin (STY). The Mreggs was lowest on the variety Sinnweyin (SNY; 13.2) and highest on the variety Hmawbisan (HBS; 52.8) followed by the varieties Thihtatyin (THY) and OM 4900 (OM) (both 47.5). Differences in severity of root galling were observed among the 15 lowland varieties. Variety Thihtatyin (THY) showed the highest root galling index (6.2) while variety Shwethweyin (STY) showed the lowest root galling index (2.6). The other varieties were not significantly different nor from the variety Thihtatyin (THY) or from the variety Shwethweyin (STY).

In Table 3, fresh root weight, root length, dry shoot weight and plant height at harvest of the 15 lowland varieties in uninoculated and inoculated plants are presented. Interaction between rice varieties and nematode infection did not show a significant effect on any of these plant growth traits. Thus, the factor level means (the effect of nematode infection) were calculated for all varieties together. On average, infection with M. graminicola significantly (P < 0.05) reduced fresh root weight by 41%, root length by 19.7%, plant height by 5.8% and dry shoot weight by 28.2% at harvest.

In Table 4, the number of tillers per plant and panicles per plant at harvest of the 15 lowland

varieties and their reduction in numbers in inoculated plants are presented. For these traits, interaction between rice varieties and nematode infection did show a significant (P < 0.05) effect. Thus, the effect of nematode infection was calculated for each variety separately. The highest reduction in number of tillers was observed on the variety Shwemanaw (SMN; 50%) while the highest reduction in number of panicles was observed on the variety Sinthukha (STK; 48.9%). Nematode infection did not have a significant effect on the number of tillers or panicles in five out of the 15 lowland varieties. For the varieties Hmawbisan (HBS) and Thihtatyin (THY), a significant (P < 0.05) reduction in the number of tillers (with 32.7 and 22.2%, respectively) was observed but not in the number of panicles. For the varieties Hmawbi 5 (HB 5) and MR 9, a significant (P < 0.05) reduction in the number of panicles (with 33.3 and 25%, respectively) was observed but not in the number of tillers. On average, infection with M. graminicola significantly (P < 0.05) reduced the number of tillers and panicles by 29.1 and 21.2%, respectively.

In Table 5, a selection of yield-contributing traits and yield at harvest of the 15 lowland varieties in uninoculated in inoculated plants are presented. Interaction between rice varieties and nematode infection did not show a significant effect on any of these yield-contributing traits and yield. Thus, the factor level means (the effect of nematode infection) were calculated for all varieties together. On average, infection with M. graminicola significantly (P < 0.05) reduced the number of filled grains per panicle by 30%, percentage filled grains per plant by 11.7%, filled grain weight per plant by 34.7%, 1,000 filled grains weight by 6.5% and yield by 31.1% at harvest.

Figure 1 shows the effect of M. graminicola infection on the estimated yield of the 15 lowland rice varieties. Compared with uninoculated plants, the highest percentage yield reduction was observed on the variety Saytanar 1 (STN 1; 44%), followed by the varieties Hmawbi 5 (HB 5) and OM 4900 (OM) (42.9 and 40.7%, respectively). The lowest percentage yield reduction was observed on the variety Sinnweyin (SNY; 20.4%), followed by the variety Shwethweyin (STY; 22.6%). There was no correlation between J2 per g roots or J2 per root system or severity of root galling and the percentage yield reduction.

Field experiment with lowland rice varieties.

The initial soil population densities of M. graminicola J2 recovered from the 180 sub-plots in January before sowing of the 15 lowland rice varieties averaged 29 (100 ml soil)-1 (Table 6).

There were no significant differences in initial J2 soil population densities between the sub-plots which were designated to be treated with carbofuran or left non-treated (data not presented).

In Table 7, the soil and root population densities of M. graminicola J2, and the severity of root galling (root galling index) on 15 lowland rice varieties in carbofuran-treated plots (C+) and non-treated plots (C) at harvest are presented. The soil and root J2 population densities, and the severity of root galling were significantly (P < 0.05) influenced by both variety and carbofuran treatment but the interactions were not significant. On average, treatment with carbofuran had significantly (P < 0.05) reduced at harvest the number of J2 per 100 ml soil and the number of J2 per g roots by about 80%, the number of J2 per root system by about 70%, and the root galling index from 3.3 to 0.4. When the varieties grown in the sub-plots that had not been treated with carbofuran were analysed separately, the J2 soil population density recovered from the variety OM 4900 (OM; 5,726) was significantly (P < 0.05) higher compared with the varieties Sinthukha (STK; 292), Sinthwelatt (STL; 308) and Shwethweyin (STY; 408). The J2 population density per g roots recovered from the variety Saytanar 1 (STN 1; 3,174) was significantly (P < 0.05) higher compared with the varieties Manawthukha (MNTK; 851), Sinnweyin (SNY; 710), Yezin Lonethwe (YZLT; 700), Sinthwelatt (STL; 675), Sinthukha (STK; 557) and OM 4900 (OM; 522). The J2 population densities per root system recovered from the varieties Saytanar 1 (STN 1; 8,933) and Thukhayin (TKY; 8,478) were significantly (p < 0.05)

higher compared with the varieties OM 4900 (OM; 1,671) and Shwethweyin (STY; 1,398). The root galling index was highest on the variety Shwemanaw (SMN) and lowest on the variety Shwethweyin (STY) (4.4 vs 2.3). The root galling indices of the other varieties were not significantly different nor from the variety Shwemanaw (SMN) or the variety Shwethweyin (STY).

In Table 8, the vegetative plant growth and yield-contributing traits and yield of the 15 lowland rice varieties at harvest, and the percentage reduction between plants grown in carbofuran-treated and non-treated sub-plots are presented. Interaction between varieties and carbofuran treatment did not show a significant effect on any of these traits. Thus, the factor level means (the effect of nematode infection) were calculated for all varieties together. On average, the highest significant (P < 0.05) reduction was observed for the number of tillers (30.4%) followed by the number of panicles (27.8%) and the number of filled grains per panicle (22.1%). Plant height, percentage filled grains per plant and 1,000 filled grain weight were not significantly reduced.

Figure 2 shows the effect of M. graminicola infection on the estimated yield of the plants of the 15 lowland rice varieties grown in the sub-plots that had not received carbofuran treatment compared with the plants that had been grown in carbofuran-treated sub-plots. The highest percentage yield reduction was observed on the varieties Shwemanaw (SMN) and Hmawbi 5 (HB 5) (both 26.3%) followed by the varieties Saytanar 1 (STN 1) and OM 4900 (OM) (25.6 and 25.5%, respectively).

Fig. 3. Screenhouse experiment. Reduction in yield (%) and population densities of Meloidogyne graminicola J2 per root system recovered from nine upland rice varieties and the reference lowland rice vairiety Thihtatyin (THY) at harvest after inoculation with 3,000 M. graminicola J2 per plant. For the full names of the rice varieties see Table 1.

Table 2. Screenhouse experiment. Soil and root population densities of Meloidogyne graminicola second-stage juveniles (J2), severity of root galling (RGI) and nematode multiplication factor (Mfeggs) on 15 lowland rice varieties at harvest after inoculation with 3,000 M. graminicola J2 per plant

Rice variety No. of plants (n) Population densities ofM. graminicola RGI Mf -eggs

J2 (100 ml soil)-1 J2 (g root) -1 J2 (root system) 1

TKY 6 230±129 ab 6,994±4,864 bc 67,315±54,007 ab 5.3±2.1 ab 23.6

MNTK 6 201±184 ab 4,224±3,178 ab 84,708±57,839 ab 3.5±0.5 ab 29.2

SMN 6 389±335 ab 6,154±1,802 bc 127,188±69,714 ab 5.6±2.2 ab 44.3

STK 6 154±278 ab 5,025±4,032 abc 96,115±72,313 ab 3.3±2.1 ab 32.8

YTNT 6 176±201 ab 7,188±4,799 c 118,681±76,180 ab 5.0±1.9 ab 40.4

OM 6 362±295 ab 6,747±2,166 bc 137,148±57,617 ab 5.5±1.5 ab 47.5

MR 9 6 318±169 ab 4,006±1,300 ab 62,090±27,939 ab 4.7±1.4 ab 22.3

SNY 6 110±115 ab 2,917±2,159 a 37,989±19,893 a 3.8±0.8 ab 13.2

HB 5 6 119±115 ab 7,299±2,484 c 81,427±55,780 ab 5.2±1.2 ab 27.7

STL 6 117±148 ab 7,979±8,522 c 81,020±38,560 ab 3.8±2.3 ab 27.6

YZLT 6 95±82 ab 5,997±3,745 abc 100,682±102,347 ab 4.1±0.9 ab 34.0

HBS 6 52±59 a 7,436±2,665 c 157,705±49,306 b 5.4±1.9 ab 52.8

STN 1 6 99±67 ab 7,273±2,809 c 108,732±42,910 ab 5.8±1.5 ab 36.7

STY 6 281±198 ab 2,029±1,510 a 39,247±23,472 a 2.6±1.4 a 14.5

THY 6 911±770 b 7,743±3,697 c 128,753±123,248 ab 6.2±2.1 b 47.5

Average 234±237 5,945±3,881 95,764±67,584 4.6±1.9 32.9

Data represent means ± SD. Means in the same column followed by the same letter are not significantly different according to Tukey's HSD test (P < 0.05).

RGI = root galling index according to 0 = no swellings or galls, 1 = 10% galls, 2 = 20% galls, 3 = 30% galls, 4 = 40% galls, 5 = 50% galls, 6 = 60% galls, 7 = 70% galls, 8 = 80% galls, 9 = 90% galls and 10 = all roots of the root system galled. For the full names of the rice varieties see Table 1.

The lowest percentage yield reduction was observed on the variety Shwethweyin (STY; 5.7%), followed by the variety Sinnweyin (SNY; 8.8%). When non-treated sub-plots were analysed separately, the percentage yield reduction of all lowland rice varieties was positively correlated with the root galling index (P = 0.01, r = 0.63, n = 90). By contrast, there was no correlation between the J2 soil and root population density or root galling index and the percentage yield reduction in the carbofuran-treated sub-plots.

Screenhouse experiment with upland rice varieties. In Table 9, the soil and root population densities of M. graminicola J2, the severity of root galling (root galling index) and the nematode multiplication factor (Mf-eggs) on nine upland rice varieties and the reference lowland variety Thihtatyin (THY) at harvest are presented. The soil population density was significantly (P < 0.05) higher in the rhizosphere of the variety Motehsoema Kyaekyay (MSMKK), which had the highest soil population density of the upland varieties (425 J2 (100 ml soil)-1), compared with variety Paullar (PL), which had the lowest soil population density (30 J2 (100 ml soil)-1). The soil population densities of the other varieties were not significantly different from the variety Motehsoema Kyaekyay (MSMKK), which had the highest soil population density, or from the variety Paullar (PL). At harvest, no

significant differences in J2 per g roots were observed among the nine upland varieties and the reference variety Thihtatyin (THY). The highest J2 population density per root system was recovered from the variety Khaukphephan (KPP; 167,573) which was significantly (P < 0.05) higher compared with all other varieties, except the varieties Kaukme (KM) and Myeshel (MS; 170,485 and 171,514, respectively). The Mreggs was lowest on the variety Paullar (PL; 39.8) followed by the reference variety Thihtatyin (THY; 40.5) and highest on the variety Khaukphephan (KPP; 108.4). Differences in severity of root galling were observed among the nine upland varieties and the reference variety Thihtatyin. The root galling index was highest on the variety Thihtatyin (THY; 8.4), which was significantly (P < 0.05) higher compared with the varieties Yezin Yar 9 (YZY; 5.8), Kaukme (KM; 5.9), Paullar (PL; 4.7), Khaukparmauk (KPM; 6.4) and Khaukphephan (KPP; 3.6), which had the lowest root galling index.

In Table 10, fresh root weight, root length, dry shoot weight and plant height at harvest of the nine upland rice varieties and the reference lowland variety Thihtatyin, and the reduction in growth between uninoculated and inoculated plants are presented. Interaction between varieties and nematode infection did show a significant (P < 0.05) effect. Thus, the effect of nematode infection was

Table 3. Screenhouse experiment. Comparison of fresh root weight, root length, plant height and dry shoot weight at harvest of 15 lowland rice varieties uninoculated or inoculated with 3,000 Meloidogyne graminicola J2 per plant

Treatment No. of plants (n) Fresh root weight (g) Root length (cm) Plant height (cm) Dry shoot weight (g)

Uninoculated 90 26.8±6.7 a 31.7±6.5 a 71.2±9.2 a 7.1±2.3 a

Inoculated 90 15.8±5.2 b 25.4±5.1 b 67.1±8.6 b 5.1±1.9 b

Reduction (%) 41.0 19.7 5.8 28.2

Data represent means ± SD. Means in the same column followed by the same letter are not significantly different according to Tukey's HSD test (P < 0.05).

Table 4. Screenhouse experiment. Comparison of the number of tillers and panicles per plant of 15 lowland rice varieties at harvest uninoculated or inoculated with 3,000 Meloidogyne graminicola J2 per plant

Rice variety No. of plants (n) No. of tillers per plant No. of panicles per plant

UI I Reduction (%) UI I Reduction (%)

TKY 6 8±2 8±2 9.5 6±1 a 6±1 a 0.0

MNTK 6 14±4 8±2 41.6* 9±2 a 6±2 b 33.3*

SMN 6 13±1 7±2 50.0* 6±1 a 5±1 b 17.2*

STK 6 11±4 6±2 43.9* 8±2 a 5±2 b 48.9*

YTNT 6 7±3 8±2 (+)5.5 6±1 a 6±1 a 0.0

OM 6 8±2 7±2 10.8 6±1 a 6±1 a 0.0

MR 9 6 11±1 9±2 13.9 8±1 a 6±1 b 25.0*

SNY 6 12±4 7±1 43.1* 6±1 a 4±1 b 33.3*

HB 5 6 9±3 4±1 32.9 6±1 a 4±1 b 33.3*

STL 6 10±3 7±1 33.0* 6±1 a 4±1 b 31.7*

YZLT 6 11±3 7±2 39.3* 6±1 a 5±1 b 16.7*

HBS 6 11±2 7±2 32.7* 7±1 a 7±1 a 0.0

STN 1 6 11±1 7±2 37.0* 7±1 a 5±1 b 28.6*

STY 6 10±1 6±1 33.7* 7±1 a 5±1 b 28.6*

THY 6 9±1 7±2 22.2* 7±1 a 6±1 a 14.3

Average 90 10±7 7±2 29.1* 7±1 a 5±2 b 21.2*

Data represent means ± SD. Percentage change followed by * indicates that the change is significantly (P < 0.05) different

according to the t-test.

UI: uninoculated. I: inoculated.

(+): indicates an increase in the inoculated plants compared to the uninoculated plants. For the full names of the rice varieties see Table 1.

Table 5. Screenhouse experiment. Comparison of yield-contributing traits of 15 lowland rice varieties at harvest uninoculated or inoculated with 3,000 Meloidogyne graminicola J2 per plant

Treatment No. of plants (n) No. of filled grains per panicle Percentage filled grains per plant Filled grain weight per plant (g) 1,000 filled grains weight (g) Yield (t ha-1)

Uninoculated 90 60±13 a 85.2±7.6 a 7.5±1.2 a 21.5±3.6 a 3.3±0.6 a

Inoculated 90 42±13 b 75.2±9.9 b 4.9±1.1 b 20.1±3.4 b 2.2±0.5 b

Reduction (%) 30.0 11.7 34.7 6.5 31.1

Data represent means ± SD. Means in the same column followed by the same letter are not significantly different according to Tukey's HSD test (P < 0.05).

Table 6. Field experiment. Initial field population density (Pi) of Meloidogyne graminicola J2 recovered from 100 ml soil sub-samples from 180 sub-plots before sowing of 15 upland rice varieties in the experimental field located in the summer-irrigated lowland rice ecosystem, Hlegu region, Yangon Division, Myanmar

Rice variety to be sown in the sub-plot No. of sub-plots (n) No. of J2 (min-max) Mean J2 ± SD

TKY 12 6-55 20±18

MNTK 12 11-56 25±17

SMN 12 5-38 22±11

STK 12 10-45 25±15

YTNT 12 17-83 34±25

OM 12 3-47 43±17

MR 9 12 24-84 39±24

SNY 12 9-36 41±23

HB 5 12 10-59 28±19

STL 12 10-110 39±37

YZLT 12 13-56 28±16

HBS 12 16-73 39±25

STN 1 12 10-54 27±17

STY 12 7-69 33±23

THY 12 10-74 24±14

Total (range) Average for all sub-plots 180 3-110 29±21

For the full names of the rice varieties see Table 1.

calculated for each variety separately. Four upland varieties (Yezin Yar 9, Paullar, Kone Myint 2 and Myeshel) showed a significant (P < 0.05) reduction in fresh root weight ranging from 45.6 to 27.9%. Fresh root weight of the reference variety Thihtatyin (THY) was also significantly (P < 0.05) reduced (by 30.1%). The root length of all upland varieties and the reference variety Thihtatyin (THY), except the varieties Paullar (PL) and Khaukphephan (KPP), were significantly (P < 0.05) reduced by 13.5 to 35.8%. In six upland varieties, dry shoot weight was significantly (P < 0.05) reduced by 25.7 to 39%. Two upland varieties (Myeshel and Kaukme) showed a significant (P < 0.05) reduction in plant height (by 18.4 and 9.5%, respectively). Plant height of the reference variety Thihtatyin (THY) was also significantly (P < 0.05) reduced (by 25.1%). On average, infection with M. graminicola significantly (P < 0.05) reduced fresh root weight with 26.1%, root length by 23.2%, plant height with 5.7% and dry shoot weight by 30.7%.

In Table 11, a selection of yield-contributing traits and yield at harvest of the nine upland rice varieties and the reference lowland variety Thihtatyin, and the percentage reduction between uninoculated and inoculated plants are presented. Interaction between varieties and nematode infection did not show a significant effect on any of these yield-contributing traits and yield. Thus, the factor level means (the effect of nematode infection) was calculated for all varieties together. On average,

infection with M. graminicola significantly (P < 0.05) reduced the filled grain weight per plant by 47.6%, yield by 44.9%, number of filled grains per panicle by 33.4%, number of panicles per plant by 15%, number of tillers per plant by 14.1% and 1,000 filled grains weight by 10.9%. No significant reduction of percentage filled grains per plant was observed.

Figure 3 shows the effect of M. graminicola infection on the estimated yield of the nine upland rice varieties and the reference lowland variety Thihtatyin. Compared with uninoculated plants, the highest yield reduction was observed on the variety Kaukme (KM; 60.4%), followed by the variety Khaukparmauk (KPM; 60.0%). The lowest yield reduction was observed on the variety Khaukphephan (KPP; 9.8%). There was no correlation between J2 per g roots or J2 per root system and percentage yield reduction. However, the percentage yield reduction of all upland varieties was positively correlated with the root galling index (P = 0.02, r = 0.70, n = 60).

DISCUSSION

None of the 15 lowland and nine upland rice varieties included in our screenhouse experiments was resistant to M. graminicola. Although differences in susceptibility were observed among the lowland and upland varieties, and between the two types of rice varieties (lowland and upland),

with an Mf~eggs ranging from 13.2 to 52.8 for the lowland varieties vs 39.8 to 108.4 for the upland varieties, all varieties included in our study can be considered as highly susceptible to M. graminicola based upon their population densities at harvest. Also in the field experiment all the 15 lowland rice varieties were susceptible to M. graminicola infection.

Our study confirms the observation by Win et al. (2014) that upland rice varieties are, in general, more susceptible to M graminicola infection than lowland rice varieties. In our screenhouse experiments, the average J2 per g roots, J2 per root system and Mreggs at harvest were considerable higher in the upland varieties compared with the lowland varieties (on average 14,552 vs 5,945; 164,434 vs 95,764 and 54.3 vs 32.9, respectively). This re-confirmation is further strengthened by the difference in severity of root galling we observed between these two types of rice varieties: in the lowland varieties, the root gall index of the inoculated plants averaged 4.6 vs 6.4 in the upland varieties.

The suggestion that upland rice varieties are, in general, more susceptible to M graminicola infection than lowland rice varieties seems to be in contradiction with the results of an extensive nematological field survey carried out in several rice-based agro-ecosystems in Myanmar (Win et al., 2011). This survey showed that in the rainfed (monsoon) upland rice ecosystem the frequencies of occurrence, J2 soil and root population densities, and prominence values of M. graminicola on the upland varieties Kaukme, Khaukphephan, Myeshel and Kone Myint 2 were very low while M. graminicola J2 were not even found in the 15 fields examined in which the upland variety Paullar was grown. In the summer-irrigated lowland rice ecosystem, however, the frequencies of occurrence, population densities and root galling indices of M. graminicola were significantly higher on the lowland varieties cultivated than on the upland varieties cultivated in the rainfed upland rice ecosytem. In our study, the varieties Kaukme, Khaukphephan, Myeshel and Kone Myint 2 had an Mf~eggs ranging from 43 to 108.4, while the variety Paullar had an Mf-eggs of 39.8. In Myanmar, upland varieties are grown in the upland agro-ecosystem in a 2-year rotation system: rice followed by fallow in the first year and cultivation of various crops, such as maize, potato, garlic, ginger, pulses and oil seed crops, in the second year (Egashira & Than, 2006). Our study may support the suggestion by Win et al. (2014) that this 2-year rotation system is effective in keeping the M. graminicola population densities

low in the rainfed upland rice ecosystem in Myanmar, thus limiting in this region damage and yield losses that can be caused by this nematode species.

Win et al. (2014) examined the host response (susceptibility) to M. graminicola infection of the same lowland and upland rice varieties included in our study at 8 weeks after inoculation (WAI) when the plants were at the stem elongation growth stage. Because this study was carried out under screenhouse conditions similar to our experiments, we can compare the susceptibility to M. graminicola infection of these varieties at an early plant growth stage with their susceptibility at harvest. Among the 15 lowland rice varieties, some varieties showed a similar host response at 8 WAI and at harvest. An example of such a variety is Sinnweyin, which had the lowest Mreggs at harvest while Win et al. (2014) considered this variety as less susceptible when evaluated at 8 WAI. By contrast, some lowland varieties that were considered less susceptible when evaluated at 8 WAI appeared to be highly susceptible when their Mf-eggs was assessed at harvest and vice versa. Examples of such lowland varieties are Yatanartoe and Sinthukha, respectively. All nine upland rice varieties had a very high Mreggs at harvest and no significant differences in J2 per g roots were observed among the varieties. Only one variety (Khaukpephan) had a significantly higher numbers of J2 per root system compared with most other varieties. By contrast, based on the Mreggs assessed at 8 WAI, Win et al. (2014) considered one upland variety (Bukyauk) as less susceptible while three upland rice varieties (Khaukphephan, Yezin Yar 9 and Kone Myint 2) were considered as moderately susceptible. At harvest, the Mreggs of the variety Bukyauk was about as high as the average Mf-eggs of all upland varieties (56.4 and 54.3, respectively), while the Mreggs of the variety Khaukphephan (108.4) was the highest among the upland varieties. The other two varieties (Yezin Yar 9 and Kone Myint 2) had the lowest Mf-eggs at harvest (42.4 and 43, respectively) among the upland varieties.

These differences in susceptibility to M. graminicola infection between the early growth stages and maturity of rice plants indicate that it is very difficult to identify at an early plant growth stage (stem elongation) at 8 WAI both lowland and upland rice varieties that are less susceptible to M. graminicola infection. These differences underline the fact that one should be very careful in extrapolating host responses to M. graminicola infection of rice varieties at an early plant growth stage to their host responses over an entire crop cycle.

Table 7. Field experiment. Soil and root population densities of Meloidogyne graminicola J2, and severity of root galling (RGI) on 15 lowland rice

varieties at harvest grown in carbofuran-treated (C+) and non-treated (C-) sub-plots

Population densities of M. graminicola RGI

Rice variety J2 (100 ml soil)"1 J2 (g roots) 1 J2 (root system) 1

C+ C" C+ C" C+ C" C+ C"

TKY 214±227 ns 2,560±1,177 ab1 275±260 ns 2,052±1,123 ab1 2,515±1,753 ns 8,478±3,834 b 0.3±0.4 ns 3.9±1.5 ab1

MNTK 361±319 982±616 ab 273±317 851±276 a 1,281±1,249 2,748±1,908 ab 0.5±0.3 3.4±0.6 ab

SMN 291±287 1,004±204 ab 196±166 1,354±683 ab 954±825 4,662±3,440 ab 0.2±0.1 4.4±1.1 b

STK 156±132 292±285 a 282±138 557±386 a 1,202±658 2,315±2,847 ab 0.2±0.2 3.7±1.4 ab

YTNT 257±274 1,957±2,006 ab 242±211 1,663±2,000 ab 883±904 6,937±6,014 ab 0.6±0.6 3.0±1.1 ab

OM 343±345 5,726±3,721 b 124±122 522±578 a 698±680 1,671±1,616 a 0.6±0.4 3.9±1.1 ab

MR 9 79±57 547±316 ab 329±303 1,091±619 ab 1,841±1,642 3,819±3,014 ab 0.3±0.1 3.1±0.5 ab

SNY 104±52 1,200±1,444 ab 84±88 710±532 a 516±454 2,364±1,915 ab 0.1±0.1 2.8±0.4 ab

HB 5 312±224 1,823±679 ab 224±414 1,103±709 ab 1,228±1,195 3,857±2,511 ab 0.4±0.3 3.4±0.9 ab

STL 52±28 308±237 a 181±147 675±484 a 895±532 3,800±2,595 ab 0.3±0.3 2.9±0.3 ab

YZLT 244±302 2,678±222 ab 370±332 700±356 a 1,128±663 2,083±1,587 ab 0.1±0.1 3.0±1.2 ab

HBS 491±442 825±739 ab 354±96 1,417±740 ab 1,147±1,140 4,177±2,381 ab 0.3±0.2 3.7±0.9 ab

STN 1 662±486 2,918±688 ab 592±427 3,174±1,557 b 3,031±1,704 8,933±3,280 b 0.5±0.5 3.9±1.3 ab

STY 135±105 408±431 a 95±84 684±573 a 404±233 1,398±1,094 a 0.5±0.5 2.3±0.4 a

THY 1,793±1,666 2,422±3,931 ab 445±390 1,695±1,427 ab 2,367±2,179 7,022±6,743 ab 0.3±0.2 3.5±1.6 ab

Average for all 335±533 B2 1,649±1,845 A 257±262 B 1,288±1,141 A 1,351±1,368 B 4,284±4,410 A 0.4±0.4 B 3.3±1.1 A

varieties

P>F

Variety <0.05 <0.01 <0.01 <0.01

Carbofuran <0.05 <0.001 <0.001 <0.001

Var. x Carbof. 0.9 0.9 0.9 0.32

Data represent means ± SD (n = 6). ns: not significantly different among the varieties. C+: Carbofuran-treated; C-: non-treated.

RGI = root galling index according to 0 = no swellings or galls, 1 = 10% galls, 2 = 20% galls, 3 = 30% galls, 4 = 40% galls, 5 = 50% galls, 6 = 60% galls, 7 = 70% galls, 8 = 80% galls, 9 = 90% galls and 10 = all roots of the root system galled.

1 Means in the same column followed by the same lowercase letter are not significantly different according to Tukey's HSD test (P < 0.05).

2 Means in the same row followed by the same uppercase letter, between carbofuran-treated and non-treated sub-plots, are not significantly different according to Tukey's HSD test (P < 0.05).

For the full names of the rice varieties see Table 1.

Table 8. Field experiment. Comparison of plant growth and yield-contributing traits of 15 lowland rice varieties at harvest grown in carbofuran-treated (C+) and non-treated (C) sub-plots

Plant growth and yield-contributing traits No. of plants (n) C+ C- Reduction (%)

Plant height (cm) 450 60.9±9.6 59.5±9.6 2.4

Dry shoot weight (g) 450 12.2±2.4 9.7±1.6 20.5*

Fresh root weight (g) 450 5.2±2.4 4.3±1.7 17.3*

No. of tillers per plant 450 5±1 3±1 30.4*

No. of panicles per plant 450 4±2 3±1 27.8*

No. of filled grains per panicle 450 40±18 31±16 22.1*

Filled grains per plant (%) 450 56±21.8 53.6±22.5 4.3

Filled grain weight per plant (g) 450 3.5±2.3 3.2±2.1 8.6*

1,000 filled grains weight (g) 450 20.5±3.7 19.9±3.6 2.9

Yield (t/ha) all 3.3±1.3 2.8±1.1 16.5*

Data represent means ± SD. Percentage change followed by * indicates that the change is significantly different according to Tukey's HSD test (P < 0.05).

Table 9. Screenhouse experiment. Soil and root population densities of Meloidogyne graminicola J2, severity of root galling (RGI) and nematode multiplication factor (Mfeggs) on nine upland rice varieties and the reference lowland rice variety Thihtatyin (THY) at harvest after inoculation with 3,000 M. graminicola J2 per plant

Rice variety No. of Population densities of M. graminicola RGI Mf

plants (n) J2 (100 ml soil)-1 J2 (g roots) 1 J2 (root system) -1 -eggs

BK 6 161±136 ab 15,481±7,858 a 166,857±139,386 a 7.3±0.4 ab 56.4

YZY 9 6 214±130 ab 11,140±2,483 a 124,019±53,320 a 5.8±1.1 ac 42.4

KM 2 6 344±348 ab 10,768±3,996 a 123,738±80,176 a 7.0±1.0 ab 43.0

KM 6 202±122 ab 14,865±5,017 a 170,485±63,932 ab 5.9±0.9 ac 56.8

MS 6 163±167 a 17,818±6,848 a 171,514±78,737 ab 7.1±0.9 ab 58.0

PL 6 30±22 c 12,069±2,309 a 118,994±14,205 a 4.7±1.3 cd 39.8

MSMKK 6 425±212 b 15,089±9,465 a 167,573±107,627 a 7.0±0.7 ab 58.0

KPP 6 375±546 ab 16,495±7,052 a 319,660±209,073 b 3.6±0.9 d 108.4

KPM 6 140±31 ab 14,271±5,403 a 150,764±78,230 a 6.4±1.1 a 51.0

THY 6 263±286 ab 16,599±12,216 a 117,438±78,320 a 8.4±0.4 b 40.5

Average 60 232±259 14,552±6,882 164,434±112,955 6.4±1.5 54.3

Data represent means ± SD. Means in the same column followed by the same letter are not significantly different according to Tukey's HSD test (P < 0.05).

RGI = root galling index according to 0 = no swellings or galls, 1 = 10% galls, 2 = 20% galls, 3 = 30% galls, 4 = 40% galls, 5 = 50% galls, 6 = 60% galls, 7 = 70% galls, 8 = 80% galls, 9 = 90% galls and 10 = all roots of the root system galled. For the full names of the rice varieties see Table 1.

At 8 WAI (Win et al, 2014), the average Mreggs of the 15 lowland rice varieties (also inoculated with 3,000 M. graminicola J2 per plant) was 2.1 vs 32.9 at harvest (our study). This is an increase of 15.7 times. The average Mreggs of the nine upland rice varieties (and the lowland rice variety Thihtatyin; also inoculated with M. graminicola 3,000 J2 per plant) was 3.8 at 8 WAI vs 54.3 at harvest (Win et al., 2014 vs our study, respectively). This is an increase of 14.3 times. One should be careful to combine the results of separate experiments (although carried out under very similar screenhouse

conditions) but this observation may indicate that the population dynamics of M. graminicola during a crop cycle might be similar in the two types of rice varieties and that the differences in soil type and water regime between the lowland and upland varieties experiments did not have an effect on the nematode population dynamics of M. graminicola once the J2 had infected (penetrated) the roots. This is an aspect that should be further investigated.

A comparison of the M. graminicola J2 population density per g root at 8 WAI (Win et al., 2014) with the population density at harvest (our study)

Table 10. Screenhouse experiment. Comparison of fresh root weight, plant height and dry shoot weight of nine upland rice varieties and the reference lowland rice variety Thihtatyin (THY) at harvest uninoculated or inoculated with 3,000 Meloidogyne graminicola J2 per plant

Rice variety Fresh root weight (g) Root length (cm) Plant height (cm) Dry shoot weight (g)

UI I Reduction (%) UI I Reduction (%) UI I Reduction (%) UI I Reduction (%)

BK 13.5±7.7 12.3±4.8 8.8 28.3±2.0 21.0±3.6 25.9* 115.7±4.4 117.2±16.4 (+)1.3 8.4±3.5 5.1±1.4 39.0*

YZY 9 24.8±4.6 13.5±3.8 45.6* 30.5±2.5 25.6±2.6 16.1* 102.7±13.1 90.2±5.6 12.1 11.2±4.1 7.3±1.8 34.6

KM 2 22.3±2.2 12.7±2.7 43.3* 38.8±3.6 25.2±3.3 35.2* 113.5±5.7 107.7±4.4 5.1 13.2±1.9 8.7±1.9 34.5*

KM 17.4±4.5 15.2±5.2 12.4 34.2±3.5 25.7±2.9 24.9* 129.3±3.7 117.0±2.9 9.5* 11.4±2.1 7.3±1.3 36.3*

MS 18.1±2.2 13.1±4.1 27.9* 34.5±5.4 23.7±2.5 31.4* 121.5±7.5 99.2±13.3 18.4* 10.6±2.0 7.7±5.0 27.0

PL 24.1±7.9 13.5±1.6 44.1* 30.3±6.0 29.8±3.4 1.6 128.7±7.5 122.2±11.2 5.1 8.3±1.1 5.4±0.8 35.1*

MSMKK 21.2±4.8 16.8±1.9 20.5 36.0±3.1 25.5±2.6 29.2* 115.7±8.1 115.3±8.5 0.3 12.5±2.3 9.4±1.9 24.4

KPP 23.4±3.4 22.0±5.6 6.0 36.3±2.9 31.2±6.4 14.2 130.3±7.1 117.8±23.8 9.6 13.8±1.0 10.3±3.1 25.7*

KPM 18.4±5.8 12.0±5.8 34.8 33.3±2.1 28.8±4.0 13.5* 136.0±18.4 113.3±19.7 16.7 11.2±1.6 7.6±0.9 32.2*

THY 14.3±3.8 10.0±2.1 30.1* 31.7±1.6 20.0±2.5 36.8* 89.5±4.6 67.0±10.6 25.1* 5.1±1.3 3.5±1.3 31.4

Average 19.5±6.0 14.4±4.7 26.1* 33.4±4.5 25.6±4.8 23.2* 118.3±15.9 111.5±15.7 5.7* 10.7±3.2 7.4±2.8 30.7*

Data represent means ± SD (n = 6). Percentage reduction followed by * indicates that the change is significantly different according to the t-test (P < 0.05).

UI: uninoculated. I: inoculated.

(+): indicates an increase in the inoculated plants compared to the uninoculated plants. For the full names of the rice varieties see Table 1.

Table 11. Screenhouse experiment. Comparison of yield-contributing traits of nine upland rice varieties and the reference lowland rice variety Thihtatyin at harvest uninoculated or inoculated with 3,000 Meloidogyne graminicola J2 per plant

Yield-contributing traits No. of plants (n) UI I Reduction (%)

No. of tillers per plant 90 5±2 a 4±1 b 14.1

No. of panicles per plant 90 4±2 a 3±1 b 15.0

No. of filled grains per panicle 90 55±21 a 37±17 b 33.4

% filled grains per plant 90 77.4±12.3 a 76.7± 14.3 a 0.8

Filled grain weight per plant (g) 90 5.7±1.6 a 3.0±1.6 b 47.6

1,000 filled grains weight (g) 90 27.2±5.6 a 24.3±5.6 b 10.9

Yield (t/ha) 90 1.9±0.8 a 1.0±0.7 b 44.9

Data represent means ±SD. Means in the same row followed by the same letter are not significantly different according to Tukey's HSD test (P < 0.05). UI: uninoculated. I: inoculated.

shows that the pathogen pressure on roots was much higher in the upland rice varieties compared with the lowland rice varieties: on average 839 vs 5,945 (g root)-1 (which is 7.1 times higher) at 8 WAI and 4,421 vs 14,552 (g root)-1 (which is 3.3 times higher) at harvest. These numbers illustrate the higher susceptibility to M. graminicola infection of the upland varieties compared with the lowland varieties. Again, one should be careful to combine the results of different experiments but this observation may also indicate that the pathogen pressure on a root system is higher in mature rice plants than in younger plants (4,421 vs 839 J2 (g root)-1, respectively, for the lowland varieties and 14,552 vs 5,945 J2 (g root)-1, respectively, for the upland varieties). Sedentary endoparasitic nematodes, such as Meloidogyne spp., complete their life cycles inside the roots and increase their root population densities until maturity of their host plants. Differences in fresh root weight at harvest between the two types of rice varieties are not at the origin of the differences in pathogen pressure observed in our study as the fresh root weights of the upland varieties were only slightly lower compared with the lowland varieties: 26.8 vs 20.4 g for the uninoculated plants and 15.8 vs 14.6 g for the inoculated plants, respectively. Meloidogyne graminicola causes galling of the root tip, thus limiting root growth (lengthening) and this, in turn, will result in an increased in root damage towards the end of the crop cycle. At 8 WAI (Win et al., 2014), the percentage reduction in root length in inoculated plants was 23.8 and 30.8% for the lowland and upland varieties, respectively, compared with 19.7 and 23.2%, respectively, at harvest.

At 8 WAI (Win et al, 2014), the root gall index of the inoculated plants of the lowland varieties averaged 5.4 vs 7.3 in the upland varieties. At harvest, the root gall indices were 4.6 and 6.1, respectively. This comparison shows that the severity of root galling caused by M. graminicola can already be high during the early growth of rice plants and shortly after nematode infection. The same observation was made during the population dynamics study reported by Win et al. (2013).

The observation that in both screenhouse experiments of our study, the percentage reductions caused by M. graminicola infection of all ten plant growth and yield-contributing traits measured (with the exception of percentage filled grains per plant for the upland varieties) were significant when all lowland and all upland rice varieties were combined indicates that M. graminicola can cause substantial damage to all these traits. It is difficult to state which traits were the least and the most affected. For both lowland and upland varieties, the lowest percentage reductions (< 12%) were observed for plant height, percentage filled grains per plant and 1,000 filled grains weight. For both lowland and upland varieties, the percentage reduction of some traits was high and similar (such as number of filled grains per panicle reduced by 30 and 33.4%, respectively) but for some other traits the percentage reduction was higher in the lowland varieties compared with the upland varieties and vice versa. Examples of the former are fresh root weight (41 vs 26.1%), number of tillers per plant (29.1 vs 14.1%) and percentage filled grains per plant (11.7 vs 0.8%). Examples of the latter are filled grain weight per plant (34.7 vs 47.6%) and 1,000 filled grains weight (10.9 vs 6.5%). Since within the two types of

rice varieties significant differences in percentage reduction of plant traits between uninoculated and inoculated plants among the varieties were also observed, we suggest that the effect of M. graminicola on the plant growth and yield-contributing traits examined might be highly genotype-dependent.

Our observation that M. graminicola caused a high percentage reduction in both types of rice varieties in fresh root weight (41% and 26.1% in the lowland and upland varieties, respectively) and dry shoot weight (28.2 and 30.7% in the lowland and upland varieties, respectively) is in agreement with several earlier reports that this nematode species affects the root development of rice plants, which in turn leads to poor water and nutrient uptake and a reduced shoot weight (Hunter, 1958; Babatola, 1984; Soomro & Hague, 1986; Soomro, 1987; Plowright & Bridge, 1990). Some of these authors also reported stunting of M. graminicola-infected rice plants but in our study plant height at harvest was reduced with somewhat less than 6% in both screenhouse experiments and with only 2.4% in the field experiment. Israel et al. (1963), Prot et al. (1994), Rao et al. (1998) and Soriano et al. (2000) have also reported a high percentage reduction in number of tillers and panicles per plant. In our study, in both the screenhouse and the field experiment, these two traits were reduced by 20 to 30% in the lowland varieties. However, in the screenhouse experiment with the upland varieties these traits were less affected (14.1 and 15%, respectively).

The average yield of the uninfected/less infected plants of the lowland rice varieties was 3.3 t ha-1 in both the screenhouse and the field experiment vs 1.9 t ha-1 of the uninfected plants of the upland rice varieties. This is in agreement with the fact that the yield of lowland rice is, on average, higher than the yield of upland rice (Galinato et al, 1999). In the screenhouse experiments, infection with M. graminicola caused on average a yield reduction with 31.1% in the lowland varieties vs 44.9% in the upland varieties. This indicates that in the upland varieties the higher nematode population densities per root unit and per root system resulted in a higher yield loss compared with the lowland varieties. In the field experiment, carbofuran treatment resulted in a 16.5% increase in yield. This is less than observed in the screenhouse experiment with the lowland varieties and can be explained by the fact that the percentage reduction of most plant traits was substantially lower in the field experiment compared with the screenhouse experiment. Only the percentage reduction in number of tillers per

plant was similar (about 30%) in both experiments. One of the most important yield-contributing traits is filled grains weight per plant and this trait was reduced by 34.7% in the screenhouse experiment vs only 8.6% in the field experiment.

In addition to the observation that yield losses are usually more marked under screenhouse/ glasshouse and microplot conditions (De Waele, pers. communication), this difference between the screenhouse and field experiments can be caused by several factors or a combination of these factors. Firstly, in the carbofuran-treated sub-plots the number of J2 per root system was reduced by about 70% and the pathogen pressure per root by about 80%. However, carbofuran-treated plants had on average 257 J2 (g root)-1 and 1,351 J2 root system-1. That still a relatively high number of nematodes were found in the carbofuran-treated sub-plots can have been caused, inter alia, by re-infestation of M. graminicola of these sub-plots. Although 2 kg a.s. ha-of carbofuran was applied before sowing and at the maximum tillering stage of the rice plants, summer rice in this region is irrigated by downstreaming water from field to field and this practice may favour dissemination of M. graminicola from infested to uninfested fields as flooding is known to be the main mean of dissemination of this nematode species (MacGowan & Langdon, 1989; Bridge et al, 2005). In addition, the efficacy of carbofuran to control M. graminicola was reported to decline 20 days after application in paddy rice (Krishna-Prasad & Rao, 1982). Also, Duxbury (2002) reported that carbofuran treatment did not control M. graminicola very effectively, allowing the nematode to re-establish in the soil and infect rice roots. Secondly, the initial nematode population density in the field was much lower compared with the nematode inoculum density in the screenhouse experiment (200 J2 (100 ml soil)-1 vs 29 J2 (100 ml soil)-1). The reduction in yield in our field experiment is in agreement with the results of a field experiment carried out by Padgham et al. (2004) in a M. graminicola-infested lowland (but rainfed) rice field in Bangladesh. These authors reported that nematicide application resulted in a yield increase of 16 to 20%. In M. graminicola-infested upland rice fields in Thailand, nematicide application resulted in a yield increase of 12 to 33% (Arayarungsarit, 1987). Based on the results of an outdoor raised-bed experiment, De Waele et al. (2013) reported that M. graminicola caused, on average, a 30% yield reduction of 19 aerobic rice varieties grown under aerobic conditions in a sandy loam soil.

Meloidogyne graminicola was the only species of plant-parasitic nematodes observed in this field during the duration of the experiment. The initial J2

soil population ranged from 3 to 110 J2 (100 ml soil)-1 with an average of 29 J2 (100 ml soil)-1. A low J2 soil population density of M. graminicola before the beginning of the summer-irrigated rice growing season in a double rice cropping sequence after harvest of the rainfed (monsoon) rice has also been observed during the population dynamics study of Win et al. (2013). Compared with the screening experiment with lowland varieties, a lower root population density of M. graminicola J2 and a lower severity of root galling at harvest were observed in the field. This was also the case for the carbofuran-treated and untreated plants in the field experiment. The low J2 soil population density in the field at the start of the summer-irrigated rice growing season multiplied to population density levels that caused considerable damage to plant growth and yield. Poudyal et al. (2005) reported that a soil population density of 0.1 J2 (g soil)-1 at the beginning of the rice growing season may result in 31% yield loss in the rice variety Masuli, while Cuc & Prot (1992) considered a M. graminicola population density of 100 J2 (g root)-1 (between flowering and maturity) as a high infection level on irrigated lowland rice. On average, in the field experiment, the M. graminicola J2 soil population increased about 50 times in the untreated sub-plots and about 10 times in the carbofuran-treated sub-plots. This observation indicates that all lowland rice varieties included in the field experiment were good hosts of M. graminicola.

The results of our study make it possible to examine the effect ofM. graminicola on the yield of an aerobic upland rice variety (Yezin Yar 9) grown under upland conditions in a saturated soil. The effect on this variety was very high since the yield was reduced by 41.8%.

The results of our study also allow the identification of rice varieties that are less sensitive or tolerant to M. graminicola infection. On the basis of the results of the screenhouse experiment no such lowland varieties could be identified because the percentage yield reduction of all 15 lowland varieties included in the experiment was higher than 20%. In the field experiment, however, the lowland varieties Sinnweyin and Shwethweyin combined a relatively low number of M. graminicola per g roots (< 1,000 J2) and a relatively high number of M. graminicola J2 per root system (2,364 and 1,398, respectively) with a percentage yield loss which was less than 10%. In the screenhouse experiment with upland varieties, the traditional upland variety Khaukphephan combined the highest reproduction factor (108.4) with the lowest percentage yield reduction (9.8%) while the percentage reduction in

yield of the other eight upland varieties was higher than 30%. This may indicate that the variety Khaukphephan is tolerant to M. graminicola infection. Although the yield of uninfected plants of the variety Khaukphephan (1.3 t ha-1) was the lowest among the upland varieties (average 1.9 t ha-1) the tolerance of this variety should be further investigated.

The results of our study also allow the identification of rice varieties that are either highly susceptible or hypersensitive to M. graminicola infection, such as for example the lowland variety Hmawbi 5 and most of the upland varieties examined. Cultivation of these varieties should be avoided on fields infested with M. graminicola, especially when no appropriate management practices, such as a 2-year rotation system for the upland varieties, are applied.

In none of the three experiments was a significant correlation observed between soil and root J2 population densities of M. graminicola and percentage yield reduction. Coyne & Plowright (2000) suggested that in rice the nematode population densities at harvest may be poor indicators of their economic importance in tropical climates since nematodes may develop rapidly and complete multiple life cycles during one crop cycle. This is certainly the case for M. graminicola, which can complete up to six life cycles during one rice growing season (Fernandez et al., 2014).

By contrast, the severity of root galling may be a good indicator of the yield loss M. graminicola can cause since in the field experiment with the lowland rice varieties and in the screenhouse experiment with the upland rice varieties a positive correlation was found between the root galling index of infected plants and the percentage yield reduction. Amarasinghe et al. (2007) reported that in rice the observation that a higher number of galls per plant induced by M. graminicola reduced plant growth and yield, resulted more from the greater disturbance to water and nutrient uptake by the root system.

ACKNOWLEDGEMENTS

This study was supported by a Flemish Interuniversity Council (VLIR-UOS) Ph.D. scholarship to P.P. Win. The authors express appreciation to the Plant Protection Division (Yangon), Myanmar Rice Research Centre (Hmawbi), Rice Division, Department of Agricultural Research (Yezin) and the Ministry of Agriculture and Irrigation, Myanmar, for the facilities and assistance in conducting the experiments.

REFERENCES

Amarasinghe, L.D., Kariyapperuma, K.A.D.P.S. & Pathirana, H.N.I. 2007. Study on approaches to integrated control of Meloidogyne graminicola in rice. Journal of Sciences of the University ofKelaniya 3: 29-46.

Arayarungsarit, L. 1987. Yield ability of rice varieties in fields infested with root-knot nematode. International Rice Research Newsletter 12: 14.

Babatola, O. 1984. Rice nematode problems in Nigeria: their occurrence, distribution and pathogenesis. Tropical Pest Management 30: 256-265.

Bos, L. & Parlevliet, J.E. 1995. Concepts and terminology on plant/pest relationships: towards consensus in plant pathology and crop protection. Annual Review of Phytopathology 33: 69-102.

Bridge, J. & Page, S.L.J. 1980. Estimation of root-knot nematode infestation levels on roots using a rating chart. Tropical Pest Management 26: 296-298.

Bridge, J. & Page, S.L.J. 1982. The rice root-knot nematode Meloidogyne graminicola on deep water rice (Oryza sativa subsp. indica). Revue de Nematologie 5: 225-232.

Bridge, J., Plowright, R.A. & Peng, D. 2005. Nematode parasites of rice. in: Plant Parasitic Nematodes in Subtropical and Tropical Agriculture (M. Luc, R.A. Sikora & J. Bridge Eds). pp. 87-130. Wallingford, UK, CAB International.

Coyne, D.L. & Plowright, R.A. 2000. Nematode threats to intensifying smallholder upland rice production in the Guinea savannah of Cote d'lvoire. Tropical Science 40: 67-74.

Cuc, N.T.T. & Prot, J.C. 1992. Root-parasitic nematodes of deep-water rice in the Mekong Delta of Vietnam. Fundamental and Applied Nematology 15: 575-577.

De Waele, D. & Elsen, A. 2007. Challenges in tropical plant nematology. Annual Review of Phytopathology 45: 457-485.

De Waele, D., Das, K., Zhao, D., Tiwari, R.K.S., Shrivastava, D.K., Vera-Cruz, C. & Kumar, A. 2013. Host response of rice genotypes to the rice root-knot nematode (Meloidogyne graminicola) under aerobic soil conditions. Archives of Phytopathology and Plant Protection 46: 670-681.

Duxbury, J.M. 2002. Sustainability of post-green revolution agriculture: the rice-wheat cropping system of South Asia. Annual report (2001/2002) submitted to the Soil Management-Collaborative Research Support Program. USA, the University of Hawaii. 43 pp.

Egashira, K. & Than, A.A. 2006. Cropping characteristics in Myanmar with some case studies in Shan State and Mandalay Division. Journal of the Faculty of Agriculture of Kyushu University 51: 373-382.

Fernandez, L., Cabasan, M.T.N. & De Waele, D. 2014. Life cycle of the rice-root-knot nematode Meloidogyne graminicola at different temperatures under non-flooded and flooded conditions. Archives of Phytopathology and Plant Protection 47: 1042-1049.

Galinato, M.I., Moody, K. & Piggin, C.M. 1999. Upland Rice Weeds of South and South East Asia. The Philippines, International Rice Research Institute (IRRI). 156 pp.

Hunter, A.H. 1958. Nutrient absorption and translocation of phosphorus as influenced by the root-knot nematode (Meloidogyne graminicola). Soil Science 85: 246-250.

Israel, P., Rao, Y.S. & Rao, Y.R.V.J. 1963. Investigations on nematodes of rice and rice soils. Oryza 3: 125-128.

Jain, R.K., Khan, M.R. & Kuman, V. 2012. Rice root-knot nematode (Meloidogyne graminicola) infestation of rice. Archives of Phytopathology and Plant Protection 45: 635-645.

Kreye, C., Bouman, B.A.M., Castañeda, A.R., Lampayan, R.M., Faronilo, J.E., Lactaoen, A.T. & Fernandez, L. 2009a. Possible causes of yield failure in tropical aerobic rice. Field Crops Research 111: 197-206.

Kreye. C., Bouman, B.A.M., Faronilo, J.E. & Llorcaz, L. 2009b. Causes for soil sickness affecting early plant growth in aerobic rice. Field Crops Research 114: 182-187.

Kreye, C., Bouman, B.A.M., Reversat, G., Fernandez, L., Vera-Cruz, C., Elazegui, F., Faronilo, J.E. & Llorca, L. 2009c. Biotic and abiotic causes of yield failure in tropical aerobic rice. Field Crops Research 112: 97-106.

Krishna-Prasad, K.S. & Rao, Y.S. 1982. Effect of few systemic pesticides as soil treatments on the growth and development of Meloidogyne graminicola in rice roots. Indian Journal of Nematology 12: 14-21.

MacGowan, J.B. & Langdon, K.R. 1989. Host of the rice root-knot nematode Meloidogyne graminicola Golden & Birchfield, 1965 (Nematology circular № 172). USA, Florida Department of Agriculture & Consumer Services. 4 pp.

Netscher, C. & Erlan, X. 1993. A root-knot nematode, Meloidogyne graminicola, parasitic on rice in Indonesia. Afro-Asian Journal of Nematology 3: 90-95.

Nombela, G. & Romero, D. 1999. Host response to Pratylenchus thornei of a wheat line carrying the Cre2 gene for resistance to Heterodera avenae. Nematology 1: 381-388.

Padgham, J.L., Duxbury, J.M., Mazid, A.M., Abawi, G.S. & Hossain, M. 2004. Yield losses caused by Meloidogyne graminicola on lowland rainfed rice in Bangladesh. Journal of Nematology 36: 42-48.

Plowright, R.A. & Bridge, J. 1990. Effect of Meloidogyne graminicola (Nematoda) on the establishment, growth and yield of rice cv. IR 36. Nematologica 36: 81-89.

POUDYAL, D.S., POKHAREL, R.R., SHRESTHA, S.M. & Khatri-Chetri, G.B. 2005. Effect of inoculum density of rice root-knot nematode on growth of rice cv. Masuli and nematode development. Australasian Plant Pathology 34: 181-185.

Prot, J.C. & Matias, D.M. 1995. Effects of water regime on the distribution of Meloidogyne graminicola and other root-parasitic nematodes in a rice field toposequence and pathogenicity of M. graminicola on rice cultivar UPLRi-5. Nematologica 41: 219-228.

Prot, J.C., Soriano, I.R.S. & Matias, D. 1994. Major root parasitic nematodes associated with irrigated rice in the Philippines. Fundamental and Applied Nematology 17: 75-78.

Ramakrishna, A. & Sharma, S.B. 1998. Cultural practices in rice-wheat-legume cropping systems: effect on nematode community. In: Proceedings of a Regional Training Course "Nematode Pests in Rice-Wheat-Legume Cropping Systems" (S.B. Sharma, C. Johansen & S.K. Midha Eds.). pp. 73-79. Hisar, India, CCS Haryana Agricultural University.

Rao, Y.S., Jayaprakash, A. & Mohanty, J. 1998. Nutritional disorders in rice due to infestation by Heterodera oryzicola and Meloidogyne graminicola. Revue de Nematologie 11: 375-380.

SOOMRO, M.H. 1987. The effects of plant parasitic nematodes and plant growth regulators on root growth of graminacious plants. Ph.D. Dissertation, the University of Reading, UK, 323 pp.

Soomro, M.H. & Hague, N.G.M. 1986. The effects of Meloidogyne graminicola on the root growth of rice

and Echinochloa colonum. Pakistan Journal of Botany 26: 441-449.

Soriano, I.R., Prot, J.C. & Matias, D.M. 2000. Expression of tolerance for Meloidogyne graminicola in rice cultivars as affected by soil type and flooding. Journal of Nematology 32: 309-317.

Tandingan, I., Prot, J.C. & Davide, R.G. 1966. Influence of water management on tolerance of rice cultivars for Meloidogyne graminicola. Fundamental and Applied Nematology 19: 189-192.

Taylor, S.P., Hollaway, G.J. & Hunt, C.H. 2000. Effect of field crops on population densities of Pratylenchus neglectus and P. thornei in southeastern Australia, part 1: P. neglectus. Journal of Nematology 32: 591-599.

Whitehead, A.G. & Hemming, J.R. 1965. A comparison of some quantitative methods of extracting small vermiform nematodes from soil. Annals of Applied Biology 55: 25-38.

Win, P.P., Kyi, P.P. & De Waele, D. 2011. Effect of agro-ecosystem on the occurrence of the rice root-knot nematode Meloidogyne graminicola on rice in Myanmar. Australasian Plant Pathology 40: 187-196.

Win, P.P., Kyi, P.P. & De Waele, D. 2013. Population dynamics of Meloidogyne graminicola and Hirschmanniella oryzae in a double rice cropping sequence in the lowlands of Myanmar. Nematology 15: 795-807.

Win, P.P., Kyi, P.P., Maung, Z.T.Z. & De Waele, D. 2014. Evaluation of the host response of lowland and upland rice varieties from Myanmar to the rice root-knot nematode Meloidogyne graminicola. Archives of Phytopathology and Plant Protection 47: 869-891.

Yoshida, S. 1981. Fundamentals of Rice Crop Science. The Philippines, International Rice Research Institute (IRRI). 269 pp.

Pa Pa Win, Pyone Pyone Kyi, Zin Thu Zar Maung, Yi Yi Myint and D. De Waele. Оценка потенциала вредоносности и возможных потерь урожая от рисовой галлообразующей нематоды Meloidogyne graminicola на различных сортах риса в Мьянме.

Резюме. В условиях теплиц, была проведена оценка возможной вредоносности M. graminicola для новых высокопродуктивных сортов поливного риса, а также традиционных и улучшенных сортов суходольного риса. Полевые эксперименты, проведенные в дельте р. Иравади, показали, что ни один из 15 сортов поливного и 9 сортов суходольного риса не был устойчив к M. graminicola. Различия в устойчивости были выявлены между отдельными сортами, и в целом между поливными и суходольными сортами. Так, показатель MTeggs (фактор умножения числа личинок 2-й стадии) составлял от 13.2 до 52.8 для поливных сортов против 39.8-108.4 для суходольных сортов. В целом, суходольные сорта оказались более чувствительными, чем поливные. Наибольшие различия между этими сортами наблюдали в снижении таких показателей как сырой вес корней (41.0 против 26.1% для поливных и суходольных, соответственно), число побегов на растении (29.1 против 14.1%), процент наливающихся зерен на растении (11.7 против 0.8%, соответственно) и средний вес зерна (34.7 против 47.6 г.). Значительные различия в показателях роста между сортами как поливного, так и суходольного риса указывают на генетическое определение уровня устойчивости. Поражение M. graminicola вызывало среднее снижение урожая в 31.1% у поливных сортов, и 44.9% - у суходольных. Обработка карбофураном приводила в среднем к повышению на 16.5% урожая у поливных сортов риса.