Soybean germplasms evaluation for acid tidal swamp tolerance using selection index

Availability of fertile land on the island of Java in Indonesia decreases due to the shifting from agricultural land to non-agricultural land. Hence, an extensification of soybean culture to outer Java suboptimal land areas is needed, such as tidal swamp which occupies approximately 20.192 million hectares. The main limitations in this soil are soil acidity, Fe toxicity and excess water. To develop soybean varieties tolerant to acid tidal swamp, tolerant soybean gene resources are needed. Hence, glasshouse and field experiments were carried out to identify tolerant gene resources. The glasshouse experiment has been conducted using 185 genotypes of germplasm at the Indonesian Legume and Tuber Crops Research Institute, Malang, East Java. Selection was carried out by using a selection index method. The glasshouse experiment was followed by field experiment at the Belandean research station, Banjarbaru, South Kalimantan, using the best 17 genotypes selected from the glass house trial. Results showed that there was variability of response of each genotype to acidity and Fe toxicity. Therefore, assessment of soybean tolerance to acidity and Fe toxicity should be conducted by root growth. Based on selection index criteria, varieties of Lawit and Menyapa served as check tolerant varieties and showed lower growth than the 17 selected genotypes. In the field experiment, genotype MLGG 1087 was identified as the most tolerant and can serve as a gene resource tolerant to acid tidal swamp because it has the highest relative root growth on root dry weight, and the highest average of root and shoot dry weight.


Introduction
In an effort to fulfill soybean demand which reaches more than 2 million tons per year, growers encounter constraints related to the narrowing of fertile areas in Java island due to conversion of agricultural land area to nonagricultural land.This is reflected in a reduction in soybean planting areas from 1995 to 2004 by 6.1% per year and a decrease in soybean harvesting areas during the period 2000-2004 at a rate of 9.7% per year.Because of the decrease in harvested areas, the 2004 national production only reached 0.71 million tons. 1 Hence, the fulfillment of soybean demand should be conducted through both intensification and extensification.The decrease in agricultural areas in Java Island suggested that extensification only can be conducted on the outer Java Island.However, the attempts made in the outer Java Island have mainly been with suboptimal land.One of the constraints in this suboptimal land is soil acidity, because the soils are dominated by ultisols, oxysol, and histosol.In Indonesia, some of the lands classified as suboptimal encompass acid soil and swamp areas.Indonesia has very wide swamp area, i.e. 33.4 million hectares; consisting of 20,192 million hectares of tidal area and 13,283 million hectares of waterlogged area. 2 Tidal swamp is a marginal land and a fragile ecosystem.Its suitability for agricultural commodities in the original natural condition is classified as marginally suitable (S3) and moreover included in the category presently not suitable (N1).Many plant growing constraints on tidal area caused by saturated and anaerobic rhizosphere (mud, reduction reaction, oxygen and H2S gas deficiencies), existing of pyrite or sulfidic materials, Al, Fe, Mn toxicities, very acidic reaction of soil, and low natural fertility (P, N and K deficiencies, and poor bases). 3On land where pyrite has overcome oxidation, in aerobic conditions such as on shallow pyritic swamp areas, many problems will be presented; some of them are Al, Mn and ferry (Fe 3+ ) toxicities because of low soil pH, P deficiency, and perhaps also other macro nutrients deficiencies.While, in water logged conditions, when the water table reaches the surface or during rain, increasing pH caused by a reduction process leads to decreasing Al and Mn toxicities but increasing Fe 2+ , H 2 S, CO 2 and water soluble organic acids toxicities. 4 Beside soil chemistry problems, an problems with water availability also have important role.Linkemer et al. 5 reported that soybeans' early vegetative period and early reproductive stages were most sensitive to waterlogging.Further, it is reported that fewer pods per reproductive node lead to a decrease in pod production and finally yield reduction.At least two days of waterlogging during the final vegetative phase can reduce seed yield up to 18%; while at an early reproductive phase it can decrease seed yield up to 26% 6,7 also reported a significant negative correlation of flooding to the number of pods and seed yield.This case correlated with plant capacity to survive in oxygen deficiency condition. 8he low pH in soil due to the high H + ions concentration can affect plant growth and development by an H + ions toxic effect.However, the direct effect of H + ions on acid soil is difficult to determine, because in low pH conditions this correlates with Al, Mn, Fe and other mineral toxicities which are soluble in high concentrations, and essential nutrient availability such as Ca, Mg, Mo, and P which are in suboptimal condition. 9,10Acid soil effects showed in soybean by different root growth responses, where tolerant genotypes had higher root length, and susceptible genotypes had suppression on growth root. 11On an Al-tolerant soybean cultivar, Schenkel 12 reported that decreasing the nutrient solution from pH 6 to 4.3 decreased total dry matter per pot up to 63% over the cultivar's average.
The complexity of problems on acid tidal swamp areas suggested a step by step problem solving approach.The main problem on acid tidal soil in South Kalimantan is Fe toxicity, so this problem needs to be the main solution by utilizing tolerant gene resources which serve as material in developing a tolerant variety.

N o n -c o m m e r c i a l u s e o n l y
Iletri collection, including two adaptive tidal swamp varieties, Lawit and Menyapa, as controls.These 185 genotypes are not only from native germplasm but also from overseas introductions.Factorial randomized compelete block design with two replications were used in this experiment.The first factor was low pH and consisted of two levels, i.e. a) aerated waterlogging with pH 7 and b) unaerated waterlogging with pH 4.0.The second factor was genotype, and consisted of 185 genotypes.Aeration was proposed to provide root respiration and to avoid hypoxia and anoxia.
Seeds from each genotype were sterilized with 1% sodium hypochlorite for ten minutes, rinsed in deionised water and then germinated in sand with a nutrition solution containing 30 mg N L -1 (NH 4 NO 3 ), 1 mg P L -1 (NaH 2 PO 4 2H 2 O), 30 mg K L-1 (K 2 SO 4 7H 2 O), 2 mg Fe L -1 (FeSO 4 7H 2 O), 0.5 mg Mn L -1 (MnSO 4 4H 2 O), 0.5 mg B L-1 (H 3 BO 3 ), 0.2 mg Zn L -1 (ZnSO 4 7H 2 O), 0.01 mg Cu L -1 (CuSO 4 5H 2 O), and 0.005 mg Mo L -1 ((NH 4 ) 6 Mo 7 O 24 4H 2 O). 13 The acidity of the solution for sand culture was controlled at pH 7.After five days, seedlings were transplanted into aerated water logging with pH 7 or unaerated waterlogging with pH 4.0.As long as plants were in normal solution, the solution was aerated using an aerator to avoid hypoxia and anoxia.During the stress treatment, the solution was not aerated and the pH of the solution was controlled between 4.0±0.1.The nutrition concentration was held at the previous concentration.
Plants were observed at age 0 and 7 days after transplanting (DAT).After rinsing with deionised water, plants were divided into root and shoot.Each organ was weighted after dried in 80°C for 48 hours.Relative growth of each organ was measured according to Nursyamsi 14

Field evaluation
Field evaluation was conducted at the Belandean research station, Banjarbaru, South Kalimantan from May to July 2007, using randomized complete block design with two replications.The tested genotypes in field evaluation were selected from glasshouse experimental evaluation.Stress treatment was conducted at one level, i.e. acid waterlogging.Land for field evaluation was chosen that was suitable for the research goal to obtain soybean germplasm tolerant to acid tidal swamp.Before planting, land was cleared from weeds.Following clearing, a pool was prepared of approximately 1 m depth.Water irrigation was made from the pool with water from the tidal swamp.Water was flooded into the evaluation pool for a minimum of two weeks before soybean planting was conducted, in order to obtain an evaluation location that was appropriate with the natural condition of a tidal swamp.Similar planting was conducted with the glasshouse evaluation, where seeds were sowed into sand and then transplanted into treatment field when aged 5 days.Observation was carried out at age 0 and 14 days.

Glasshouse evaluation
Research was conducted by previously sowing the seeds in sand.After five days plants were transplanted into the treatment pool.Of the 185 soybean seed genotypes that were tested, 5 genotypes could not grow in neutral and acid stress treatment: MLGG 0713, MLGG 0812, MLGG 0816, MLGG 0836 and MLGG 0932.It was assumed that the 5 genotypes were sensitive to waterlogging or sensitive to transplanting conditions.Lack of oxygen was considered the main problem associated with waterlogging. 15n this research, waterlogging treatment with and without aeration had an affect on plant growth, especially on leaf chlorosis.However, the effect of waterlogging did not affect the plants' life.Boru et al. 16 reported similar results while, during two weeks, the treatment without oxygen (100% gas N2) or 80% reduction in oxygen had no impact on plant survival.[18]

Relative growth
The rate of relative weight root growth varied from -0.64-13.50g.week -1 with an average of 1.29 g.week -1 (Table 1).Negative values showed that root growth of sensitive genotypes was terminated, the root was broken so the mass was decreased.This condition also occured with other observed parameters.On shoot measurments, such as shoot dry weight, hypocotyls length, epicotyls length and plant height, the negative value was probably due to the shrinking of mass or length of plant, as the result was shifted to other plant parts.On peanut, Kuskopf et al. 19 stated that plants were allocated more N to roots than to shoots for amelioration of environment stress.This increased some organs and decreased other plant parts.While on root parameters, a negative value occured due to the brittleness of plant organ, so weight and size of root become lower than the early growth.
Genotypes of MLGG 0731 and MLGG 0511 had highest rate of relative root dry weight growth, i.e. 13.50 g.week -1 and 11.21 g.week -1 , respectively (Table 2).On the other genotypes the rate of root dry weight growth was not as high as these 2 genotypes.While on shoot dry weight character, MLGG 0647 demonstrated the highest rate of growth reaching 10.03 g.week -1 .Other rates of growth such as root, hypocotyls and epicotyls length, and plant height did not have same direction with weight root growth and shoot at the same genotypes.It indicated that there was no linkage among observed characters.The implication of such gene resources in plant breeding was to join the traits by crossing many parents with specific traits.
Root dry weight observation was the major observation concern in this research.Under the stress condition, it was seen that roots in all tested genotypes did not grow normally.This result suggested that the successes of treated stress condition, and genotype having good root were selected.Based on passive observation, it was seen that soybean was tolerant to acid waterlogging not only shown on root but also through all plant quantitave characters.The symptoms of toxic poisoned plants were chlorosis and leaf twisted 20 and also hypocotyls and epicotyls shorterning.Leaf chlorosis was caused by transports of gibberellins and cytochinine acids to leaf, where these occurred to protect leaf membrane integrity.Reducing nitrogen transport from leaf to root also caused chlorosis as a result of decreasing protein and chlorophyll syntheses. 21 transport was very important because accu-  o n l y mulation of N was associated with plant agronomic improvements.22 Beside, the oxidative stress also caused the decrease of total formed chlorophylls.23 Hence, relative growth observation conducted not only on root but also on shoot dry weight, root length, hypocotyls length, epycotyls length and plant height.

Selection index
To evaluate tolerant plants to acid waterlogging stress based on those six characters, selection index 24 was used (with weighting relative growth of root dry weight relative: root length: shoot dry weight: hypocotyls length: epicotyls length: plant height = 1:1:0.75:0.5:0.5:0.5).Based on the six characters, selection index criteria was arranged as follows: Y = 0.2223X 1 + 0.1830X 2 + 0.0405X 3 + 0.0224X 4 + 0.0262 X 5 + 0.0415X 6 where: X 1 = relative growth of root dry weight X 2 = relative growth of root length X 3 = relative growth of shoot dry weight X 4 = relative growth of hypocotyl length X 5 = relative growth of epicotyl length X 6 = relative growth of plant height According to the selection index formula with limit of selection criteria over 0.75, 17 identified genotypes tolerant to acid waterlogging were obtained (Table 3).Genotype MLGG 0731 was on the first rank with an index value of 3.055 followed by MLGG 0511 and MLGG 0170, while genotypes on 15, 16 and 17 ranks were MLGG 0238, MLGG 0061 and MLGG 0592.These 3 genotypes' index selection value were 1.099; 0.962; 0.781 respectively, where values of index selection were lower than MLGG 0731 and MLGG 0511 which was on the first and second rank (3.055 and 2.459).Hence, those 2 genotypes can serve as gene resources for acid waterlogging tolerance.
The performance of 5 varieties (Lawit, Malabar, Menyapa, Panderman, and Sinabung) is provided in Table 3.Of those 5 varieties, Lawit was more tolerant than the other 4 varieties with a selection index value of 0.6006.This was followed by Panderman, Sinabung, Malabar and Menyapa.Lawit and Menyapa varieties in this research were commercial tolerant varieties, because these 2 varieties were released as tol-erant varieties in tidal acid swamp.However, in this research, Lawit and Menyapa were not included in the best 10% according to the selection index.This showed that the gene resource being evaluated was more tolerant (than Lawit and Menyapa) to acid waterlogging and is still numerous in the Indonesian germplasm collection.If compared between 2 tolerant varieties, Lawit and Menyapa with selected gene resources in this research, it was seen that Lawit and Menyapa had already died (7 day), while genotypes MLGG 0731 and MLGG 0511 were still living in acid waterlogging stress conditions.Tolerance differentiations among genotypes was caused by root character differentiation, e.g.differentiation of increasing pH in root area. 25In acid tolerant soybean genotypes, the increasing pH of the rhizosphere was faster than the sensitive genotypes. 26This was part of apoplasmic tolerance mechanism, i.e. plant ability to exudate or control toxic element such as Fe into hairy root cells.This mechanism included organic acid exudation, pH increasing of rhizospher and lignifications 27 which appeared on each tested genotype.

Field evaluation
The field evaluation was carried out in a similar way to the glasshouse evaluation using formerly sowed seed soybean in sand media.After five days, plants were transplanted into treatment media.The differences between the glasshouse evaluation and this field evaluation was that stress treatment was longer (14 days).When the field evaluation was conducted, water acidity degree analysis showed pH 3.5.So this evaluation showed that selected genotypes from the field will have higher tolerance than glasshouse selected genotypes.
According to root dry weight criteria, MLGG 1087 genotype had first rank position with an average root dry weight 0.024 g; followed by MLGG 0511 and MLGG 0731 with rate 0.016 and 0.013 g, respectively.But based on root length average, genotype MLGG 0592 had first rank followed by MLGG 0511 and MLGG 0009, while root length of MLGG 1087 was shorter than the average of all genotypes.On shoot dry weight characters, genotype MLGG 1087 also ranked first with an average of 0.072 g, where this value was higher than two times the average of all genotypes.For hypocotyls length characters, MLGG 0238 had the longest hypocotyls (18.12 cm) followed by MLGG 0592 (17.84 cm) and MLGG 0257 (17.68 cm).Genotype of MLGG 1087 with highest shoot dry weight had lower hypocotyl length than average genotypes.It means that MLGG 1087 had hypocotyl diameter higher than other genotypes (Table 4).Genotypes MLGG 1087 had the highest shoot dry weight.This may be due to a) this genotype can eliminate Fe through oxidation in rhizosphere and exudation of Fe from plant tissues, or b) Fe is inactivated in plant tissues.Wasiaturrohmah 26 reported the occurrance of faster pH increase of identified Fe tolerant genotypes in the rhizosphere.A similar case was found in rice, where differences among varieties in excretion of OH-lead to pH differences in rhizosphere.A variety exudating more OH -could increase pH and absorb less Fe.On the other hand, a variety exudating less OH - could not increase soil pH leading to a greater absorbtion of Fe. 28 Differences in Fe absorbtion in root and leaf by plant depend on the concentration.Mehraban et al. 23 stated that up to 250 mg/L Fe content accumulated in root was higher than in leaf, but in concentrations of 500 mg/L Fe content in leaf was higher than in root.

Conclusions
Waterlogging affected leaf chlorosis on plants, but did not cause plant death; whilst Fe toxicity and media acidity affected growth of tested genotypes.The variety Lawit was more tolerant than the variety Menyapa, which indicated that Lawit was more suitale as check variety for acid tidal tolerance.Based on the selection index on the rate growth, 17 tolerant genotypes were identified, where they had relative growth higher than varieties of Lawit and Menyapa.Genotypes MLGG 1087 and MLGG 0511 were identified as the gene resources tolerant to acid tidal swamp based on root dry weight characteristics.The tolerant genotype MLGG 1087 was also supported by the highest average of root and shoot dry weight.
There is a need to screen tolerant gene resources using other germplasm collections to find other soybean genotypes tolerant to acid tidal swamp in order to increase the probability of developing soybean varieties tolerant to acid tidal swamp.Genotype MLGG 1087 can be used as an acid tidal swamp tolerant gene resource since it has the highest root and shoot dry weight.
as follows: where: RG = Relative growth DW pH4 = Dry weight at pH 4 treatment DW pH7 = Dry weight at pH 7 treatment DW T0 = Dry weight at 0 day after transplanting