Characteristics of Superior Soybean Breeding Lines Tolerance to Rust ( Phakopsora pachyrhizi )

Soybean rust caused by Phakopsora pachyrhizi is one of the most important diseases which limits soybean production. The aim of this study was to e�aluate the resistance of 28 superior soybean lines and their tolerance to rust. The study was conducted at a screen house and arranged in a completely randomized design (CRD); three replications. All genotypes tested were artificially inoculated with P. pachyrhizi, and a set of un-inoculated genotypes was planted as a comparison. Number of pustules was recorded weekly, and resistant criteria was rated based on the International working group on soybean rust IWGSR method. Lesion color (LC), sporulation le�el (SL), number of uredia (NoU), frequency of pustule which had uredia, and yield were also recorded. Among 28 genotypes tested, only one was categorized as resistant and 2 genotypes were susceptible. Resistant genotypes had few pustules, lower AUDPC �alues, low disease se�erity, and Reddish Brown lesion type. Soybean rust affected yield components, i.e. number of intact pods and yield per plant. Yield loses due to rust in this study �aried from 5-89%, and the a�erage was 51%. The set of lines from Tanggamus pedigree showed more resistant to rust but less tolerant compared to Sinabung pedigree.

tance to rust may be complicated due to lack of knowledge of the host-pathogen interaction and en�ironmental factor.Therefore, non-host specific resistance and tolerant culti�ars could be an alternati�e.Indonesian Legume and Tuber Crops Research Institute (ILETRI) ha�e some superior breeding lines deri�ed from superior culti�ars, namely Tanggamus (Tgm), Anjasmoro (Anj), Burangrang (Brg), Sinabung (Sin), Argomulyo (Arg), and Malabar (Mal) to impro�e the superiority of its parental.This study aimed to e�aluate the resistance of soybean lines against soybean rust and its potential loss due to this disease.This information is crucial for supporting data in releasing a new �ariety and also important to de�elop good agricultural practices suitable for culti�ar characteristics.

METHODS
Twenty fi�e soybean genotypes consisted of 23 lines and two culti�ars (Anjasmoro and Grobogan) were planted at a screen house to e�aluate its resistance to soybean rust.Soybean seeds were sowed in plastic pots with 5 kg mixtures of soil and cow manure (2:1 �/�) with NPK fertilizers of 1:2:2 (w/w/w) (5 grams per pot).Two ml l -1 of Captan was applied at 14 and 21 days after planting (dap) to reduce soil borne pathogens.The a�erage temperatures were recorded during the study (25-28°C) as well as relati�e humidity (80-85%).
Artificial inoculation with P. pachyrhizi urediospores suspension at concentration of 104 spore ml-1 was conducted twice, at 21 and 28 dap.The uredia was har�ested from susceptible �ariety, Ringgit, which was planted four weeks prior to the study.Prior to inoculation, infected lea�es with mature urediospores were placed on wet sterile papers in trays to maintain the humidity.The infected lea�es then were incubated for 24 hours at room temperature.Inoculum was prepared by suspending the urediospores in sterile water.As many as 20µl Tween 20 was added in the suspension and mixed well and then filtered through a cheesecloth.The urediospores were sprayed on the abaxial surface of the lea�es in the e�ening using hand sprayer.The experiment was arranged in a completely randomized design (CRD) with three replications.E�ery genotype was planted in six pots and a pot consisted of two plants.A separate experiment was conducted without artificial inoculation as a check.
Response of soybean genotype to rust was e�aluated starting from se�en days after inoculation (dai).Numbers of pustules were recorded

INTRODUCTION
Asian soybean rust (ASR) caused by Phakopsora pachyrhizi is one of the most important pathogens which limits soybean production.The ASR causes significant yield loss up to 80% under hea�ily infection, depending on the susceptibility of the genotype (Twizeyimana et al., 2009).P. pachyrhizi infection also causes premature defoliation that effects the number of intact pods and yield reduction (Kumudini et al., 2008), reduces number of pods and seed sizes (Li et al., 2010).Early infection may not de�astate the crops, but may reduce the o�erall yield more than that of the late infection.
The a�ailability of resistant or tolerant cul-ti�ars is a key component of rust integrated control (Twizeyimana et al., 2009;Hartman et al., 2011).Howe�er, soybean breeding for resistance to soybean rust is complicated because of the lack of information about the host-parasite interaction (Pham et al., 2010) and the limitation of parent material resistant to rust.Tolerant culti�ars could be one of the alternati�e choices to reduce yield loss on soybean caused by the rust disease.Tolerance can be defined as the plant capacity to resist the pathogen de�elopment, without significant reduction in yield or quality of the product (Scafer, 1971;Politowski & Browning, 1978).The best technique to de�elop new culti�ars which are resistant to rust is to screen the a�ailable germplasm and to create pedigree from the existing cul-ti�ars and genotypes.
There are some important parameters to classify the tolerance of soybean to rust.Araujo & Vello (2010) mentioned that the intensity of pathogen sporulation and periodical se�erity assessment were important parameters for the classification of genotypes into tolerant or susceptible to rust.Moreo�er, Hartman et al., (2011) said that the number of uredinia per cm2 was also important besides the lesion color and types (Yamanaka et al. 2010).Sometimes, response of soybean to rust was ambiguous or partially resistant as a result of genotype response to the en-�ironmental factors, i.e. temperature, humidity, and plant's physiological factors.The �ariability in genotype reaction was also influenced by the interaction between the host genotype, soybean rust pathotype, and en�ironmental conditions (Slaminko et al., 2008).Hartman et al., (2011) reported that there was �ery high �irulence �ariability in ASR population especially in Asia as the center of origin and di�ersity of both the rust pathogen and its host soybean.
In Indonesia, soybean breeding for resis-weekly.The ASR infection le�el and soybean genotype resistance were rated using the modified three digits of International Working Group of Soybean Rust (IWGSR).Disease se�erity was also rated at 49 dap based on the leaf area co�ered by the P. pachyrhizi according to Godoy et al., (2006) methods (Figure 1).Lesion color (LC), sporulation le�el (SL), number of uredia per lesion (NoU), and frequency of lesions which had uredinia (% LU) were also e�aluated and classified into six categories in accordance to the method of Yamanaka et al. (2010), where 1 (�ery dark) to 6 (�ery light).The SL of each lesion was classified into 0 (no spores) to 3 (abundant spores) (Figure 2).Area Under Disease Progress Cur�e (AUDPC) was calculated according to Simko & Piepho (2012) to estimate the disease progress on each genotype.(Tolerance = × 100) and rust tolerance index (RTI = yield on inoculated plant / yield on un-inoculated plant) as formulated by Kawuki et al. (2003).

RESULTS AND DISCUSSION
Soybean rust symptom was signed by the appearance of P. pachyrhizi pustule on the abaxial surface of leaf started �aried from V2 stage to V6 depending on the genotypes.A�erage number of pustules at the first obser�ation, 7 days on Argomulyo/Sinabung (Arg/Sin) pedigree was the highest (26 pustules cm-1).All Sinabung pedigrees (Sin/Arg, Sin/Mal, Mal/Sin, L.Jat/Sin, and Arg/Sin) had relati�ely higher number of pustule compared to that of Tanggamus lines.These lines had a hea�y pustule density since the number of pustules was more than 16 pustules cm-2.In contrast, all Tanggamus pedigree lines (Tgm/Anj and Tgm/Brg) had less pustules and been categorized as light to medium pustule density which had less than 9 pustules cm-2 except Tgm/Anj-743.
The pustule number of genotypes tested did not consistently show the same trend on weekly obser�ation.At 21 days, when data recorded were used as justification for resistant criteria, the number of pustules on all genotypes did not show significant differences, it �aried from 3 to 8 pustules cm-2.Howe�er, P. pachyrhizi pustules had co�ered almost all the lea�es (Table 1).
In general, the increase of the number of pustules was in line with the plant's age and the duration of infection.In this study, the numbers of pustules did not constantly increase o�er period of time due to the number of pustules.This was also influenced by the en�ironmental factors such as temperature (canopy temperature), humidity (leaf wetness), and light.Medeiros et al., (2008) stated that in an experiment conducted in South Africa, the number of pustules per lesion, as well as lesion size increased with the increase of leaf wetness duration at 85% and 95% RH, and the temperature was 21-24°C.The a�erage temperature during this study was 25-28°C and 80-85% of relati�e humidity which was less fa�orable for pustule growth and de�elopment.
The de�elopment of rust disease was not only influenced by the increase in the number of pustule, but also the position of pustule or uredia on the lea�es.In this study, the pustule obser�ed at the third upper side of plant since 2 weeks after inoculation represented the de�elopment of rust disease which was relati�ely fast, in contrast to the rust se�erity and the number of pustules on all genotypes tested which were relati�ely low.The upper third of the leaf canopy had been infected by P. pachyrhizi since 14 days, and all pustules were sporulated, causing the increase in infection site and the reduction of latent period.According to IWGSR resistant criteria, of the 28 genotype tested, only 1 breeding line (Tgm/Anj 743) was categorized as resistant (R) to the rust.19 genotypes were moderately resistant, and the other 6 genotypes were moderately susceptible, whereas 2 genotypes (L.Jat/Sin-85 and Grobogan �ariety) were susceptible.None of the genotype tested was categorized as immune (Table 1).
Rust se�erity illustrated by leaf area co�ered by the P. pachyrhizi pustules showed light to medium, and the infection rates �aried from 0.6 to 25.7% (Table 1).The a�erage of rust se�erity on susceptible genotypes was 20.7%, 8.5% on moderately susceptible genotypes, and 5.9% on resistant genotypes which represented the relati�ely slow expansion of urediospores.The se�erity of the rust on Sinabung pedigrees was hea�ier than that of the Tanggamus pedigrees e�en though the highest was on Grobogan culti�ar.Rust se�erities on Sinabung pedigrees were more aggres-si�e, and the pustules co�ered broader leaf area.In this case, the effect of genetic materials which were partially resistant played an important role.Genotypes which were partial resistant expressed incomplete resistance which were characterized by pathogen colonization.This may occur in inoculated plants, but the extent of colonization was limited compared to that on fully susceptible genotypes, non-pathotype specific, and reduced selection pressure by allowing limited pathogen propagation to occur (Hartman et al., 2011).
Three lesion types were obser�ed on all breeding lines including reddish brown (RB), Tan, and mixtures of RB and Tan (Figure 3).Most of genotype tested had mixed RB and Tan lesion color while most of genotypes tested (68%) were moderately resistant to rust.In contrast to the study conducted by Kato & Yorinori (2008), resistant genotypes ha�e RB lesion, and susceptible genotypes ha�e Tan lesion, and some ha�e mixture of RB and Tan.
Responses of all genotypes tested to rust also �aried in terms of the lesion types, lesion color, sporulation le�els, number of uredia, and frequency of lesions with uredia (Fig 4).The lesion color of the genotypes tested �aried from �ery dark (1) to light (5), depending on their parental.Most of Tanggamus pedigrees had light lesion color (4 to 5 color standard categories) and Sinabung pedigrees ha�e darker color lesion, categorize as 2 to 3 on Kato & Yorinori color standard (Figure 4a).Sporulation le�el on all genotypes tested ranged from none (0) to abundant (3).Ele�en genotypes had moderate sporulation le�el (scored as 0-2 and 1-2).11 genotypes had abundant sporulation (scored as 1-3), and only 6 genotypes had less sporulation (Figure 4b).Number of uredia �aried from 0 to 5 per pustule.The highest num-ber of uredia on Sinabung pedigrees was 5, whereas on Tanggamus pedigrees were only 3 uredia per pustule (Figure 4c).In line with the number of uredia, frequency of lesion which had uredinia on Tanggamus pedigrees was lower than that on Sinabung pedigrees.Frequency of lesion which had uredinia on Tanggamus pedigree �aried from 47% to 100%, and the a�erage was 78%.On the other hand, frequency of lesion which had uredinia on Sinabung pedigree �aried from 68% to 100%, and the a�erage was 95% (Figure 4d).
Lesion color is known to be controlled by resistant genes, and it is important to obser�e the lesion color when selecting the resistant genotypes.In this study, it was difficult to group all genotypes tested into resistant criteria based on its lesion color due to the large �ariation among samples in e�ery genotype.Yamanaka et al., (2010) explained LC was easily influenced by en�ironmental factors, and it has a low correlation with other resistant characters including SL, NoU, and % LU, so it is important to complete LC data with SL, NoU, and % LU data.Spore production was considered to be more important to be identified because it represented the potency of urediospores to infect the crop.The higher le�el of sporulation, the bigger chance of disease to spread throughout the crop was.The chance of pathogen to spread out was also supported by the higher number of uredia and the frequency of lesions which had uredia.Most of Sinabung pedigrees were categorized as moderately susceptible genotypes which had higher SL, NoU, and % LU compared to those of Tanggamus pedigrees.It was illustrated that susceptible genotypes had higher SL, NoU, and % LU than resistance genotypes.
Disease progress and rust de�elopment o�er period of time could be represented by AU-DPC �alue.Area under Disease Progress Cur�e (AUDPC) �alues calculated from the number of Figure 3. Lesion type of ASR on the two sets of the soybean breeding lines, (a) Reddish Brown, (b) Tan, and (c) mixture between Reddish Brown and Tan pustules per cm2 ranged from 11.5 to 390.9 (Table 2).Susceptible genotypes had higher AUDPC �alue than that of the resistance genotypes.Ho-we�er, not all the susceptible breeding lines had high AUDPC �alues as obser�ed in Anjasmoro culti�ar.The positi�e correlations were obser�ed between AUDPC and sporulation le�el (r = 0.5), between AUDPC and number of pustules (r = 0.6), and between AUDPC and disease se�erity (r = 0.4).These correlations showed that rapid ad�ance of the disease was caused by the increase of sporulation which resulted in increasing of the number of pustules and more secondary infection sources.In line with study conducted by Cherif et al., (2010) and Maphosa, Talwana, & Tukamuhabwa (2013) AUDPC was related to factor effected resistance to rust such as lesion number, sporulation, and disease se�erity e�en tough, sometimes AUDPC �alue was ambiguous due to the fact that it was highly influenced by the en�ironment (Mueller et al., 2009;Cherif et al., 2010;).In this study, the highest AUDPC �alues did not represent the highest rust se�erity, the highest pustules, and the highest sporulation, and sometimes the low AUDPC occurred in genotypes which had high sporulation and high disease se�erity.
Soybean rust infection reduced yield and yield components such as number of intact pods, number of empty pods, and weight of seed per plant (Table 2).Number of intact pods per plant �aried from 6 to 22 pods.The a�erage number of intact pods on the susceptible genotypes was lower than that of resistant genotypes.Number of intact pods on resistant and moderately resistant genotypes �aried from 11 to 22 pods per plant, and those of the susceptible �aried from 6 to 20 pods per plant.Besides ha�ing number of intact pods, the number of empty pods was also recorded, and the result showed that the a�erage number of empty pods per plant on susceptible genotypes (0.9) was higher than that on resistant genotypes (0.7).
A�erage yield per plant was low due to the rust infection.Stress amount caused by rust infection which was illustrated by SI was 0.61.It me- ans that the disease stressor was relati�ely high.The yield �aried from 2.7 to 10.6 grams per plant on the inoculated set, and it was higher on the un-inoculated set.The yield �aried from 7 to 20.7 grams per plant.Yield loses due to rust in this study �aried from 5-89%, and the a�erage was 51%.Yield loss on Sinabung pedigree was lower than that on Tanggamus pedigree, the a�erage of yield loss on Sinabung pedigree was 9%, and yield loss on Tanggamus pedigree was 76%.Disease se�erity, as measured by the AUDPC, was positi�ely correlated to the yield loss (r = 0.7).In most Tanggamus pedigrees with low AUDPC �alues, a big differences in yield were obser�ed among infected and uninfected crops; howe�er, not all Sinabung pedigrees which had higher AUDPC �alue had lower yield losses.As described abo�e, AUDPC �alue was highly influenced by genotype x en�ironment interactions.Howe�er, in this study the influence of genotype factor was dominant since all breeding lines were culti�ated in the same en�ironment.The large �ariation was found on the AUDPC �alue.Howe�er, lack association was noticed between the area under the disease Rust Tolerance Index �alue on Tanggamus pedigree �aried from 0.11 to 0.35, lower than that of Sinabung which was from 0.73 to 1.19.Three dimensional diagram illustrated based on RTI, yield in stress plant, and yield potential was shown in Figure 5. On three dimensional diagram, all Tanggamus pedigrees were grouped into one group, which was separated from Sinabung pedigrees, whereas existing culti�ars (Anjasmoro and Grobogan) were in the same group with the Tanggamus pedigrees.This diagram illustrated that tolerance of Tanggamus pedigree are relati�ely lower than Sinabung pedigrees.Tolerance to soybean rust in genotypes tested was shown by Sinabung pedigrees, which had yield loss which was lower due to rust.It means that the yield on rust infected plants was not significantly different with those on un-infected plants.As described by Jar�ie (2009) tolerance is genotypes with yield on stressed plant which is not significantly different with it on healthy plant, as it implies that genotypic adaptation to that en�ironment.In a case of more se�ere disease symptoms, breeding lines which were categorized as tolerant showed a lower reduction of yield.
Tolerance to rust culti�ar could be an al-ternati�e since resistant culti�ar is difficult to de�elop due to big �ariability in genotype x en-�ironment interaction, rust pathogen pathotype, and the lack of information about genetic materials which contain resistant genes.In this study, the set of breeding lines from Tanggamus pedigrees showed more resistant to rust compared to the set of Sinabung pedigrees, ne�ertheless, the yield loses on Tanggamus pedigrees were higher than those of Sinabung pedigrees.Breeding lines crossed from Tanggamus and Sinabung culti�ars could be proposed to de�elop new culti�ars resistant to rust.
Soybean rust has widely spread in Indonesia, but the study of characteristic of genotypes resistant and tolerant to rust is still limited.The �ariation in reaction patterns of Indonesian soybean genotypes would be different if it was compared with soybean from other places, representing that they might contain new sources of rust resistance that useful to create a new culti�ar resistance to rust.

CONCLUSION
Resistance to moderately resistant soybean genotypes ha�e characterized by low number of pustules, lower AUDPC �alues, low disease se�erity and had Reddish Brown lesion type.Among 28 genotypes tested, only one genotype (Tgm/ Anj-743) was categorized as resistant, whereas 19 genotypes; most of them were Tanggamus pedigree, were moderately resistant.6 genotypes; most of them were Sinabung pedigree, were categorized as moderately susceptible, and 2 genotypes (L.Jat/Sin-85 and Grobogan) were susceptible.Tolerant genotypes showed by Sinabung pedigree (Sin/Arg-8, Sin/Malabar-16, Sin/Malabar-19, Arg/Sing-34, Arg/Sin-47, Argo/Sin-52, Mal/ Sin-66, Mal/Sin-68, L.Jat/Sin-85, and Arg/ Sin-98) had lower yield loses and higher tolerant index.E�en though the set of superior breeding lines from Tanggamus pedigree showed more resistant to rust it was less tolerant compared to Sinabung pedigree.Breeding lines crossed from Tanggamus and Sinabung culti�ars could be proposed to generate new culti�ar that was resistant to rust.

Figure 2 .
Figure 2. Standards of lesion color (A) and sporulation le�el (B)(Yamanaka et al., 2010) Yield components, including number of pods, number of empty pods and the weight of seed per plant were recorded from both inoculated and un-inoculated sets.Stress index (SI) was calculated using formula of SI =

Figure 5 .
Figure 5. Three dimensional diagrams illustrated based on RTI, yield in stress plant, and yield potential on soybean lines

Table 2 .
The Area under Disease Progress Cur�e (AUDPC) �alue, number of pods, number of empty pods, and yield per plant of two sets soybean breeding lines and the resistance of the genotypes tested to rust.