Evaluation of Yield and Pigment Content of Eleven Genotypes of Green and Purple Pod Yard Long Bean (Vigna unguiculata (L.) Walp.)

The consumption of yard long bean in Indonesia is high and it has been increasing continuously, but the production and harvest area has been decreasing. New, superior long bean varieties with higher productivity are required to meet the increasing demands. The aim of this study was to evaluate the morphological and yield characters of selected purple and green long bean genotypes. The research was conducted from February to May 2018 in the Madiun district, East Java, Indonesia. The experiment was laid out in a randomized complete block design with genotypes as the single factor treatment, replicated three times. Eleven yard long bean genotypes were tested, consisting of four new genotypes, F7-0130144U-16-1-1, “F7-013014-4U-16-1-2”, “F7-0130144U-16-1-3”, and “F7-013014-7P-4-1-1”, and seven control genotypes, “KP13”, “KP14”, “KP Putih China”, “KP Putih China”, “Borneo”, “Sabrina” and “Parade”. The new genotypes, “F7-013014-4U-16-1-1, “F7013014-4U-16-1-2”, “F7-013014-4U-16-1-3”, and “F7-013014-7P-4-1-1” had at least one superior characters compared to the control genotypes, i.e. earlier fl owering and time to harvest, longer leaves, greater pod weight, longer pods, larger pod diameter, and higher contents of anthocyanin, carotene, chlorophyll a, chlorophyll b, and total chlorophyll.


Introduction
Long bean consumption in Indonesia in 2016 was 1,158 million tons per year, ranked 4 th of the most consumed vegetables (BPS, 2016). The average long bean production and harvest area had decreased signifi cantly from 2009 to 2014, with an increase in productivity (Directorate General of Horticulture, Ministry of Agriculture, 2015). The increase productivity of long bean is due to the uses of superior varieties with higher productivity (Syukur et al., 2015). Superior long bean varieties were selected through plant breeding to improve the growth characters, yield, and adaptability (Syukur et al., 2015).
One of the desired characters in long bean breeding is to increase the pod pigment content, including anthocyanin, chlorophyll and carotenoids. Anthocyanin is classifi ed as an antioxidant , whereas carotenoid content is related to vitamin A activities (Syahputra et al., 2008). Carotenoid content in long bean correlated signifi cantly with chlorophyll a, chlorophyll b and total chlorophyll content (Basrowi, 2017). The purple pod long bean are more tolerant to pests and diseases due to their thicker and harder skin pods which may repel pests, and can adjust well to limited soil water environment . Higher nutritional content in long bean pods will increase the consumer demands for long beans (Ardian et al., 2016).
The breeding of long beans is intended to produce varieties with a higher biomass, better nutritional content with acceptable taste, more attractive shapes and colors, and longer freshness (Syukur et al., 2015). The breeding lines long bean needs to be evaluated to determine the productivity and adaptability before releasing new long bean varieties. The yield evaluation is one of the plant breeding stages after selection which includes the preliminary evaluation, and an advanced evaluation (Syukur et al., 2015). According to Kuswanto (2005), long bean yield evaluation was determined based on fresh pod yield, pod quality and taste. This study was aimed evaluating the morphological and yield characters of purple and green long bean's genotypes as compared as compared to the existing long bean varieties and the parental genotypes, i.e. "KP13", "KP14", "KP Putih China", "KP Ungu China", " Borneo", "Sabrina" and "Parade".

Materials and Methods
This research was conducted in Madiun Regency, East Java with an altitude of 60 m above the sea level and located in -7.535 LS and 111.656 BT from February to May 2018, with monthly rainfall of 409. 2, 440.8, 498.8 and 195.7 mm per month, respectively. The maximum temperature ranged from 27.8 to 29.5 °C with the relative humidity of 88.5 to 91.1% (BMKG, 2018).
Plant beds were prepared and added with 1 kg of manure and 200 g of dolomite per plot; mulch was applied one week before planting. Seeds were directly sown into the plots using a spacing of 50 cm x 50 cm. Plant maintenance includes weed control, fertilization with NPK Mutiara 16-16-16, tying up the tendrils, watering, and pest control when necessary. NPK fertilizer was diluted at 5 g per liter of water with a volume of 250 ml per plant, applied once a week. Harvesting was conducted at three day intervals according to the criteria used for fresh bean consumption, i.e. when the end of the pods have been fully fi lled, but before seeds in the pods became too large and protruded out of the pods.
Scoring was conducted on stem diameter (cm), length of a leafl et from trifoliate leaves (cm), length of petiole (cm), time to fl ower (DAP), time to harvest (DAP), pod length (cm), pod diameter (cm), seeds per pod, 1,000 seeds weight (g), pod weight (g), number of pods per plant (g), pod weight per plant (g), productivity (t.ha -1 ), pod content of anthocyanin, carotene, chlorophyll a, chlorophyll b and total chlorophyll. Measurements were made on 10 randomly selected sample plants from each plot when the plants had entered the generative phase. Pigment content of the long bean pods were analyzed using the Sims, DA and Gamon method at IPB Postharvest Laboratory (Sim and Gamon, 2002). Samples for pigment content analysis were collected from two randomly selected pods of each genotype. Data were tested by ANOVA using SAS 9.0. Signifi cant differences between means were further tested using Duncan Multiple Range Test (DMRT) at α 5%. Correlation and cluster analysis used single linkage, Euclidean distance with Minitab 18.
Soil analysis was tested at Laboratory of Soil Science and Land Resources Department, Faculty of Agriculture, IPB. Soil samples were collected from three different spots which were then mixed into one sample. Determination of particle size distribution, pH, organic-C, total N, electrical conductivity, water-soluble cations and anions, cation exchange capacity, free Fe oxides, available N, P and K, available Fe, Mn, Cu, Zn, Mo and B in soil used the methods described in Hardjowigeno (2004).

Result and Discussion
Analysis of variance of the 11 long bean genotypes is in Table 1. The highly signifi cant characters were time to fl ower, time to harvest, pod weight, pod length, pod diameter, number of seeds per pod, the content of anthocyanin and chlorophyll b. Characters with signifi cant results were the number of pod per plant, productivity, chlorophyll a, total chlorophyll and carotenoid content, while other characters were not signifi cantly different (Table 1).
The coeffi cient variance (CV) shows the extent of environmental effects on the growth parameters, as also reported in Pradipta et al. (2010). Time to harvest character had the smallest CV value, whereas the largest CV value was in pod chlorophyll a content. Table 2 shows that the long bean genotypes fl owered at 29 to 39 day after planting (DAP), which were quite early compared to those reported in Basrowi (2017). The white pod genotype had the earliest time to fl ower, whereas the green pod genotype had the longest time to fl ower. "F7-013014-7P-4-1-1" had the earliest time to fl ower, i.e. three days earlier compared to the introductory genotype "KP Putih China". "Sabrina" and "Parade" were the latest time to fl ower. The time to fl ower of "F7-013014-4U-16-1-3" was earlier than "Sabrina" and "Parade". "F7-013014-4U-16-1-2" had the similar time to fl ower with of "KP 14" and KP Putih China, and signifi cantly earlier than "Sabrina" and "Parade". The growers usually preferred crops with early fl owering (Putri et al., 2015). Time to fl ower and time to harvest in this study were four and six days earlier than those reported in Basrowi (2017). According to Septeningsih et al. (2013), the difference in time to fl ower was affected by genetic and environmental factors. Cahyaningrum et al. (2014) reported that heavy rain conditions can lead to fl ower drops. Rainfall in Basrowi's research (2017) was 78-116.8 mm higher than the research in Madiun.
The median values of pod length were 40.16 to 56.55 cm ( Table 4). The longest and shortest pod was "F7-013014-4U-16-1-2" and "KP Putih China", respectively. The pod length of "F7-013014-4U-16-1-2" and "F7-013014-4U-16-1-3" were not signifi cantly different from the comparative genotypes of "Borneo", "Parade", "Sabrina" and "KP 13", but were signifi cantly longer than the other genotypes. The pod lengths of all test genotypes were signifi cantly longer than some control genotypes. One of the main criteria in the selection of superior long beans varieties was the character of pod length . Consumers prefers long bean with pod lengths of 60 to 90 cm (Kuswanto et al., 2009). The length of pod can affect the weight of pod (Cahyaningrum et al., 2004), but longer pods do not necessarily had more seeds (Ardian et al., 2016).
The pod diameter ranged from 5.16 to 6.67 cm. The pod diameter of "F7-013014-4U-16-1-2" was signifi cantly larger than those of the other genotypes, while "KP Putih China" was signifi cantly smaller ( Table 4). The result showed that the white pod such as "KP Putih China" and "F7-013014-7P-4-1-1" had pods with a relatively smaller diameter than the green and purple pods.
The number of seeds per pod ranged from 14 to 19 seeds. The control genotype of "KP 13" had the greatest number of seeds per pod, which was signifi cantly more than all test genotypes. "F7-013014-4U-16-1-3" had the least number of seeds. "F7-013014-4U-16-1-2" had the fewest number of seeds per pod but it had the longest pods among the other genotypes. "KP 13" and "F7-013014-4U-16-1-3" although had a signifi cantly more number of seeds per pod, had similar pod length.

Number of Pods per Plant, Pod Yield per Plant, and Productivity
The number of pods per plant ranged from 14.2 to 30 ( Table 5). The number of pods of "F7-013014-4U-16-1-1" and "F7-013014-4U-16-1-3" was not signifi cantly different from those from the control genotypes, therefore the yield of the tested genotype was comparable to yield of the existing commercial varieties. The number of pods per plant of "F7-013014-7P-4-1-1" was signifi cantly lower than "KP 14", but it was not signifi cantly different from the number of pods of the other control genotypes. The number of pods for the tested genotype "F7-013014-4U-16-1-2" was signifi cantly lower than "KP Putih China", "KP 13", "KP Putih China" and "KP 14" (Table  5). This might have been caused by aphid (Aphis craccivora) infestation which was observed across all the replications so many of the pods from this genotype were not harvestable.
The productivity of the 11 long bean genotypes ranged from 9.10 to 15.07 t.ha -1 ( Table 5). The yields of long bean in this study were smaller than the potential yields reported by The Indonesian Ministry of Agriculture and Basrowi (2017). Based on the description of long bean varieties released by the Ministry of Agriculture the yield of " Parade" (2006), "Borneo" (2010) and "Sabrina" (2012) was 12 to 25 t.ha -1 , 18 to 19.2 t.ha -1 and 20 to 24 t.ha -1 , whereas in this study they were 11.47 t.ha -1 , 15.07 t.ha -1 and 11.47 t.ha -1 , respectively. The yield evaluation study conducted in Bogor, West Java, (Basrowi 2017) reached 22.99 t.ha -1 whereas in Madiun, East Java, the highest productivity was 15.07 t.ha -1 . The differences in the yield in the two locations were possibly due to the environment effects that were greater than the genetic effects, and interaction of both. Different environment might result in differences in the metabolic processes, including photosynthesis rate, which eventually affected yield (Cahyaningrum et al., 2014).
The low yields in this study were likely caused by Sclerotium rolfsii which had attacked since the crops were at the early generative phase. The disease infestation was made worse by the saturated soil environment due to high rainfall intensities in February to April 2018. Too wet media causes sub-optimal plant growth due to less oxygen availability (Hendriyani, 2009). In addition, soil Fe in the experimental site was quite high (Table 6). In rice, high concentration of soil Fe reduced growth, especially in the number of tillers and shoot dry weight (Lubis and Noor, 2012).
A similar study conducted in Bogor (Basrowi, 2017) showed that "F7-013014-7P-4-1-1" had the highest productivity of 22.99 t.ha -1 , whereas in this research the yield only reached 9.33 t.ha -1 . The difference in the productivity was likely due to the different harvest criteria, because the pods of "F7-013014-7P-4-1-1" hardened and the seeds enlarged quickly upon ripening.The swollen pods are heavier, which in turn increased productivity.

Chlorophyll a, Chlorophyll b, Total Chlorophyll, Carotene and Anthocyanin Content of Long Bean Pods
The chlorophyll a in "F7013014-4U-16-1-2" seeds was signifi cantly higher than that of the other genotypes,and chlorophyll b in "F7013014-4U-16-1-2" and "F7013014-4U-16-1-3" was signifi cantly higher compared to the other genotypes (Table 7). "KP Putih China" had the lowest content of chlorophyll b.The high content of chlorophyll a or chlorophyll b indicates effi ciency in photosynthesis (Mescht et al., 1999).
The total chlorophyll of "F7013014-4U-16-1-2" seeds was the highest and it was signifi cantly higher than those from the other genotypes. The total chlorophyll of "F7013014-4U-16-1-3" was not signifi cantly different from the control genotypes " Borneo","Parade","Sabrina" and "KP 14", but it was signifi cantly higher than the other genotypes. The genotype of "KP Putih China" had the lowest chlorophyll content.
The genotype with the highest carotenoid content was "F7-013014-4U-16-1-2", whereas the lowest was in "KP Putih China". The carotenoid level of "F7-013014-4U-16-1-3" seeds was not signifi cantly different from "Borneo", but it was signifi cantly higher than other genotypes. There were 10 types of carotenoids in vegetables that play important roles in human's health, including improving visual function, immune system and resistance to infection (Merdekawati and Susanto, 2009). Mortensen (2006) reported that beta carotene is a precursor of vitamin A. Table 7 shows that long beans with purple pod have high anthocyanin content, whereas the whitishgreen pods had the lowest. "KP 14" had the highest anthocyanin content, whereas "KP Putih China" had the lowest. The high anthocyanin content in "KP 14" correlated positively with the darker color of the pods. A similar result was reported by  that the intensity of purple seeds correlates with anthocyanin content.
The anthocyanin content of "F7-013014-4U-16-1-1" and "F7-013014-4U-16-1-2" was signifi cantly lower than "KP 14" and "KP Putih China" but higher than the other genotypes. The result of this study was different from the fi ndings of Baswori (2017) where the anthocyanin content of "F7-013014-4U-16-1-2" was higher with a difference of 0.0152 mg.g -1 . The different result was possibly due to the long duration of transport when shipping samples to Bogor prior to pigment analysis. Sample shipment took about two days; during which the samples were not stored in cold storage. According to Hayati et al. (2012), anthocyanins stability is infl uenced by temperature and storage environment. The high temperatures cause anthocyanin degradation that effect on the loss of color in anthocyanins, e.g. longer storage duration causes a decrease in anthocyanin color and stability of the petals extract of rosella fl ower (Hibiscus sabdariffa L.).
The soil base saturation of this study was 23% (Table 6). According to the Soil Research Center (1983) in Hardjowigeno (2010) soil base saturation of >20 % was classifi ed as low category. The content of C-organic, N, P, K, Na, Mg, Ca, and CEC in the study was low according to Hardjowigeno (2010). Soil with low saturated base usually had high levels of leaching, so the cations (Ca ++, Mg ++, K +, Na +) of the soil were low (Hardjowigeno, 2010). The soil Fe and Mn in this study was 85.18 and 95.69 ppm (Table 6), which according to the Soil Research Center (2005) is classifi ed as very high. High soil Fe content is likely due to the ferrolysis which occurred due to the reduction and oxidation processes as a result of repeated fl ooding and drying (Hardjowigeno, 2004). High soil Mn content is usually found in poorly drained and inundated land (Makarim, 2005) which is similar to the soil condition in this study. High content of Fe and Mn could reduce plant productivity (Makarim, 2005).
Pod length, pod diameter, the content of chlorophyll a, chlorophyll b, total chlorophyll and carotenoids were highly correlated with the weight per pod. Pod with long and large size would have a greater weight per pod compared to short and small pods. Similar result was reported by Cahyaningrum et al. (2014) that pod weights were affected by pod length. The heavier pods also contain higher chlorophyll a, chlorophyll b, total chlorophyll and carotenoid. This result is inversely proportional to the character of the number of pods per plant which had a negative correlation to pod length, pod diameter, weight per pod, and level of chlorophyll a, chlorophyll b, total chlorophyll, and carotenoid. A larger number of pods per plant will have less weight of seeds per pod, smaller pod size and pod diameter, and a lower content of chlorophyll and carotenoids.
The pod length character does not correlate signifi cantly with number of seeds per pod, but  correlated with chlorophyll and carotene content. The length of pods did not affect the number of seeds per pod. These results are different from Basrowi (2017) who reported that the length of pod affected the number of seeds per pod. The character of pod weight per plant correlated signifi cantly (~100%) with productivity, as productivity values are the result of conversion from pod weight per plant.
The total chlorophyll content correlated signifi cantly with chlorophyll a, chlorophyll b and the carotene content. According to Pradnyawan et al. (2015), chlorophyll a and chlorophyll b are parts of the chlorophyll which will affect the photosynthesis. The total chlorophyll content had no signifi cant effect on anthocyanin content, whereas the anthocyanin content did not show any signifi cant correlation with all characters. Pods with high total chlorophyll content will likely to have high content of chlorophyll a, chlorophyll b and carotene (Pradnyawan et al., 2015).

Cluster Analysis
Cluster analysis was used to determine grouping based on the closeness in the character or inter character relationships (Suketi et al., 2010). The distance between clusters shows the high order between clusters (Vailapalli et al., 2014). Figure 1 show the similarity based on 16 quantitative characters of long bean genotype which ranges from 67.54 to 95.60%. The highest similarity level was 95.60%. The highest similarity level of 95.60% was between "KP 13" and "F7-013014-4U-16-1-3". According to Hadi et al. (2014), this high percentage indicates the similarities of a number of characters between the variables. This high similarity was possibly because the "KP 13" was one the "F7-013014-4U-16-1-3" parents. The difference in the character between the two genotypes was only on the number of seeds per pod, the content of chlorophyll a and b and total chlorophyll.