Amorphophallus muelleri blume diversity based on morphological characteristics in South Sulawesi

Porang is an alternative carbohydrate source plant that contains the highest glucomannan among other Amorphophallus species in Indonesia. Amorphophallus muelleri Blume is an important species of the genus Amorphophallus, which belongs to the same subspecies A. bulbifer and A. yuloensis. A. muelleri Blume has good disease resistance and heat tolerance tendency, and high KGM content. Kinship relationships can be used as a basis for assembling more potential genotypes. To find out the kinship between existing varieties we must exploration and identify. This research was carried out using qualitative methods analysed in the numerical taxonomy and multivariate analysis system version 2.02i (NTSYS) program. The genetic materials used in this study were explored from wild growing habitat in agroforestry areas in South Sulawesi, Indonesia. The results obtained that the MR6 and BL10 genotypes which had a similarity coefficient value of 1.00, between genotypes MR4, MR9, MR10, BT4 and BT10 had a similarity coefficient value of 1.00, between genotypes BT5 and BT7 have a similarity coefficient value of 1.00, and between genotypes BT1, BT2, BT9 have a similarity coefficient value of 1.00. Furthermore, BL1, BL4, and BL6 contribute significantly of the variation as showed by PC1.


Introduction
Porang (Amorphophallus muelleri Blume) is an alternative carbohydrate source plant that contains the highest glucomannan among other Amorphophallus species in Indonesia [1].Porang were first recorded in the Andaman Islands in India.Then it is predicted to spread eastward through Myanmar and Thailand to Indonesia [2].In Indonesia, Porang grows wild and is cultivated in some province such as: Sumatra, Madura, Bali, West Nusa Tenggara (NTB), and mostly in Java [3].The distribution of Porang in East Java includes Nganjuk, Madiun, and Bojonegoro regencies in wild forests and agroforestry areas [4].Agroforestry is a method of growing trees on agricultural land.Agroforestry was first pioneered by the Canadian International Development Centre, which is an institution tasked with identifying development priorities in the forestry sector in developing countries in the 1970s.The identification 1230 (2023) 012133 IOP Publishing doi:10.1088/1755-1315/1230/1/012133 2 results show that forests in developing countries have not been used optimally, including Porang plants that grow wild in the forest [5].
Amorphophallus muelleri Blume is an important species of the genus Amorphophallus, which belongs to the same subspecies Amorphophallus bulbifer and Amorphophallus yuloensis.Among the three species, only Amorphophallus muelleri Blume has a variety with a chromosome number of 2n = 26, which could possibly be hybridized with Amorphophallus konjac and Amorphophallus albus.Amorphophallus muelleri Blume has good disease resistance and tendency to heat tolerance, and high KGM content.Therefore, Amorphophallus muelleri Blume is an important source of germplasm for genetic improvement of other plant subspecies such as Amorphophallus konjac and Amorphophallus albus [6].
Porang germplasm has high genetic diversity, so it is necessary to increase the benefits of Porang plants through plant breeding programs.One example is through the analysis of kinship based on morphological.The kinship relationships can be identified based on the similarity of characters with the assumption that different characters are caused by differences in genetic composition [7].This research is to study the genetic diversity of Porang plants in several agroforestry areas in South Sulawesi and know similarities Porang plant genotypes in different area based on morphological characters.

Sampling method
The sampling technique used in this study is simple random sampling, which is a method of taking samples from members of the population using a completely randomized design without regard to the level in the population members on 10.000 meters square area.The sample of Porang plants used was in the vegetative growth phase with a tuber weight less of 110.5 gr.Sampling was documented using a camera application that was supported by the coordinates feature.Furthermore, all samples of Porang plants were observed for morphological characters.

Morphological character observation
Observation of morphological characteristics was carried out the descriptors.Observations are consisting stem color, stem surface, leaf surface, leaf blade shape, number of leaflets, leaf edge color, leaf color, stem position, outer color of stem tubers, bulbil outer color, color of tuber flesh, bulbil flesh color, color of tuber chips, stem tuber shape, stem tuber surface, bulbil surface, bulbil position, bulbil shape, stem tuber tissue, and bulbil tissue.

Data analysis
The data obtained from the results qualitative data in the descriptor than transform to binary data.Similarity coefficient was calculated using the Unweight Pair Group Arithmetic Method (UPGMA) to Estimation of genetic similarity is obtained using NTSYS-pc analytical software version 2.02i (Rohlf, 1998).Clustering and multivariate analysis was performed by morphological data using the online tool ClustVis [8].

Results and discussion
Recapitulation of the morphological characters of thirty Porang genotypes (Table 1) in the leaf color section had the highest percentage in light green color of 50%, and the lowest percentage in yellowish green color of 16.67%.Stem color had the highest percentage in light green at 46.67%, and the lowest percentage was in green with white spots at 23.33%.Bulbil flesh color obtained the highest percentage in the yellow color of 96.67% and the lowest percentage in the orange color of 3.33%.The outer color of the tubers had the highest percentage in yellow-brown color of 63.34% and the lowest percentage in brown color of 13.33%.The highest percentage of color in tubers was white-yellowish at 43.33%, and the lowest percentage was at yellow at 26.67%.The leaf color had quite a variety of colors: green, light green, and yellowish green (Table 1).This difference is thought to be due to the difference in the total of chloroplasts in each plant.Based on the research that 30% shaded stands are more maximal than the photosynthetic process that occurs in 80% shaded stands [9].The sunlight that enters the shaded stand is 30% greater than the sunlight in the 80% shaded stand which tends to be little because the Porang plants are blocked by plants as canopy.The diversity of leaf color that occurs is influenced by the process of photosynthesis, the sampling location in Bantaeng district there are small amounts of shaded stands include Tectona grandis, Cinnamomum cassia, and Aleurites moluccana.It made Porang has dark green leaves.Furthermore, yellowish green due to the intensity of sunlight with large amounts.Based on Maros, there are large numbers of shaded stands include Tectona grandis and yellow Bambusa vulgaris so that Porang has a light green leaf color due to the intensity of sunlight in small amounts.Meanwhile, the exploration in Bulukumba Regency contained a small number of shaded stands include Syzygium aromaticum, Tectona grandis, Dypsis lutescens, and Bambusa maculate.It made Porang has a light green leaf due to the intensity of little a total of sunlight.
Character variations of A. muelleri were also found based on the color of the stems observed.Based on the results of observations, two color variants of Porang stems were found consist leaves with rhombus-shaped patterns, and rhombus-shaped patterns with linear lines [7].Meanwhile, that the stems of A. muelleri have purplish smooth bark and white spots [10].In addition, that A. muelleri has green stems with large prismatic patterns, green trunks with small prismatic patterns, and green trunks with striped large prismatic patterns [11].In this study, it was found that variant I (Group I) was Porang which had green stems with white spots (Table 1).While Varian II (Group II) is A. muelleri which has stems with light green to dark green colors (Table 1).
Bulbils appear from the leaf axils, but not all leaf axils produce bulbils.Bulbil flesh has a variety of colors consist yellow and orange (Table 1).The color of the tubers of A. muelleri leaves is light yellow.That the color in the tubers of A. muelleri leaves is yellow [12].The bulbil flesh is light brown [13].In this study, there was a difference in the color of the bulbil flesh in one out of thirty genotypes (3.33%), it happened because one genotype observed was a side bulbil, not a main bulbil.It grows more slowly so that the main bulbil has a larger size than the side bulbils.The main bulbil is a bulbil that is on the main stem under the axils of the leaves, so that the photosynthetic results stored in the main bulbil are very large.Photosynthetic results stored in the main bulbil have a higher glucomannan content than the side bulbils, although not significantly [14].Tissues that have a high glucomannan content are indicated in yellow.Based on this study, it was found that one out of thirty genotypes (3.33%) had an orange color in the side bulbils, thus indicating lower glucomannan content in the side bulbs compared to the main bulbs in twenty-nine genotypes (96.67%).The genetic relationship of thirty genotypes based on morphological characters (Figure 1) has a similarity level at a coefficient of 0.81 -1.00.The results of the analysis of the level of the coefficient of similarity of the genotypes of Porang were grouped into 2 groups.Group I consisted of 15 genotypes (MR6 and BL10) and (MR4, MR9 and MR10), (MR1, MR5, MR2, BL9, BL3, BL8, MR3, BL7, MR7, and MR8) both was had coefficient 1.00 and 0.85-0.94 of similarity coefficient, respectively.Group II consisted of 15 genotypes (BT4 and BT10), (BT5 and BT7), (BT1, BT2, and BT9) which had a similarity coefficient of 1.00.Furthermore, (BL1, BL4, BL6, BL2, BL5, BT3, BT6, and BT8) which had a coefficient of similarity between 0.83 and 0.94.
The data matrix plot based on morphological descriptors was subjected to principal component analysis (PCA) for estimating genetic differentiation among 30 genotypes of Porang are showed in Figure 2. The scatter plot based on these components disclosed a pattern of four groups.The plot shows that the genotypes, BL1, BL4, and BL6 were distinctly different from each other.Calculating principal components was using singular value decomposition (SVD) iteratively until estimates of missing value converge [15].X axis and Y axis showed principal component 1 (PC1) and principal component 2 (PC2), both was explained 25.6% and 16.8% of the total variance, respectively.These results indicate that the genotypes, BL1, BL4, and BL6 contribute significantly of the variation as showed by PC1.Then, BT1 was the main component of the PC2.Distribution of genotypes did not according to geographical origin that showed in the matrix plot of thirty genotypes.It same in other research in thirty-seven morphological characters of sesame did not reflect their geographical origin [15].The 30 genotypes were grouped into five major clusters based on heat map and dendrogram (Figure 3).Cluster I consist eight genotypes (BL7, MR10, BL3, BL8, BL9, BL10, MR2, and MR6).Cluster II consist eight genotypes (MR5, BL2, MR7, MR8, MR4, MR9, MR1, and MR3).Cluster III consist four genotypes (BT8, BT10, BT4, and BL5).Cluster IV consist six genotypes (BT1, BT5, BT2, BT9, BT3, and BT7).Cluster V consist four genotypes (BT6, BL1, BL4, and BL6).

Conclusions
BL1, BL4, and BL6 contribute significantly of the variation as showed by PC1.Then, BT1 was the main component of the PC2.Distribution of genotypes did not according to geographical origin that showed in the matrix plot of thirty genotypes.