Morphological and Physiological Characterization of Cassava Genotypes on Dry Land of Ultisol Soil in Indonesia

Indonesia has a large cassava diversity, but the tolerant cultivars on drought areas have not been well recorded. Candidate mapping can begin with morphological and physiological characterization. This study aimed to map cassava’s genetic diversity, determining the key phenotype to distinguish genotypes, physiological adaptation, and high-yield candidates under environmental stress. A total of 29 genotypes were clustered into 5 groups. A specific group for genotype from same site was not found. The differences and relations among genotypes were very clear, demonstrating cassava’s genetic diversity in Indonesia. The key group characteristics are upward petiole orientation (G1), nine lobes (G2), prominent foliar scars (G3), winding lobe (G4), and elliptic-lanceolate (G5). A total of 19 genotypes had a number of storage root > 10 storage roots, 20 genotypes had a weight of storage root > 2kg/plant, and 3 genotypes had > 4kg/plant. Morphological and physiological trait determination is relevant to contribute to high-yield cassava breeding in dry areas. The morphological characteristics of well-adapted plants were plant height, lobe characteristics, and petiole orientation, while the physiological traits were chlorophyll index, transpiration rate, and photosynthesis rate.


Introduction
Cassava is an important food source in Indonesia. It is cultivated on over 1.3 million hectares, both on mono and mixed crops [1], gradually spreading into the less densely populated areas in Indonesia [2]. Cassava tubers have a high nutritional content, mostly from carbohydrates, which varies depending on the speci c plant part (root or leaves), geographic location, variety, plant age, and environmental conditions [3]. Carbohydrates comprise about 32%-35% of its fresh weight and 80%-90% when dry. Starch is composed of 80% carbohydrates, 3% amylopectin, and about 17% amylose, although it also contains low quantities of sucrose, glucose, fructose, and maltose [4].
So far, cassava production is dominated by small farmers with low productivity because of the low use of improved cultivars and fertilization in areas which are not always environmentally apt. In addition, climatic changes such as water de cit present looming challenges to food production, even when cassava is considered tolerant to water de cits [5].
Adaptable cultivars are an opportunity to produce highyield cassava despite these conditions. Indonesia has a large cassava diversity, both in farmer-cultivated land and research elds. However, this diversity has not been well explored to nd high-yield cultivar candidates in dry areas.
is mapping of potential candidate can begin with the morphological and physiological characterization of cassava genotypes, as already shown for genetic clustering [6][7][8].
Morphological characteristics are important attributes that contribute to gas exchange and plant metabolism, physiological characteristics, that can increase the yield. erefore, morphological, physiological, and yield characteristics are important selection criteria in breeding programs [9]. e present study aimed to characterize the genetic diversity of cassava in Indonesia, pinpoint the key phenotype to distinguish among genotypes and physiological adaptation signal to environmental stress, and determine a highyield candidate in environmental stress conditions.

Site and Climate.
e research was conducted in Jonggol Teaching and Research Farm, Department of Agronomy and Horticulture, Faculty of Agriculture, Bogor Agricultural University. e site is located at 6˚28ʹ12″S 107˚01ʹ47″E. Figure 1 shows the climatic conditions during the research. e monthly averages of temperature and relative humidity were 21.4°C and 86.3%, respectively.

Experimental Design.
e eld experiment was arranged in randomized complete block design with ve replications, and the factor, i.e., cassava genotypes, consisted of 29 genotypes. Cassava genotypes were collected from both farmers and Indonesia Legumes and Tuber Crops Research Institute (ILETRI) ( Table 1).

Photosynthesis Rate Measurement.
e photosynthetic rate was estimated using LI-COR (LI-6400XT Portable Photosynthesis System, LI-COR Inc., Lincoln, NE, USA). e estimation was conducted to the tenth leaf from shot, ve plants per genotype, and measured between 10:00 and 12:00 am.
2.6. Chlorophyll Content Measurement. Chlorophyll content was estimated using the chlorophyll content meter (CCM-200 plus, Opti-Sciences Inc.). e measurement was conducted on the tenth leaf from shoot on the tagged plant, ve plant per genotypes, and measured between 10:00 and 12:00 am.

Data Analysis.
Before data analysis, data were separated into a vegetative and a generative dataset and examined for outliers. Data are expressed as the mean ± standard error and were analyzed statistically using Minitab® Statistical Software Ver.18. Cluster analysis was performed using PBSTAT (http://www.pbstat.com/). e cluster analysis for plant genetic diversity used both quantitative and qualitative data. Clustering used Gower dissimilarity mode and neighbor joining clustering method. e correlation among characteristics was analyzed using the Pearson correlation using opensource R statistic (0.05).

Morphological Characteristics.
is morphological characterization aimed to highlight the variations and identify germplasms for plant breeding [11]. Table 3 provides the details of 14 morphological characteristics, including the color, pattern, and shape for leaf, petiole, lobe, and stem for 29 cassava genotypes. e corresponding data were interpreted on frequency (%; number genotypes/29 genotypes; Table 4). Generally, leaf characteristics were light green color apical leaves, absent pubescence on apical leaves, average leaf retention, lanceolate and elliptic-lanceolate shape of central leaflet, light green leaf color, and green color leaf vein. Color of apical leaves was recorded light green (65.5%), purplish (6.9%), and dark green (27.6%). No genotype showed pubescence on apical leaves. e leaf retentions were average (72.4%) and better than average (27.6%). e shapes of central leaflet were elliptic-lanceolate (37.9%), obovate-lanceolate (13.8%), and lanceolate (48.3%), while the leaf colors were light green (65.5%), dark green (17.3%), and purple-green (17.2%), and color of leaf vein was green (93.1%), reddish green in less than half of the lobe (3.4%), and all red (3.4%).
Almost all genotypes had purple petiole color and horizontal orientation, smooth lobe margin, and seven lobes. e petiole color distribution was purple (65.5%), yellowish green (17.2%), greenish red (13.8%), and red (3.4%), while the orientation of petiole is both horizontal (72.4%) and inclined upwards (27.6%). e lobe margin was smooth (65.5%) and winding (34.5%), while the number of leaf lobes was seven lobes (58.6%) and nine lobes (41.4%). Stem phenotypes are large, prominent of foliar scars, silver and green-yellowish stem exterior, medium-long distance between leaf scars, and straight kinds of growth habit. Prominence of foliar scars was both prominent (79.3%) and semiprominent (20.7%). e color of stem exterior was silver (44.8%), green-yellowish (37.9%), and others. e distribution of distance between leaf scars was short (41.4%), medium (51.7%), and long (6.9%), while the kinds of growth habit were straight (93.1%) and zigzag (6.9%). Table 5 provides the quantitative evaluation of the characteristics of the 29 genotypes. Malang 1 and UJ 5 were the tallest that are about 280 and 270 cm, respectively. In contrast, the smallest, about 160 cm, was identified on Barokah. Twenty genotypes have a plant height of more than 2 m, while about 9 genotypes have height of less than 2 m. Ten genotypes were recorded with no branching, while 19 genotypes were recorded to have branching. e height of first branching range about 0.2-2 m, whereas the highest first branching was recorded on Adira 1 about 2 m, followed by Kubar about 1.1 m, and other genotypes height of first branching mostly were obtained about 0.2-0.6 m. e shortest petiole length was recorded on Kunir, Genjah Urang, and Vati 2 (about 27 cm). In contrast, the longest petiole length was recorded on Malang 4, Daplang Gureh, and Mangu (about 44 cm). Leaf lobe characteristics including length, width, and ratio were recorded as about 17-25 cm, 4-7 cm, and 3-5 cm, respectively.

Genotype Clustering.
Morphological classification is important in plant breeding to emphasize the variability and relationships between genetic lines. Accessions sharing many similarities are closely related [11]. Conversely, accessions showing many differences show distant relationships [12]. e characterization of genetic diversity by morphological characteristic is a cheap and proven method [6,7] that also shows a significant correlation to agronomic traits that could be used to evaluate the potential production and can be used in plant breeding [8].
Group 1 showed similarity on most qualitative characteristics, except for PC, LC, and CLV. Similarly, group 2 showed similarity on most qualitative characteristics, except  International Journal of Agronomy 5 CSE, DLS, CAL, SCL, and OP. Group 3 shared all qualitative characteristics, except CAL, LR, and CSE. In contrast, in group 4, similarities were found only for PAL, CLV, and OP characteristics and in group 5 on PAL, LC, CLV, GHS, and NLL. To distinguish among groups, ≥1 characteristics need to be used. However, the number of leaf lobes could be a specific characteristic to differentiate for group 2, prominence of foliar scars for group 3, petiole orientation for group 1, lobe margin for group 4, and shape of central leaflet for group 5.
Despite the lack of specific groups in same site genotypes, differences and relationships among genotypes were clear, indicating the genetic diversity of cassava germplasm in Indonesia. Nevertheless, lobe characteristics, prominence of stem, and petiole orientation can distinguish the genotypes in each group. e key characteristics of groups include inclined upward orientation of petiole (G1), nine lobes (G2), prominent foliar scars (G3), winding lobe (G4), and elliptic-lanceolate (G5). e large variations of quantitative characteristics were evaluated among the genotypes. e
Physiological changes follow morphological changes in plants responding to environmental stress [14]. As reported by [15], CI can serve as adaptation indicator to environmental stress. CI is involved in gas exchange functions such as photosynthesis and transpiration, and [16] considered it as a possible indirect indicator of photosynthetic capacity. Environmental stress could signi cantly decrease CI, indicating impaired photosynthesis [17]. It is di cult for plant to escape the environment in which they are currently growing, so they must passively adapt to changing or even adverse conditions; therefore, evolution can occur in the process of their long-term adaptation [18].

Yield Characteristics.
Yield characteristics consider both storage units and storage root weight. Figure 4 shows both characteristics for all genotypes. e highest weight corresponded to UK 1 (5 kg) and the lowest to Manalagi (1.1 kg), while Darul Hidayah (14) had the highest number of storage roots and Daplang G (6) the lowest. Eight genotypes had a storage root weight >3.5 kg plant −1 , namely, Ketan, IR, Kunir, Mangu B, UK 1, UJ 3, Malang 1, and Genjah Bayam.    is indicates that some genotypes can be potential candidates for dry land conditions. Under water de cit conditions, almost all cassava genotypes resulted in <10 storage roots and about 5-8 root storage plant −1 [19]. In addition, 20 genotypes had >2 kg plant −1 storage root, with 3 having >4 kg plant −1 . Since, the weight of storage root of cassava in dry land is usually 1-1.8 kg plant −1 [19]. is result shows the characteristic candidates of high production in dry land.

Correlation among Characteristics.
e correlation among characteristics was analyzed to determine relationships between characteristics. Two signs are positively correlated if one increases quantitatively and other decreases quantitatively or if both decrease [12]. e correlation among characteristics was low to moderate ( Figure 5).   Moderate correlations were found for HP * PL, HP * PR, CI * PR, CI * WSR/P, and NSR/P * WSR/P, while low correlations were identi ed for HP * TR, PL * TR, PL * CI, PL * NSR/P, PL * WSR/P, CI * TR, and CI * WSR/P. Vegetative growth as indicated by both HP and PL a ects photosynthesis and transpiration, whereas higher plant will increase photosynthesis, and a longer petiole would increase transpiration. Since CI a ects photosynthetic, TRs further support to increase the weight of storage root. Negative correlations between transpiration rate and yield and photosynthesis rate and NSR/P were found as well.
Positive and moderate correlations among CI, photosynthesis rate, transpiration rate, and storage root weight could be considered storage root accumulations. e correlation of plant growth to physiological activities [20], plant growth to yield characteristics [21], and physiological      activities to yield [22], among yield characteristics [23], had been reported. Morphological and physiological description is an important first step for successful high-yield cassava breeding for dry areas. e morphological characteristics of well-adapted plants were plant height, lobe characteristics, and petiole orientation, and the physiological traits were CI, TR, and photosynthesis rate.

Conclusions
e initial 29 genotypes were clustered in 5 groups. A specific group for genotype from same site was not found. e differences and relations among genotypes were very clear. It indicates the presence the genetic diversity of cassava germplasm in Indonesia. e key characteristics of groups were upward petiole orientation (G1), nine lobes (G2), prominent foliar scars (G3), winding lobe (G4), and ellipticlanceolate (G5). Nineteen genotypes had >10 storage roots, and 20 genotypes had >2 kg/plant of storage root, of which 3 had >4 kg/plant. e morphological and physiological traits finding was very important to early data to contribute to the successful high-yield cassava breeding for dry areas. e morphological characteristics were plant height, lobe characteristics, and petiole orientation, while the physiological traits include CI and photosynthesis rate.

Data Availability
All the data used to support the findings and conclusions of this study are included within the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.