Variance analysis and heritability estimation of growth and yield parameter of sweet cassava promising clones in two environments

Planting a good variety is important thing in increasing the productivity of cassava’s productivity. Selection and evaluation are needed for development of new variety. The evaluations were done in Malang (Entisol soil) and Probolinggo (Alfisol soil). The evaluation used an RCBD with three replications. Plants were planted on rows with distance 100 cm between rows and 80 cm within rows. Twelve clones (8 promising clones and four released variety) were used in this evaluation. Results can be seen that the variance of genotype was bigger than the variance of genotype x environment interaction for fresh tuber yield, starch content, long of tuber, diameter of tuber, tuber number/plant in eight months and plant height in 4 months, means genetic potency of clones tested more important compared to effect of genotype x environment interaction. This result indicated that clone tested have low chance to be a specific adaptation. The genotypic variance and the heritability for fresh tuber yield, diameter of tuber and tuber number/plant in eight months in Malang were higher than that in Probolinggo.


Introduction
Cassava is important crops in Indonesia.The harvesting area of cassava in Indonesia was 666,186 hectares in 2021 [1].Indonesian farmer's average land ownership is 0.3 hectare, so around 2,220,620 farmers planted the cassava in 2021.Planting cassava requires around ten labourers per hectare, so around 6.7 million were involved in planting and harvesting of cassava.Cassava is used as raw material for industries.Many industries used cassava as a raw material; among them are cassava starch industry, feed industries, and food industries.So many labourers are involved in planting, harvesting, and processing of cassava.
Most cassava are planted on Alfisol and Entisol soil.Around 45% of cassava plants in Indonesia are planted on these soils [2].These soils are in Java.The soil parent material of Alfisol soil is limestone.Acidity of Alfisol soil is diffrent from Ultisol soil, so the kind and dynamics of nutrient may be also different from ultisol soil.In general, Alfisol soil is productive soil.Base cation status is from medium to high, in general; texture is generally profitable for plants [3].Entisol soil has light texture, and capacity to hold the water is low; the content of organic matter is low; availability of P and K is also low, and the soil is easy to be erosion [4].The Percentage of sand on the topsoil of Entisol soil (79%) higher than Alfisol soil (5.5%) [5].
Many farmers get their income from cassava plantations.Cassava tuber has the potential for food diversity, industry raw material for, and for feeds.Productivity is an important parameter for farmers.The availability of new varieties, which are superior to the existing varieties, is expected to increase productivity.There are steps in the development of new varieties, one is selection, either single plant selection or single raw selection.Related to selection, it was reported that removing the bad genotype IOP Publishing doi:10.1088/1755-1315/1312/1/012052 2 was better than selecting good genotype on initial selection [6].Indirect selection for tuber yield through harvest index was more effective compared to direct selection based on tuber yield, especially during early stage.The harvest index had a high heritability value every selection stage consistently.On the other hand, biomass and tuber yield had low heritability.It was reported that the diameter of the tuber, number of tuber/plant and harvest index can be used as criteria of selection for cassava selection for high yield [7]. From cassava breeding activities before, there are some promising clones.A promising clone is defined as clone that shows potential for certain desirable traits.The promising clones were from the original variety UJ5 and local variety (Cecek Ijo).Mutation breeding was used for the development of promising clones.Promising clones need to be tested in some environments before releasing those clones as a new variety.This paper reported the result of variance analysis and estimation of heritability of growth and yield parameters for cassava and their performance in two environments.

Experimental design and cassava cultivation methods
A randomized Complete Block Design was used in these evaluations with three replications.Plot size was 500 cm x 480 cm with distance between rows 100 cm.Plants were planted with distance 80 cm within row. 100 kg urea plus 100 kg Sp-36 and 100 kg KCl per hectare were used in these evaluations.

Measurement of parameters and data collection
Data for Fresh tuber yield, starch content, starch yield, long of tuber, diameter of tuber, number of tuber/plant and plant height in four months were observed.Plants were harvested in eight months.The gravity system was used as basis for data of starch content.Data for long of tuber, diameter of tuber and number of tuber/plant data used tubers with weight more than 200 gram.The diameter of the tuber was measured in the middle of tuber.

Data analysis
Analysis of data used MSTAT-C and excel software.Analysis of variance was done with assumption that genotype factor and locations were random, so combined variance analysis including expected means square is presented on Table 1, while variance analysis each location is presented on Table 2 Table 2. Variance analysis for response variable Y(i) for one environment.
The estimation of variance component for 1 location were calculated as bellows: ŝ 2 e = M3 , as estimation of error variance (V 2 E) ŝ 2 P = ŝ 2 g + ŝ 2 e/3, as estimation of phenotype variance (G 2 P )

Variance component and heritability
The Genotype Variance (ŝ2G), the variance of genotype x environment interaction (ŝ2GE), the error variance (ŝ2e) for quantitative characters is on Table 3. Genotype variance was bigger than the variance of genotype x environment interaction for tuber yield, starch content, starch yield, long of tuber, diameter of tuber and plant height in 4 months.The importance of the genetic potency of cassava clones exceeds the impact of genotype x environment interaction.This statement implies that the tested promising cassava clones have a low likelihood of specific adaptation in certain environments.The same result was reported for cassava for characters of starch yield, fresh tuber yield, number of tuber/plants, harvest index, starch content in 6 and 9 months, also starch content in 9 months, and leaf area/plant in 6 months [9].Kasno [10] also reported the same thing for groundnut for characters of a total pods, number of filled pods, number of seeds and weight of 100 seeds.Adie [11] also reported the same for soybean for characters seeds weight/plant, weight of 100 seeds, plant height, number of filled pods, number of empty pods, productive internodes, and days to mature of pods.The genotypic variance for fresh tuber yield, diameter of tuber and tuber number/plant in eight months in Malang were higher than that in Probolinggo (Table 4).The ratio of genetic variance to phenotypic variance is called broad heritability.Heritability is one of them which is the most important parameters in plant breeding [12].Information of this value is important for breeding program Planning.Classification of heritability is high (more than 0.5), medium (0.2-0.5) and low (less than 0.2) [13].Priyatno [14] and Miftahorrachman & Sulistyowati [15] also used this classification.This classification classified tuber yield, starch content, tuber long, tuber diameter was classified as high heritability.Starch yield and tuber number/plant) were classified as medium heritability.Heritability was also used as basis for selecting the selection location.A good location for selection is the location that can give the high heritability.The heritability for fresh tuber yield, diameter of tuber and tuber number/plant in eight months in Malang were higher than that in Probolinggo.This means Environment in Malang was better than environment in Probolinggo for doing selection.The fresh tuber yield is an important characteristic for farmers.This character is one of the characters used as basis for variety selection.The Interaction of genotype and environment was not significantly difference for the fresh tuber yield (Table 5).The implication was a good clone on Entisol soil would generally be good in Alfisol soil.However, another experiment showed interaction between genotype and environment for tuber yield [16,17,18,19,20].The genotype factor was significantly difference for the fresh tuber yield (Table 5).This difference indicates that the fresh tuber yield of the tested clones varied significantly between at least two of the clones under consideration.The fresh tuber yield in 8 months of tested clones ranged from 29 to 52.83 t/ha with mean 42.21 t/ha (Table 6).Variety UK1 Agritan as a check produced the highest tuber 52.83 t/ha, however this check has a taste slight bitter in roasted tuber.Related to the tuber yield, the promising clone CId150-152 had the same yield as UK1 Agritan statistically, and clone UJ5d50-35 had the same yield with CI d50-152 statistically.The Gajah variety as check produced the lowest tuber in 8 months.However, Shanti and Nirmala [21] reported that Gajah variety produced tuber in 9 months in Kalimantan Timur 43.91 t/ha, equal to UJ5.Starch content is also an important characteristic for farmers.The price of fresh tuber with high starch content will be higher than that of fresh tuber with low starch content.Starch content determines the starch yield.There was a positive correlation between starch and fall rain during 6-9 months after planting.However, the correlation was negative during the 1 or 2 months before harvest [22].The interaction of genotype and environment was not significantly different for the starch content (Table 5).
The implication was clone with high starch content in Malang will be high starch content in Probolinggo generally.It was reported that there was interaction between genotype and environment for starch content [23].The genotype factor was significantly differed for the starch content (Table 6).Starch content ranged from 17.35% to 20.87% with a mean of19.16%.UJ5d50-55 as promising clone had the highest starch content (20.87%).UJ5 as a check had a 20.27% starch content.It was reported that UJ5 had high starch content.However, the taste of roasted tuber was bitter [24].The promising clone UJ5d50-35 had a 20.57% starch content; it was same with UJ5d50-55 statistically.This clone had the same fresh tuber yield statistically with UK1 Agritan in Malang, the taste of roasted tuber of UJ5d50-35 was not bitter.The multiplication between fresh tuber yield and starch content characterizes starch yield.The interaction of genotype and environment was not significantly different for the starch yield (Table 5).The implication was a good clone in starch yield in Malang will be good in Probolinggo.However, it was reported that there were interactions between genotype and environment for starch yield [25,26].The genotype factor was significantly differed for the starch yield (Table 6).The starch yield of tested clones was significantly differed between minimal two tested clones (Table 6).The starch yield in 8 months of tested clones ranged from 5.322 to 9.228 t/ha with mean 8.05 t/ha (Table 6).Variety UK1 Agritan as a check produced the highest 9.228 t/ha.However, this check has a taste slightly bitter in roasted tuber.Regarding the starch yield, the promising clone, UJ5d50-35 had the same yield as UK1 Agritan statistically.
The promising clone UJ5d59-35 produced 50.58 t/ha; this value was statistically equal to the top value in Malang.This clone had high starch content (20.57%); it was also equal to the highest value statistically.This clone has the potential to be released as a new variety.However, this clone needed to be evaluated in a minimal eight environments before releasing this clone because the fresh tuber yield is affected by interaction between genotype and environment [16,17].Farmers consider the fresh tuber yield in selecting the variety.3.3.Tuber long, diameter of tuber, number of tuber/plant and plant height Tuber long was positively and significantly correlated with tuber yield genetically and phenotypically [26].Interaction of genotype and environment was not significantly difference for tuber long (Table 5).The impllication was a good clone in long tuber in Malang will be good in Probolinggo generally.However, an other experiments showed an interaction between genotype and environment for tuber long [19].The genotype factor was significantly difference for tuber long (Table 5).Tuber long of tested clones was significantly different between minimal two tested clones.Tuber long of tested clones ranged from 21 to 28.5 cm with a mean of 24.1 cm (Table 7).Clone CId50-144 had the highest tuber length.
The diameter of tuber was positively and significantly correlated with tuber yield genetically [27].Interaction of genotype and environment was significantly difference for diameter of the tuber (Table 5).Another experiments also showed interaction between genotype and environment for the diameter of the tuber [19].The genotype factor was significantly difference for the diameter of the tuber (Table 5).The tuber diameter of tested clones ranged from to 4.2 cm to 5.2 cmwith a mean of 4.5 cm in Malang, while in Probolinggo ranged from 3.8 cm to 4.8 cm with mean of 4.3 cm (Table 7).Variety UK1 Agritan as a check had the highest diameter of the tuber.This variety can be used as source of gene for high diameter of tuber character.
The number of tuber/plants was positively and significantly correlated with tuber yield genetically [27].The interaction of genotype and environment was not significantly difference for number of tuber/plant (Table 5).The genotype factor was significantly differed for number of tuber/plant (Table 5).Number of tuber/plants of tested clones was significantly different between the minimal two tested clones.The number of tuber of tested clones ranged from 7 to 10 tubers with mean of 9 tubers (Table 8).Variety UJ5 as a check had the highest amount of tuber/plant.This variety can be used as source of gene for high number of tuber/plant character.
Plant height in 4 months was negatively significantly correlated with tuber yield genetically and phenotypically [27].The interaction of genotype and environment was not significantly different for plant height in 4 months (Table 5).Sholihin [23] reported an interaction between genotype and environment for plant height for seven months.The genotype factor was significant difference for plant height in 4 months (Table 5).Plant height in 4 months of tested clones was significantly differed between the minimal two tested clones.Plant height in 4 months of tested clones ranged from 121cm to 176 cm with a mean of 148 cm (Table 8).Clone CId50-144 Variety had the lowest plant height.This variety can be used as source of gene for low plant height character.

Table 1 .
[8].Combined variance analysis for response variable Y(i) on two environments.

Table 3 .
Variance component of quantitative characters of sweet cassava promising clones in two environments.

Table 4 .
Heritability, genotype variance, phenotype variance estimation of quantitative characters of sweet cassava promising clones in Malang and Probolinggo.

Table 5 .
Combined variance analysis of quantitative characters of sweet cassava promising clones in two environments.

Table 6 .
The fresh tuber yield, starch content, starch yield of sweet cassava promising clones.

Table 7 .
Tuber long, diameter of tuber of sweet cassava promising clones.

Table 8 .
Number of tuber/plant and plant height in 4 months of sweet cassava promising clones.