Determination of agronomic properties of tobacco (Nicotiana tabaccum L.) voor-oogst on krosok production using path analysis

Tobacco is an important trade commodity in Indonesia because it supports the country’s economy through taxes and excise. The path analysis is used to measure the direct and indirect effects of agronomic character parameters that are strongly correlated with the results. The path analysis in this study aims to find the agronomic characteristics of tobacco that can be used as selection criteria to get high krosok production. The number of samples used was 50 plants. Observation parameters included plant height, stem diameter, number of leaves per plant, foot leaf length, foot leaf width, middle leaf length, middle leaf width, shoot leaf length, shoot leaf width, leaf wet weight, yield, heavyweight. The results showed the wet weight of the leaves can be used as direct selection criteria. The fresh weight of the leaves has a high positive correlation (Rx10y = 0.812), a high positive direct effect (Px10y = 0.924), and a contribution value of 75.007%.


Introduction
One of the important trading commodities in Indonesia is tobacco. This summer plant is one of Indonesia's many natural resources that has been popular for decades. Indonesian tobacco consumers come from all over the world on average, tobacco is generally used as a cigar wrapper and also as a raw material for clove cigarettes. Currently, the use of tobacco can be used as a raw material for biopesticides, bioinsecticides, and most importantly in the cigarette industry.
The development of superior tobacco varieties can be through plant breeding programs. The assembly of tobacco varieties in plant breeding can be done through a deeper study of the agronomic traits that affect the yield of krosok. Krosok is a dried tobacco leaf used as a cigar. Krosok or dry weight leaf is also the desired result in the tobacco drying process. This agronomic trait is a selection criterion for enhancing or enlarging new varieties [1]. For soybean commodity, to determine the relationship between yield components and agronomic properties can be done through correlation analysis. The agronomic properties that correlated positively with soybean were seed weight per plant, namely flowering age, harvesting age, number of productive branches, number of one seed pods, number of two seed pods, number of three seed pods, number of filled pods, number of pods on branches, total number pods, the total number of seeds and the weight of one hundred grains [2]. The emergence of agronomic traits will be followed by other characteristics so that the emergence of correlations between agronomic traits. Based on [3] the correlation between agronomic properties and results, a closeness relationship can be determined by calculating the value of the correlation coefficient. The strength of the correlation relationship can be further tested through path analysis. Path analysis is a measurement of the direct and indirect effect of agronomic character variables which can have a strong correlation with yield power [1]. Based on the research [4], it was stated that those which had a positive direct effect on the fruit weight of pineapples were plant height, fruit stalk length, fruit diameter, pulp thickness, and total acid. The number of shoots and the harvest age had an indirect effect through fruit diameter and total acid on fruit weight.
Previous research on tobacco breeding resulted [5] in agronomic characters affecting the yield and quality of dry chopped tobacco with total effects of 92.8% and 76.7%, respectively. The five main agronomic factors affecting dry chopped yield are leaf area per blade, number of leaves, specific leaf weight, plant height, and leaf dry weight. Thus, this study aims to determine the direct and indirect effect of an observed agronomic character of tobacco on on the results of krosok of voor-oogst tobacco.

Material
This research was conducted from May to August 2019 in the Jember State Polytechnic. The tools used are pruning shears, bamboo baskets, sacks, scales, sujen bamboo (the bamboo is thick), slop, tarpaulin, tape measure. The material used by the Voor-Oghst tobacco plant is ready for harvest.

Methods
This study used a cross-analysis test of the agronomic traits of voor-oogst tobacco. The tobacco observed was Jepon voor-oogst tobacco. The spacing used is 100 cm x 50 cm. The number of samples used was 50 plants. Observations consist of 11 independent variables and 1 dependent variable. The variables observed were plant height, stem diameter, number of leaves, leg length, foot leaf width, middle leaf length, middle leaf width, shoot length, shoot width, leaf wet weight, yield, the weight of krosok.

Correlation Analysis
Further data analysis through calculating the correlation coefficient to determine the closeness of the relationship between independent and dependent variables. According to [6]

= coefficient of path analysis of inverse correlation between independent variables
with correlation with independent variables between dependent variables = correlation matrix between independent varibles = matrix invers from = Coefficient of correlation between independent variable with dependent variable

Directly contribution and residual
Directly contribution can be calculated using the following formula: 3 = Coefficient of correlation between independent variable with dependent variable = coefficient of path analysis of inverse correlation between independent variables with correlation with independent variables between dependent variables The effect that cannot be explained in a model is called the path coefficient of residual. The path coefficient of residual can be calculated using the following formula: 4

Correlation Between Tobacco Agronomic Characteristics and Krosok Production
There is a correlation between the agronomic properties of tobacco and there is a correlation between the agronomic properties of tobacco and krosok  Table 1 shows the value of the correlation coefficient between the agronomic properties of tobacco plants and the production of tobacco krosok. From the results of the correlation analysis, it can be seen that the largest positive correlation coefficient is the leaf wet weight (RX10Y = 0.812), which means that there is a strong influence between leaf wet weight and krosok production. So that if there is an increase in diameter, leaf number, yield, leaf width, middle leaf length, shoot length, plant height, leg leaf length, leaf wet weight, middle leaf width, leaf width, it will also be followed by increased production from krosok. tobacco. The increase is adjusted for the magnitude of the correlation.

Analysis of Across the Agronomic Traits of Tobacco with Krosok Production
Determining the best character, it is not enough if only by correlation analysis. In correlation analysis, it has a weakness, namely that it is not sufficient to describe the relationship between the components of the results.  Table 2 describes the relationship between the agronomic properties of tobacco plants and result of krosok, directly or indirectly. Through path analysis, the correlation value between independent variables (X) and dependent variables (Y) can be separated into the direct effects of a dependent variable and indirect effects through other dependent variables.
Based on table 2, it is stated that there is a direct effect of the agronomic properties on the weight of the tobacco crop. The direct positive effect is plant height, leaf width, shoot length, wet weight, yield. The negative direct effect was stem diameter, count of leaves, leaf length, middle leaf length, middle leaf width, and shoot width. From the path analysis, the biggest direct effect on krosok production was leaf wet weight, which was 0.924, which means that the effect of the path analysis was categorized as high. The existence of a high direct effect indicates that other agronomic properties are fixed, the increase in leaf wet weight will increase the weight of tobacco krosok.

Directly contribution and residual
Based on table 3 to see the large and small contribution of the agronomic properties of tobacco plants to krosok. The results of the correlation coefficient and path analysis can be calculated for the total and direct contribution value. The direct contribution is the product of the correlation coefficient with the direct effect of the agronomic character on the yield of tobacco krosok. This means that if the value of directly contribution is greater, the agronomic character will have a high effect on the result of crosok tobacco. The total positive contribution to the production of krosok tobacco includes plant  on table 3, the total contribution given by the agronomic character to the production of krosok tobacco was 98.687%. This means that the observed agronomic character affects 98.687% of the production of krosok tobacco. From the total results, there is a residual value of 1.313%. This means that there are external factors of 1.313% that influence tobacco production. Residual or residual value is the residual value that is influenced by external factors. External factors that influence are environment, climate, soil, temperature, cultivation and harvesting systems. These external factors affect the reduction in the total contribution value.

Selection Criteria
Knowledge of the correlation and direct and indirect effects of yield components on the yield of tobacco krosok can be determined by the appropriate yield component characteristics to be used as selection criteria for results [7]. According to [8] a character can be used as a selection criterion, if the correlation value and direct effect are almost as high and have the same sign.
There are 3 things to consider in the selection of the selection criteria, namely the magnitude of the direct influence on the results (Pxy), the magnitude of the correlation between traits and results (Rxy), the difference between the correlation to the results and the direct effect on the results ((Rxy -Pxy) <0.05) . Based on the results obtained, several characters that meet the selection criteria can be selected. For leaf wet weight with a high correlation coefficient value (RX10Y = 0.812) and cross analysis value having a large effect (PX10Y = 0.924) on krosok production, the difference between the two is also small, namely <0.05. So that the agronomic character of the wet weight can be used as a selection criterion for the production of krosok tobacco. Because wet weight is highly correlated and has a high direct effect on tobacco yield. Not only seen from the correlation coefficient but also seen from the direct contribution, the wet weight of the leaves gave the largest contribution, which was around 75.007% to the tobacco krosok. So it is hoped that the more wet weight of tobacco, the weight of krosok will also increase.
The increase in leaf wet weight can be seen when the plants have not been harvested, namely by looking at the value that has a positive correlation between height and wet weight, namely the length