Yield Components of Potato [Solanum tuberosum (L.)] and Their Relationship with Tuber Yield

Potato [Solanum tuberosum (L.)] is the edible tuber and a member of Solanaceae family. It is also called as earth apple, the potato is world’s fourth largest food crop after wheat, rice and maize.  China is the biggest producer of potato worldwide, with about one third of the world’s potato produced in the China and India. According to FAO estimates, in 2019, over 370 million metric tons of potato were produced worldwide, a substantial increase from a production volume of 333.6 million tons in 2010.
 
Root and tuber crops have contributed significantly to staple food requirements in many developing countries, ensuring food security at national and household levels. The major roots and tuber crops used in Nigeria include: cassava (Manihot esulenta), yam (Dioscorea spp), sweet potato (Ipomoea batatas), coco yam (Colocasia esculenta) and Irish potato (Solanum tuberosum). Potato have been part of the regular feeding habit of many Nigerians. The crop is a major contributor to cross-substitution when other food items are in short supply (Ndor, 2013). Tuber yield in potato is a function of positive association of tuber related traits. Understand the interplays of these traits is useful for a meaningful selection process and tuber yield improvement (Sandhya Kiranmai et al., 2016).
 
In a study of some growth indices and their interrelationships with yield, number of leaves and plant height were reported to positively correlated to tuber yield (Kareem, 2014). Several authors have reported findings on relationships among important yield components and yield (Supriatna et al., 2019; Sandhya Kiranmai et al., 2016). However, there were no adequate information on the character associations as well as percentage contributions of the various yield related components to tuber yield of potato in the study area. Therefore, this study was planned to assess the extent of relationships of the various yield related components andto evaluate their direct and indirect contributions to tuber yield of potato.
 

The experiment was conducted during 2017-18 and 2018-19 dry seasons at the Teaching and Research Farm of the Faculty of Agriculture, Bayero University, Kano (11°58'N and 8°25'E) andthe Teaching and Research Farm Kano University of Science and Technology, Wudil (11°25'N and 9°25'E) 400-430 m above sea level. The average temperature of the study areas was 26°C. These locations fall in the Sudan savanna agro-ecological zone of Nigeria.
 
Soil samples were collected from the experimental fields at 0-30 cm depths prior to planting. These were bulked and composite samples used to determine their physical and chemical properties (Table 1). The experiment involved a fractional factorial design of five (5) planting times (late October, early November, mid-November, late November and early December), two methods of propagation (whole seed and cut seed) and three plant densities [66,666 (20 cm), 43,333 (30 cm) and33,333 (40 cm)]. The design was generated using the design of experiment (DOE) platform of JMP 14 according to the D-optimality criterion (Atkinson and Donev, 1989).
 
Plot size was 3 X 4 m long consisting of four ridges. A distance of 1meter was maintained between plots and 1meter distance between blocks. The planting materials were sourced from National Root Crop Research Institute (NRCRI) sub station, Jos. Well sprouted tuber seed of potato var: Marabel was sown at a depth of 10 cm. First weeding was done manually using hand hoe at three weeks after planting andthe subsequent weeding were done when the need arose. NPK 15:15:15 was applied at the rate of 240kg/ha (Ugonna et al., 2013).
 
Data collection and analysis
 
Data were collected on the average tuber weight, number of tubers per plant, tuber size, number of marketable tubers, number of non-marketable tubers and tuber yield. Simple correlation analysis was carried out to determine the relationships between the measured variables and the tuber yield. Simple correlation coefficients between the tuber yield (Y) and yield components (X) and within the yield characters themselves were worked out using the following equation after Poolman (1959).
   
Where,
Y = Correlation coefficient.
SPxy  = Sum of products of x and y.
ssx  = Sum of squares of x.
ssy  = Sum of squares of y. 

The calculated coefficients were further used to develop the following simultaneous equations in order to partition the correlations into cause and effect by working out the path coefficients (P_i).
r₁₆ = p₁ +p₂r₁₂ + p₃r₁₃ + p₄p₁₄ + p₅r₁₅
r₂₆ = p₁r₁₂ +p₂ + p₃r₂₃ + p₄r₂₄ + p₅r₂₅
r₃₆ = p₁r₁₃ +p₂r₂₃ + p₃ + p₄r₃₄ + p₅r₃₅
r₄₆ = p₁r₁₄ +p₂r₂₄ + p₃r₃₄ + p₄ + p₅r₄₅
r₅₆ = p₁r₁₅ +p2r₂₅ + p₃r35 + p₄r₄₅ + p₅
 
From the above equations p₁,  p₂,  p₃,  p₄ and p₅ are the path coefficients (direct effect) while p₁r₁₃, p₁r₂₃, p₁r₃₄, p₁r₄₅, p₂r₂₃, p₂r₂₄, p₂r₂₅, p₃r₂₅, p₃r₃₄, p₃r₃₅ and p₄r₄₅ are the indirect effects while r₁₂ …… r₅₆ are the correlation coefficients. The individual and combined percentage contributions of any two characters were also computed using the following relation as described by Gomez and Gomez (1984). Where,
E = Per cent individual contribution.
E_ij = Combined percent contribution of characters i and j.
R_ij = Coefficient of correlation between i and j.
P_i and p_j = Path coefficients of characters i and j.
 
The residual factor (R_x), which represents the unaccounted error by the direct and combined effects, is calculated as:  R_x = 1 - (p₁r₁₆ + p₂r₂₆ +p₃r₃₆ + p₄r₄₆ + p₅r₅₆), while the sum of the percent contribution (individual and combined) as well as the residual should add up to 100.
 

The results of the soil analysis of the two experimental sites are presented in Table 1. The results showed that the soils at Wudil contains 59.81% sand, 21.32% silt and 18.84% clay. Therefore, soil texture was classified as sandy clay. The soil at BUK contains 64.20% of sand, 19.43% of silt and 16.37 of clay; hence, the soil texture was classified as clayey sand. It was observed that soil at Wudil was slightly acidic (6.41) and neutral (7.35) at BUK. The results also indicated that total nitrogen was high in both BUK and Wudil. The available phosphorus was medium (11.14) at Wudil and low (2.39) at BUK. Other significant differences in micronutrients of the soils of the two sites were observed especially Cu, Mn and Fe which were relatively higher at Wudil.


 
Simple correlation between yield components and tuber yield 
 
Significant correlation was observed between yield components and tuber yield of potato (Table 2). The results from correlation analysis indicated a strong relationship between average tuber weight, number of tubers per stand, tuber size and number of marketable tubers to total tuber yield. Maity and Chatterjee (1997) also reported number of tubers per plant are closely connected with the yield of potato tubers. Similar observation was reported for strong positive correlation of number of roots per plant and root weight to root yield of sweet potato (Yahaya et al., 2015). However, a strong negative correlation exists between number of non-marketable tubers to all other yield component and tuber yield. All other yield components have positive correlation wiyh each other. This indicated that all these characters were important for tuber yield enhancement. Similar association was reported by Majid et al. (2011) and Lemma Tessema et al. (2020). However, a strong negative correlation exists between number of non-marketable tubers and tuber weight.


 
Direct, indirect and total contributions of some yield components to tuber yield
 
The direct, indirect and total contributions of yield components to tuber yield of potato is presented in Table 3. The total contribution of average tuber weight to tuber yield was significant (0.8910) while the direct contribution was (0.4082). This corroborates with the results of Yahaya et al. (2015) who reported root weight as the highest direct contributor to root yield in sweet potato. The indirect contribution of average tuber weight via number of tubers, tuber size, number of marketable tubers and number of non-marketable tubers were observed to be -0.0354, 0.0697, 0.4852 and-0.0368 respectively. Islam et al. (2002) reported that average tuber weight and number of tubers had positive and high direct effects on tuber weight. For this reason, these traits could be used more significantly for potato improvement. 


 
The result of the study further revealed that total contribution of number of tubers to tuber yield was 0.4890. Hossain et al. (2000) reported similar result. When these were portioned into direct and indirect contribution, it was observed that -0.0753 was directly contributed through number of tubers. However, only 0.1919, 0.0384, 0.3542 and -0.0202 were contributed indirectly through average tuber weight, tuber size, number of marketable tubers and number of non-marketable tubers respectively. These findings were in accordance with the results of Galarreta et al. (2006).
 
The total contribution of tuber size to tuber yield of potato was observed to be 0.8600. Out of this, only 0.0741 was directly contributed by tuber size. Similarly, 0.3841, -0.0391, 0.4779 and-0.0331 were indirectly contributed by tuber size through average tuber weight, number of tubers, number of marketable tubers and number of non-marketable tubers respectively.
 
The result of the study further indicated that 0.9210 was the total contribution on number of marketable tubers to tuber yield. Out of which 0.6065 was directly contributed by number of marketable tubers. However, 0.3266, -0.0439, 0.0584 and -0.0265 were indirectly contributed by number of marketable tubers through average tuber weight, number of tubers, tuber size and number of non-marketable tubers respectively. The path coefficient analysis revealed that the direct effect on tuber yield was positive on number of marketable tubers, whereas all other characters evaluated under study exhibited direct effects (Sahu et al. 2017).
 
The total contribution of number of non-marketable tubers to tuber yield was -0.6730. Out of this 0.0449 was directly contributed by number of non-marketable tubers. Similarly, -0.0335, 0.0338, 0.0610 and -0.0 3573 were indirectly contributed by number of non-marketable tubers through average tuber weight, number of tubers, tuber size and number of marketable tubers.
 
Direct and combined contributions (%) of yield components to tuber yield
 
When the individual percentage contributions of yield components were examined, it was observed that the percentage (direct) contribution of average tuber weight was 16.6646% (Table 4). Similarly, the percentage (direct) contribution of number of tubers, tuber size, number of marketable tubers and number of non-marketable tubers to tuber yield were 0.5675%, 0.5485%, 36.7897% and 0.2023% respectively. The positive direct effect on number of tubers on tuber yield was in agreement with the findings of Alam et al. (1998) and Parida et al. (1999).


 
The combined contributions of average tuber weight and number of tubers was negative (-1.4453%). Similar trend was observed for the combined effects of average tuber weight and number of non-marketable tubers, number of tubers and tuber size, number of tubers and number of marketable tubers, tuber size and number of non-marketable tubers and number of marketable tubers and number of non-marketable tubers in which 1.5001%, 0.5250%, 3.6553%, 0.2722% and 2.2288% were contributed, respectively. Lavanya et al. (2020) reported that numbers of tubers, marketable yield, number of stems and tuber weight were the most influencing factors to improve the tuber yield. Yahaya and Ankrumah (2017) also reported that the greatest combined contributions of yield characters to grain yield in soybean were observed from number of pods per plant and number seeds per pod.
 
The result of the study further indicated that the combined contributions of average tuber weight and tuber size to tuber yield was 2.8450. However, 19.8084%, 0.2768% and 3.5938% were contributed by the combined effects of average tuber weight and tuber size, number of tubers and number of non-marketable tubers as well as tuber size and number of marketable tubers were contributed, respectively. Out of all these contributions, 28.3301% could not be accounted for andtherefore regarded as residual. Burhan (2007) reported that tuber yield was identified by tuber weight and average tuber weight since these characters had a positive and significant direct effect on tuber yield.
 

Significant and positive correlations were observed between average tuber weight, number of tubers, tuber size and number of marketable tubers to tuber yield of potato. Upon partitioning the correlation coefficients into direct, indirect and combined effects, the average tuber weight has the highest direct contribution to tuber yield. Average tuber weight and number of marketable tubers gave the highest indirect as well as combined contributions to tuber yield of potato. These traits were most influencing factors for improvement of tuber yield.
 
 

The authors hereby acknowledge the support of the Center for Dry land Agriculture (CDA), Bayero University, Kano for providing the financial assistance to conduct this research.