Interrelationship and Path analysis of Tuber yield and related traits in yam (Dioscorea spp.) from Ethiopia

M. Tewodros¹*, M. Firew², H. Shimelis³, G. Endale⁴

¹Jimma Agricultural Research Center, P.O. Box 192, Jimma, Ethiopia,

²Haramaya University, School of Plant Sciences, P.O. Box 138, Dire Dawa, Ethiopia

³African Centre for Crop Improvement, School of Agriculture, Earth and Environmental Sciences,

University of Kwa Zulu-Natal, Private Bag X01, Scottsville, 3209, Pietermaritzburg, South Africa.

⁴Ethiopian Institute of Agricultural Research, P.O. Box 2003, Addis Ababa, Ethiopia

*Corresponding Author E-mail: tewodrosmulualem@gmail.com

ABSTRACT:

Understanding the nature of associations among economically important traits is essential to improve selection efficiency in plant breeding programs. This study aimed to determine the magnitude of association between tuber yield and related traits and to identify the most influential character(s) involving 36 landrace collections of yams for effective selection and conservation. Field evaluations were conducted at Jimma Agricultural Research Center in Ethiopia using a 6x6 lattice design with two replications during 2015. Data on 12 qualitative and 19 quantitative traits were collected and subjected to analysis of variance, correlation and path analyses. Highly significant differences (p<0.01) were detected among collections for the studied traits. Signiﬁcant and positive correlations were detected between tuber fresh weight (TFW) with vine length (VL), days to maturity (DM) and tuber diameter (TDi). Tuber length (TL) positively and significantly correlated with leaf length (LL), vine length (VL) and internodes length (IL). Tuber diameter (TDi) positively correlated with LL, TuL, IL, DM and leaf width (LW). Leaf color (LC), leaf size (LSi), petiole color (PC), vine color (VC), tuber skin color (TSC) and tuber flesh color (TFC) had signiﬁcantly negative correlations with TFW and TuL. Path analysis revealed high direct path coefficient value (1.112) between LL and TuL. Also, positive direct path coefficient value (1.018) was exhibited between DM and TFW. Relatively high direct path coefficient value (0.356) was exhibited between leaf shape (LS) and TuL. This study revealed that selection for increased above ground biomass and days to maturity may improve genetic gain in storage tuber yield and length of tuber in yam breeding. Using the overall analyses, the following collections such as: 27/02, 56/76, 08/02, 10/002,39/87, 45/03,6/02,116, and 7/83 were selected for breeding and conservation.

KEYWORDS: Correlation analysis, path coefficients, yam, tuber yield.

1. INTRODUCTION:

Yam (Dioscorea species) belonging to the family Dioscoreaceae is cultivated in Africa, Asia, South America, Caribbean and the South Pacific islands for its storage tubers (Asieduand Alieu, 2010). The genus comprises over 600 species (Sesay et al., 2013), Only 10 are cultivated for human food and medicinal values such as Dioscore aalata, D. esculenta, D. batatas or D. opposite, D. bulbifera, D. cayenensis, D. rotundata complex, D. dumetorum, D. trifida, D. nummularia and D. pentaphylla (Girma et al. 2012; Dansi et al., 2013). Only few species like D. alata, D. bulbifera, D. cayenensis and, D. rotundata complex are the most widely cultivated species in Africa with greater economic significance (Lebot, 2009; Asiedu and Alieu, 2010; Norman et al., 2012). Ethiopia is considered to be the center of origin and diversity of most yam species (Coursey, 1967; Zeven and De Wet, 1982; Tamiru et al., 2011). In the country a large number of yam landraces are cultivated but no systematic genetic conservation has yet been done (Hildebrand et al., 2003). In south west Ethiopia, yams are predominantly grown by smallholder farmers for food, traditional medicine, and to earn cash incomes. Women farmers are the main producers and traders of yams in Ethiopia. Yam is increasingly showing high market value owing to its consumer demand for food and medicine purposes making it an ideal candidate for market-driven production (Mulualem and Weldemichel, 2013).

Yams have low tuber yield levels in Ethiopia estimated at 4 t/ha (Tamiru, 2011) when compared to West African countries where yields of 25 t/ha are reported (Asiedu and Alieu, 2010). Therefore, systematic genetic improvement of the crop is required in Ethiopia to boot productivity. A well-characterized genetic pool is required to develop high yielding yam varieties incorporating farmers’ and market demand (Dansi et al., 2013). Therefore, assessment of genetic variability and association between agro-morphological traits in the existing genetic resources remain a key component of yam breeding and conservation programs (Norman et al., 2011).

Tuber yield and yield components are important selection criteria of promising yam genetic resources. Tuber yield is directly or indirectly affected by yield components requiring selection of relevant traits positively correlated with it (Mulualem et al., 2013). Therefore, knowledge on the interrelationship and degree of association of yield and yield components is useful to improve selection efficiency in yam breeding and conservation efforts. Simple correlation analysis describes the mutual associations of variables without regard to cause and effect inter-relationship.

Component characters during selection are often inter-dependent showing direct or indirect interrelationships and influencing selection efficiency for yield and yield components. Simple correlation coefficients can be misleading if a high correlation between two traits is a consequence of the indirect effect of other traits (Dewey and Lu, 1959). Path coefficient analysis is a standardized partial regression which was developed by Wright (1921) and later described by several authors (Wright, 1934; Li, 1956; Bhatt, 1973; Mohammadi et al., 2003). Unlike simple correlation analysis, path coefficient analysis separates the total correlation coefficients into the direct and indirect effects in order to make selection more effective (Dewey and Lu, 1959; Falconer and Mackey, 1996). Path coefficients give the relative contribution of various yield determining traits, enabling breeders to decide between direct and indirect selection procedures (Paul et al., 2013). The direct and indirect effects of traits on economic yield can be determined through path analysis which has been used in a number of crops to study the relationships between yield and yield components (Christopher, 2000). There is no information regarding the application of path analysis in the selection and conservation of yam genotypes. This information could provide valuable insight when evaluating diverse genotypes of yam for targeted breeding. Further, isolating the most influential yam traits would enhance the selection responses in yam improvement programs. Therefore, the objectives of this study were to determine the magnitude of association between tuber yield and related traits and to identify the most influential character (s) involving 36 landrace collections of yams for effective selection and conservation.

2. MATERIALS AND METHODS:

2.1. Study area:

The experiment was conducted at Jimma Agricultural Research Center (JARC). The center is located at latitude 7^o 40.00' N and longitude 36^o 47’.00’ E with an altitude of 1753 meters above sea level (m.a.s.l.). The area receives mean annual rainfall of 1432mm with mean maximum and minimum temperatures of 29.2⁰C and 8.90⁰C, respectively. The soil at the study site is Eutric Nitosol (reddish brown) with pH of 5.3. These environmental conditions are conducive for yam cultivation.

2.2. Plant materials, experimental design and field management:

A total of 36 yam landraces was collected from Jimma, Sheka and Bench-maji zones of southwest Ethiopia (Table 1). The experiment was laid out in a 6 x 6 simple lattice design with two replications. Plants were field established using a 7 m long rows using inter-row spacing of 1.5 m and intra-rows spacing of 1 m. Tubers of the same size which started sprouting were used as planting material. One month after planting, seedlings were earthed up followed by frequent weeding. All other agronomical practices were followed according to the recommendations and farmers practices of the areas. Each yam plant was tended using dried coffee sticks of 3.5-4.5 m long to provide support and induce good canopy and vine development. Five middle plants within a row were sampled and tagged for data collection and final harvest.

2.3. Data collection:

Both qualitative and quantitative data were collected according to the descriptors of yam (Dioscorea spp.) developed by Bioversity International (IPGRI, 1997). The qualitative data collected with their descriptors are summarized in Table 2. These traits are the most important farmers preferred attributes of yam landraces for various purposes. For example, leaf color, leaf shape, leaf size, leaf density, petiole color, twinge direction, vine color, tuber shape, tuber skin color, tuber branching, tuber surface texture and tuber flesh color. Data on 19 quantitative traits were collected as follows.

2.3.1. Data collected on five selected plants:

· Leaf length (cm): the mean length of fully expanded young leaves was measured on the main vine from collar to the tip of the leaves at maturity.

· Leaf width (cm): the mean width of fully expanded young leaves was measured on the main vine from widest part at maturity.

· Length of leaf lobe (cm): the mean length of five fully expanded leaf lobs from the main vine of leaves was measured from the middle at maturity.

· Petiole length (cm): the mean length of petioles on the main vine was measured from the base to the point of insertion of the leaf lob at maturity.

· Distance between lobs (cm): the mean distance of fully expanded leaf lobs from the main vine of leaves was measured from the base at maturity.

· Vine length (cm): the length of the longest vine was measured with a tape from the ground level to the tip at the time of maturity.

· Number of tubers plant^-1: counted from five plants and the mean calculated

· Tuber length (cm): measured from the lower to the upper tip using venire caliper at harvest.

· Tuber diameter (cm): measured at the middle using venire caliper at harvest.

· Internodes length (cm): mean length was measured from the middle of vine at maturity.

· Number of internodes vine^-1: counted from five plants and the mean calculated

· Number of vines hill^-1: counted at maturity.

· Tip length (cm): the mean length of tip was measured from the apex at maturity.

2.3.2. Data collected on plot basis:

· Vine fresh weight: the vine of all the plants from each plot were bulked and weighed at harvest and the vine fresh weight/plot was then converted to tonnes per hectare (t/ha).

· Vine dry weight: the total vine dry weight was estimated by drying 100gm of fresh vine in a forced air circulation oven at 700C for about 72 hours and expressed in tonnes per hectare (t/ha) of the vine fresh weight.

· Tuber fresh weight: the tuber of all the plants of each plot were bulked and weighed at harvest and the tuber fresh weight/plot was then converted to tonnes per hectare (t/ha).

· Tuber dry weight: the total vine dry weight was estimated by drying 100gram of fresh vine in a forced air circulation oven at 700C for about 72 hours and expressed in tonnes per hectare (t/ha) of the tuber fresh weight.

· Days to maturity: was recorded from date of emergence to the date when the crop was ready for harvesting, i.e. tubers had become mature and the plant had started drying.

· Harvest Index (%): was calculated as a ratio of total marketable tuber yield to the sum of biomass and total marketable tuber yield and expressed in percentage.

2.4. Data analysis:

Qualitative data were analyzed using Statistical Package for Social Sciences 16.0 (SPSS, 1996). Quantitative data were subjected to analysis of variance (ANOVA) using the lattice procedure as suggested by Gomez and Gomez (1984) using SAS version 9.0 (SAS, 2000) and Genres (2008) statistical software packages. Means were separated using the Least Significant Difference (LSD) procedure at the 5% and 1% level of signiﬁcance. The Spearman’s rank non-parametric correlation coefficients were calculated to describe the degree and pattern of associations of observed qualitative and quantitative traits of yam landraces.

Path coefficient analysis was conducted using a genotypic correlation matrix set up as G= K×L both for tuber fresh weight and tuber length. In this matrix vector ‘G’ represents the genotypic correlation coefficients of storage tuber fresh weight (a) or tuber length vs. qualitative and quantitative traits. In the same vector ‘K’ is the value of genotypic correlation for all possible combinations among the traits and vector ‘L’, the path coefficients. The path coefficients were calculated as the product of vector G and each row of K^-1using the Genres (2008) statistical software. Direct and indirect path coefficients were calculated for both qualitative and quantitative traits as proposed by Wright (1934) and later described by Williams et al. (1990) using genotypic correlation coefficients. For path analysis of quantitative traits, tuber fresh weight (TFW) and tuber length (TuL) were considered as response variables whereas, leaf length, leaf width, petiole length, vine length, internodes length, number of internodes per vine, days to maturity, number of tubers per hill, TuL, tuber diameter, TFW, tuber dry weight and harvest index were considered as causal variables. For path analysis of qualitative traits, TFW and TuL were considered as response variables. Furthermore, leaf color, leaf shape, leaf size, leaf density, petiole color, twinge direction, vine color, tuber shape, tuber skin color, tuber branching, tuber surface texture and tuber flesh color were regarded as casual variables.

Table 1: Descriptions the 36 yam accessions used for the study.

S. No.	Name of landraces	Zone	District	Latitude	Longitude	Altitude
1	59/02	Jimma	Mana	07⁰40’37N	036⁰49’10E	1718
2	68/01	Jimma	Dedo	07⁰30’63N	036⁰53’45E	1784
3	6/02	Bench maji	Sheko	06⁰59’66N	035⁰34’11E	1728
4	75/02	Jimma	Kersa	07⁰40’43N	036⁰48’76E	1734
5	3/87	Jimma	Manna	07⁰40’58N	036⁰48’75E	1731
6	56/76	Jimma	Manna	07⁰41’89N	036⁰48’06E	1837
7	54/02	Bench maji	Sheko	07⁰02’03N	035⁰32’77E	1892
8	46/83	Jimma	Dedo	07⁰31’28N	036⁰53’59E	1771
9	08/02	Jimma	Kersa	07⁰40’46N	036⁰48’79E	1740
10	116	Jimma	Dedo	07⁰31’28N	036⁰53’63E	1683
11	01/75	Sheka	Yeki	07⁰11’30N	035⁰26’22E	1171
12	06/83	Jimma	Dedo	07⁰31’32N	036⁰53’64E	1692
13	17/02	Sheka	Yeki	07⁰11’27N	035⁰26’26E	1176
14	07/03	Jimma	Dedo	07⁰31’50N	036⁰53’60E	1733
15	45/03	Jimma	Mana	07⁰41’86N	036⁰48’08E	1810
16	27/02	Jimma	Sekachekorsa	07⁰35’06N	036⁰41’91E	1877
17	37/87	Jimma	Mana	07⁰41’87N	036⁰48’13E	1940
18	10/002	Bench maji	Sheko	07⁰02’91N	035⁰29’76E	1668
19	76/02	Jimma	Kersa	07⁰40’64N	036⁰48’84E	1728
20	06/2000	Jimma	Sekachekorsa	07⁰35’43N	036⁰41’86E	1850
21	7/83	Jimma	Sekachekorsa	07⁰35’06N	036⁰41’91E	1898
22	58/02	Sheka	Yeki	07⁰11’22N	035⁰26’25E	1192
23	39/87	Jimma	Sekachekorsa	07⁰35’42N	036⁰42’94E	1885
24	32/83	Jimma	Shebesombo	07⁰26’74N	036⁰24’01E	1372
25	24/02	Jimma	Shebesombo	07⁰26’75N	036⁰24’07E	1379
26	2/87	Jimma	Shebesombo	07⁰26’76N	036⁰24’12E	1365
27	60/87	Sheka	Yeki	07⁰11’72N	035⁰26’48E	1199
28	15/2000	Bench maji	Sheko	07⁰04’13N	035⁰37’74E	1320
29	34/87	Jimma	Dedo	07⁰31’37N	036⁰53’44E	1911
30	21/02	Jimma	Sekachekorsa	07⁰36’48N	036⁰45’09E	1785
31	57/76	Bench maji	Sheko	07⁰02’88N	035⁰29’74E	1654
32	0001/07	Jimma	Shebesombo	07⁰26’74N	036⁰24’12E	1367
33	0004/07	Jimma	Kersa	07⁰40’55N	036⁰48’75E	1741
34	7/84	Bench maji	Sheko	07⁰02’88N	035⁰29’74E	1661
35	7/85	Sheka	Yeki	07⁰14’30N	035⁰26’17E	1173
36	06/2001	Bench maji	Sheko	06⁰59’69N	035⁰34’09E	1387

Table 2: List of qualitative traits with detailed descriptions used in the study.

Qualitative traits	Description	Acronym
Leaf color	1= yellow green, 2= pale green, 3=dark green, 4= purplish green, 5=purple, 99=other	LC
Leaf shape	1= ovate, 2= chordate 3= chordate long, 4=chordate broad, 5= sagittate	LS
Leaf size	1= small, 2= medium, 3= large	LSi
Leaf density	1= low, 2= medium, 3= high	LD
Petiole color	1= all green with purple base, 2= all green with purple leaf junction, 3= all green with purple at both ends, 4= all purplish green with purple base, 5= all purplish green with purple leaf junction, 6= all purplish green with purple at both ends, 7= green, 8= purple, 9= brownish green, 10= brown, 11= dark brown, 99=other.	PC
Twinge direction	1= clockwise, 2= anticlockwise	TD
Vine color	1= yellowish, 2= green, 3= light green, 4=purple, 99=other	VC
Tuber shape	1= round, 2= oval, 3= oval-oblong, 4= cylindrical, 5= flattened, 6= irregular,	TS
Tuber skin color	1= grayish, 2= light brown, 3=dark brown, 4= 0ther	TSC
Tuber branching	0= none, 1= slightly branched, 2= branched, 3= highly branched	TBr
Tuber surface texture	1= smooth, 2= rough	TSte
Tuber flesh color	1= white, 2= yellowish white, 3= yellow, 4= orange, 5= light purple, 6= purple 7= purple with white, 8= white with purple, 9= outer purple/ inner white, 99= other	TFC

Table 3: Mean response of 36 yam landraces evaluated for qualitative traits.

No		Qualitative parameters
No	Name of landraces	LC	LS	Lsi	LD	PC	TDI	VC	TS	TSC	TBr	TSt	TFC
1	59/02	1.0	1.0	3.0	2.0	1.0	1.0	3.0	6.0	2.0	1.0	2.0	7.0
2	68/01	3.0	5.0	1.0	3.0	5.0	1.0	2.0	4.0	6.0	2.0	1.0	8.0
3	6/02	3.0	2.0	1.0	3.0	4.0	1.0	2.0	6.0	2.0	3.0	2.0	1.0
4	75/02	3.0	2.0	1.0	3.0	2.0	1.0	2.0	4.0	2.0	3.0	2.0	1.0
5	3/87	3.0	2.0	2.0	3.0	7.0	1.0	2.0	4.0	2.0	3.0	2.0	6.0
6	56/76	2.0	5.0	1.0	3.0	1.0	2.0	3.0	2.0	3.0	3.0	2.0	1.0
7	54/02	3.0	2.0	1.0	3.0	5.0	1.0	3.0	3.0	2.0	2.0	1.0	6.0
8	46/83	3.0	2.0	3.0	2.0	4.0	1.0	2.0	3.0	2.0	0.0	2.0	9.0
9	08/02	1.0	2.0	1.0	3.0	4.0	1.0	2.0	6.0	3.0	2.0	2.0	1.0
10	116	3.0	2.0	1.0	3.0	7.0	1.0	2.0	5.0	3.0	2.0	2.0	9.0
11	01/75	1.0	2.0	1.0	3.0	2.0	1.0	2.0	6.0	3.0	2.0	1.0	6.0
12	06/83	3.0	2.0	3.0	2.0	4.0	1.0	2.0	3.0	2.0	3.0	2.0	1.0
13	17/02	1.0	2.0	2.0	3.0	2.0	1.0	2.0	4.0	2.0	0.0	2.0	7.0
14	07/03	3.0	2.0	2.0	3.0	4.0	1.0	2.0	4.0	2.0	3.0	2.0	8.0
15	45/03	2.0	5.0	2.0	3.0	1.0	1.0	3.0	6.0	2.0	2.0	2.0	8.0
16	27/02	3.0	2.0	2.0	3.0	4.0	1.0	3.0	4.0	3.0	2.0	2.0	1.0
17	37/87	3.0	5.0	2.0	2.0	4.0	1.0	3.0	6.0	3.0	3.0	2.0	9.0
18	10/002	3.0	1.0	2.0	3.0	7.0	1.0	3.0	4.0	2.0	2.0	2.0	7.0
19	76/02	3.0	2.0	2.0	3.0	6.0	1.0	3.0	6.0	2.0	2.0	2.0	1.0
20	06/2000	1.0	2.0	2.0	3.0	6.0	1.0	3.0	4.0	3.0	2.0	1.0	8.0
21	7/83	3.0	5.0	2.0	3.0	6.0	1.0	2.0	6.0	2.0	0.0	2.0	1.0
22	58/02	3.0	2.0	2.0	2.0	6.0	1.0	3.0	4.0	3.0	2.0	2.0	8.0
23	39/87	2.0	2.0	3.0	3.0	4.0	1.0	3.0	3.0	3.0	2.0	2.0	8.0
24	32/83	2.0	1.0	2.0	3.0	4.0	1.0	3.0	6.0	3.0	2.0	1.0	6.0
25	24/02	2.0	2.0	3.0	2.0	6.0	1.0	3.0	4.0	3.0	3.0	2.0	6.0
26	2/87	3.0	2.0	2.0	3.0	6.0	1.0	3.0	6.0	2.0	2.0	2.0	1.0
27	60/87	2.0	1.0	3.0	3.0	4.0	1.0	3.0	5.0	3.0	3.0	2.0	6.0
28	15/2000	3.0	2.0	2.0	3.0	5.0	1.0	2.0	2.0	2.0	2.0	1.0	8.0
29	34/87	3.0	2.0	2.0	3.0	6.0	1.0	3.0	4.0	2.0	2.0	2.0	1.0
30	21/02	3.0	2.0	1.0	3.0	4.0	1.0	3.0	4.0	3.0	0.0	2.0	7.0
31	57/76	3.0	2.0	2.0	3.0	5.0	1.0	2.0	6.0	2.0	2.0	1.0	6.0
32	0001/07	2.0	5.0	3.0	2.0	1.0	2.0	3.0	2.0	2.0	3.0	2.0	9.0
33	0004/07	2.0	5.0	2.0	3.0	1.0	1.0	3.0	6.0	3.0	3.0	2.0	8.0
34	7/84	3.0	2.0	2.0	3.0	4.0	1.0	2.0	6.0	3.0	1.0	2.0	6.0
35	7/85	3.0	1.0	2.0	2.0	6.0	1.0	3.0	4.0	3.0	3.0	1.0	8.0
36	06/2001	2.0	2.0	2.0	3.0	6.0	1.0	3.0	5.0	3.0	0.0	2.0	7.0

LC= leaf color, LS= leaf shape, LSi= leaf size, LD=leaf density, PC=petiole color, TDI=twinge direction, VC=vine color, TS=tuber shape), TSC=tuber skin color, TBr=tuber branching, TSt=tuber surface texture and TFC=tuber flesh color.

Table 4: Partial analysis of variance showing mean squares and significant tests and summary statistics for 19 quantitative characters of 36 yam collections evaluated at Jimma, Ethiopia during 2015

No.	Quantitative character	Mean square
No.	Quantitative character	Accession (DF=35)	Error	R²	CV (%)	P-value
1	Leaf length (cm)	2.92**	0.38	0.93	5.89	<0.0001
2	Leaf width (cm)	4.92**	0.08	0.89	6.99	<0.0001
3	Petiole length (cm)	8.52**	3.05	0.82	17.04	0.0046
4	Length of leaf lobe (cm)	0.46	0.59	0.68	38.70	0.7478
5	Distance between lobs (cm)	0.37	0.56	0.55	22.04	0.8781
6	Vine length (cm)	188.1**	34.28	0.95	2.29	<0.0001
7	Internodes length (cm)	0.30**	0.11	0.88	3.54	0.0080
8	Tip length (cm)	0.23	0.45	0.48	26.64	0.9701
9	Number of internodes vine^-1	10.96**	2.56	0.88	6.17	0.0002
10	Number of vines hill^-1	0.59**	0.26	0.78	11.99	0.018
11	Days to maturity	187.32**	70.78	0.83	6.07	0.0067
12	Number of tubers hill^-1	0.10	0.12	0.63	8.16	0.7411
13	Tuber length (cm)	4.88**	1.89	0.88	3.54	0.0079
14	Tuber diameter (cm)	5.61**	1.94	0.84	9.34	0.0036
15	Vine fresh weight (t/ha)	12.31	20.45	0.52	36.58	0.9182
16	Vine dry weight (t/ha)	0.85	1.13	0.61	34.22	0.7871
17	Tuber fresh weight (t/ha)	256.07**	10.80	0.97	10.99	<0.0001
18	Tuber dry weight (t/ha)	4.58**	0.88	0.90	4.53	<0.0001
19	Harvest Index (%)	187.32**	70.78	0.83	12.29	0.0067

*and **Signiﬁcant difference at the 0.05and 0.01 probability level, DF: Degree of freedom.

Table 5: Mean response of 36 yam collections evaluated for 13 quantitative traits at Jimma in 2015.

Acc	LL	LW	PL	VL	IL	NIPV	DM	NT PH	TuL	TuDi	TFW	TDM	HI%
1.0	12.68	5.13	7.13	262.33	9.81	26.83	147.75	4.08	39.22	15.84	40.40	23.36	77.75
2.0	11.91	4.67	14.00	265.42	10.46	24.55	142.29	4.33	41.83	15.76	30.00	21.48	72.29
3.0	14.16	5.00	12.88	267.00	10.34	26.30	140.04	4.08	41.35	16.74	35.10	21.62	70.04
4.0	11.07	4.54	8.77	280.85	10.29	29.00	145.53	4.17	41.17	14.88	33.80	21.36	75.53
5.0	10.07	4.16	7.98	251.85	9.51	27.90	125.88	4.33	38.03	13.12	16.80	23.57	55.88
6.0	11.82	4.44	11.53	267.25	10.26	25.40	157.57	4.50	41.04	16.40	53.80	20.96	87.57
7.0	8.65	3.72	9.58	267.00	9.44	30.00	143.44	4.50	37.77	12.40	30.40	22.45	73.44
8.0	9.83	3.84	13.28	257.25	9.34	26.80	142.68	4.17	37.37	13.92	28.55	20.81	72.68
9.00	11.81	4.66	10.17	256.50	10.45	23.10	144.84	4.33	41.81	18.00	41.40	20.74	74.84
10.0	9.05	4.66	11.15	250.10	10.08	21.40	139.72	4.33	40.33	16.08	29.60	19.81	69.72
11.0	10.09	3.50	10.50	253.75	9.66	22.70	146.11	4.33	38.66	13.38	33.80	21.97	76.11
12.0	10.20	4.08	9.40	252.25	9.66	25.60	145.25	4.50	38.66	13.76	44.00	21.80	75.25
13.0	10.16	4.31	9.77	240.50	9.12	27.10	149.39	4.17	36.49	17.20	49.20	18.14	79.39
14.0	9.29	3.00	9.80	263.50	9.95	28.60	136.59	4.08	39.79	15.04	24.00	22.51	66.59
15.0	11.28	4.86	11.73	260.75	10.03	26.80	142.79	4.67	40.13	17.76	37.40	18.47	72.79
16.0	10.93	4.49	12.38	271.50	10.96	25.10	142.31	4.75	43.83	15.92	47.40	18.86	72.31
17.0	8.75	3.58	5.82	296.75	10.10	33.00	141.29	3.92	40.42	12.48	30.60	22.50	71.29
18.0	10.28	4.43	9.60	263.85	9.68	28.30	152.13	4.58	38.72	15.60	63.00	17.83	82.13
19.0	10.17	4.06	8.23	247.00	9.28	26.03	144.25	3.92	37.13	12.16	33.40	23.33	74.25
20.0	9.67	4.08	11.83	240.38	9.52	23.78	142.93	3.92	38.08	13.92	26.80	19.70	72.93
21.0	10.78	4.72	11.57	254.25	9.53	26.93	151.14	4.17	38.13	16.48	45.60	19.10	81.14
22.0	9.71	4.05	10.23	241.75	8.89	24.35	143.71	4.42	35.56	19.52	29.20	16.93	73.71
23.0	12.34	4.73	13.50	258.50	9.97	23.33	113.78	4.58	39.89	16.08	11.40	20.76	43.78
24.0	9.62	4.05	8.10	242.25	9.02	27.00	125.77	4.42	36.08	13.68	16.20	23.17	55.77
25.0	8.28	3.48	7.48	259.25	9.40	31.80	130.80	4.50	37.61	12.72	19.40	21.22	60.80
26.0	11.20	4.42	12.92	239.00	9.25	22.10	126.75	4.33	37.01	14.48	14.80	20.53	56.75
27.0	10.65	4.58	11.12	254.25	9.93	25.43	145.06	4.58	39.71	14.64	27.80	21.30	75.06
28.0	11.86	4.22	8.17	243.25	9.51	24.50	114.93	4.75	38.06	12.10	11.80	20.86	44.93
29.0	10.72	4.16	8.17	235.00	9.01	22.50	98.98	4.83	36.06	12.64	6.80	20.55	28.98
30.0	9.76	4.30	9.92	237.50	8.92	26.40	139.13	4.50	35.69	16.72	28.40	20.61	69.13
31.0	9.88	4.42	10.45	246.35	9.22	26.33	141.72	4.25	36.90	12.72	32.00	19.17	71.72
32.0	10.20	4.23	11.12	239.75	9.17	25.20	142.30	4.75	36.69	9.11	23.40	22.06	72.30
33.0	10.04	4.33	10.78	243.05	9.20	25.76	142.01	4.50	36.79	10.92	27.70	20.62	72.01
34.0	8.77	3.61	8.92	243.75	9.30	26.38	133.89	4.42	37.19	13.78	23.67	21.71	63.89
35.0	12.39	4.63	12.60	252.25	9.75	26.00	115.42	4.17	39.02	12.39	9.60	19.91	45.42
36.0	8.53	3.50	6.15	241.25	9.47	24.90	130.39	4.17	37.90	10.61	14.40	19.84	60.39
Mean	10.46	4.24	10.19	254.09	9.65	26.03	138.02	4.36	38.61	14.88	29.77	20.82	68.02
St.div	1.31	0.47	2.07	13.33	0.49	2.50	12.22	0.24	1.98	2.27	12.93	1.61	12.33
Min	8.28	3.00	5.82	235.00	8.89	21.40	98.98	3.92	35.56	11.47	6.80	16.93	28.98
Max	14.16	5.13	14.00	296.75	10.96	33.00	157.57	4.83	43.83	19.52	63.00	23.57	87.57

LL=Leaf length (cm); LW= Leaf width (cm); PL= Petiole length (cm); VL= Vine length (cm); NTPP= Number of tubers hill^-1;

TL=Tuber length (cm); TuDi= Tuber diameter (cm); IL= Internodes length (cm); NIPV= Number of internodes vine^-1; DM= Days to maturity; TFW=Tuber fresh weight (t/ha); TDW=Tuber dry weight (t/ha); and HI= Harvest Index (%).

St.div-=Standard deviation; Min=Minimum and Max= Maximum

Table 6: Spearman’s rank correlation coefficients showing pair-wise association of qualitative traits assessed among 36 yams landraces

Traits	LC	LS	Lsi	LD	PC	TDI	VC	TS	TSC	TBr	TSt	TFC
LC	1.00	0.013	-0.118	-0.020	0.485**	-0.157	-0.148	-0.146	-0.096	0.118	0.071	-0.139
LS		1.00	-0.132	-0.022	-0.389**	0.476**	0.028	-0.023	0.243	0.109	0..131	0.062
Lsi			1.00	-0.547**	-0.032	0.020	0.268	-0.158	-0.266	0.001	0.257	0.240
LD				1.00	0.080	-0.162**	-0.180	0.216	0.058	-0.137	-0.125	-0.276
PC					1.00	-0.433	-0.025	0.008	0.063	-0.077	-0.137	-0.034
TDI						1.00	0.204	-0.471**	-0.026	0.249	0.129	-0047
VC							1.00	-0.046	0.055	0.115	0.090	0.108
TS								1.00	0.025	-0.131	0.062	-0.157
TSC									1.00	0.037	-0.293	0.242
TBr										1.00	-0.068	-0.105
TSt											1.00	-0.253
TFC												1.00

**Signiﬁcant difference at 0.01 probability level.

Table 7: Spearman’s rank correlation coefficients showing pair-wise association of quantitative traits measured among yam landraces.

Traits

NIPV

NTPH

TuL

TUDi

TFW

TDW

1.00

0.77**

0.47**

0.15

0.45**

-0.33*

-0.08

0.05

0.45**

0.36*

0.11

-0.09

-0.08

1.00

0.45**

0.05

0.35*

-0.33*

0.10

0.12

0.35*

0.44**

0.27

-0.25

0.10

1.00

-0.02

0.29*

-0.45**

0.11

0.29

0.35*

0.11

-0.33*

0.11

1.00

0.73**

0.97**

0.34*

-0.22

0.73**

0.18

0.39**

0.20

0.34*

1.00

0.06

0.25

-0.05

0.99**

0.34*

0.30*

0.04

0.25

NIPV

1.00

0.22

-0.24

0.02

-0.16

0.17

0.25

0.22

1.00

-0.29

0.25

0.33*

0.85**

-0.13

1.00**

NTPH

1.00

-0.05

-0.03

-0.08

-0.21

-0.29

TuL

1.00

0.33*

0.36*

0.04

0.25

TuDi

-1.00

0.48**

-0.44**

0.33*

TFW

1.00

-0.29

0.85**

TDW

1.00

-0.13

1.00

*, **Signiﬁcant difference at the 0 .05and 0.01 levels of probability.

LL=Leaf length (cm); LW= Leaf width (cm); PL= Petiole length (cm); VL= Vine length (cm); IL= Internodes length (cm); NIPV= Number of internodes vine^-1, DM= Days to maturity, NTPH= Number of tuber per hill; TL=Tuber length (cm); TDi= Tuber diameter (cm); TFW=Tuber fresh weight (t/ha); TDW=Tuber dry weight (t/ha) and HI= Harvest index (%).

Table 8: Spearman’s rank correlation coefficients showing pair-wise association between qualitative traits with TFW and TuL in yam landraces.

Traits	LC	LS	Lsi	LD	PC	TDI	VC	TS	TSC	TBr	TSt	TFC
TFW	0.03	0.32*	0.06	0.26	-0.28	-0.04	-0.05	0.11	-0.16	0.24	-0.11	-0.26
TuL	-0.24	0.32*	0.06	0.24	-0.28	-0.04	0.37*	0.11	-0.16	0.24	-0.11	-0.25

*Signiﬁcant difference at the 0.05 probability level.

Table 9: Estimates of direct (boldfaced main diagonals) and alternate/indirect path coefficient values of qualitative traits with TFW (top) and TuL (bottom) in yam landraces.

Traits	LC	LS	Lsi	LD	PC	TDI	VC	TS	TSC	TBr	TSt	TFC	TFW
LC	0.177	-0.074	0.022	-0.049	-0.017	0.065	-0.027	0.007	-0.004	-0.022	0.020	-0.068	0.03
LS	0.037	0.349	-0.003	-0.014	-0.01	-0.058	-0.003	-0.016	0.016	0.015	0.001	0.009	0.32*
LsI	-0.033	-0.001	0.116	-0.031	0.005	0.090	-0.077	-0.002	0.002	0.046	-0.019	-0.033	0.06
LD	0.036	-0.021	-0.015	0.238	0.117	0.007	-0.003	-0.001	0.012	-0.050	0.001	-0.064	0.26
PC	-0.014	0.017	-0.002	-0.130	-0.213	-0.018	0.026	0.010	-0.016	0.011	0.023	0.031	-0.29
TDI	0.049	0.087	-0.045	-0.008	-0.017	-0.233	0.0704	0.001	-0.006	0.011	0.013	0.034	-0.04
VC	-0.029	0.006	0.055	0.005	0.034	0.100	-0.162	-0.012	0.036	-0.005	-0.043	-0.032	-0.05
TS	0.023	0.094	0.003	0.064	0.038	0.006	-0.033	-0.059	0.003	0.010	-0.020	-0.022	0.11
TSC	-0.010	-0.073	-0.002	-0.038	-0.046	-0.002	0.076	0.003	-0.077	0.004	0.022	-0.015	-0.16
TBr	0.021	0.028	0.028	-0.063	-0.012	-0.014	0.004	-0.003	-0.002	0.188	-0.006	0.074	0.24
TSt	0.021	0.000	0.012	0.000	0.029	0.018	-0.040	-0.006	0.010	0.007	-0.173	0.017	-0.11
TFC	-0.047	-0.013	0.015	0.061	0.026	0.032	-0.022	-0.005	-0.004	-0.055	0.011-	-0.252	-0.26
Traits	LC	LS	Lsi	LD	PC	TDI	VC	TS	TSC	TBr	TSt	TFC	TuL
LC	-0.186	-0.055	0.016	-0.074	-0.028	0.041	-0.006	0.016	-0.001	0.082	-0.026	-0.019	-0.24
LS	0.029	0.356	-0.003	-0.015	-0.009	-0.060	-0.003	-0.022	0.020	0.021	0.001	0.009	0.32*
LsI	-0.029	-0.001	0.102	-0.034	0.004	0.093	-0.099	-0.002	0.002	0.066	-0.008	-0.032	0.06
LD	0.054	-0.021	-0.014	0.257	0.103	0.008	-0.004	-0.022	0.015	-0.072	0.001	-0.063	0.24
PC	-0.028	0.017	-0.002	-0.141	-0.188	-0.019	0.033	0.014	-0.021	0.016	0.010	0.031	-0.28
TDI	0.032	0.088	-0.039	-0.008	-0.015	-0.241	0.090	0.002	-0.001	0.017	0.006	0.0340	-0.04
VC	-0.006	0.006	0.048	0.005	0.030	0.104	-0.207	-0.016	0.466	-0.007	-0.019	-0.031	0.37*
TS	0.036	0.096	0.003	0.069	0.033	0.006	-0.042	-0.082	0.004	0.015	-0.009	-0.022	0.11
TSC	-0.002	-0.074	-0.002	-0.041	-0.040	-0.002	0.098	0.003	-0.099	0.007	0.010	-0.015	-0.16
TBr	-0.056	0.028	0.025	-0.068	-0.011	-0.015	0.005	-0.004	-0.002	0.272	-0.002	0.072	0.24
TSt	-0.062	0.001	0.011	0.001	0.025	0.018	-0.052	-0.009	0.013	0.010	-0.077	0.016	-0.11
TFC	-0.014	-0.013	0.013	0.066	0.023	0.033	-0.027	-0.007	-0.006	-0.079	0.005	-0.246	-0.25

Table 10: Estimates of direct (bold and underlined) and indirect path coefficient values of TFW with quantitative traits amongst yam landraces.

Traits	LL	LW	VL	PL	DM	NIPV	NVPH	IL	TuL	TDW	HI	TFW
LL	-0.016	0. 056	0.001	0.042	0.210	-0.013	0.031	0.165	-0.399	0.080	0.004	0.11
LW	-0.022	0. 114	-0.015	0.024	1.389	-0.065	0.004	-0.838	0.039	0.062	-0.319	0.27
VL	0.008	0.053	-0.023	-0.083	0.780	0.113	0.005	-0.199	0.039	0.054	-0.350	0.39**
PL	-0.126	0.284	0.067	-0.034	0.181	-0.028	0.005	-0.474	0.454	-0.229	0.013	0.11
DM	0.054	-0.202	0.120	-0.066	1.018	-0.114	-0.001	-1.068	1.020	0.127	-0.029	0.85**
NOI	0.040	-0.010	-0.001	-0.032	0.169	0.022	-0.039	-0.049	0.164	0.003	-0.095	0.17
NVPH	-0.260	0.064	-0.187	0.027	0.458	-0.219	-0.065	0.556	-0.509	0.002	0.054	-0.08
IL	-0.195	0.305	0.081	-0.172	0.369	-0.154	0.014	-0.238	0.250	0.117	-0.064	0.30*
TuL	-0.023	0.141	0.108	-0.070	0.216	-0.046	0.016	-0.239	0.213	0.116	-0.071	0.36*
TDW	0.146	-0.129	0.002	-0.093	0.224	-0.143	-0.077	-0.484	0.242	0.053	-0.037	-0.29
HI	-0.179	0.086	0.042	0.082	1.234	-0.208	0.004	-0.254	0.060	0.060	-0.080	0.85**

LL=Leaf length(cm); LW= Leaf width(cm); VL= Vine length(cm); PL= Petiole length (cm); DM= Days to maturity,NIPV=Number of internodes/vine;NVPH= Number of vine per hill;IL=Internodes length (cm);TuL=Tuber length(cm); TDW=Tuber dry weight (t/ha) and HI= Harvest Index (%).

Table 11: Estimates of direct (bold and underlined) and indirect path coefficient values of TuL with quantitative traits amongst yam landraces

Traits	LL	LW	VL	PL	DM	NIPV	NVPH	IL	TFW	TDW	HI	TuL
LL	0.112	0.009	-0.001	-0.006	-0.001	0.299	0.007	0.009	0.007	-0.002	0.008	0.45**
LW	0.551	-0.31	0.001	-0.067	0.001	0.186	0.021	0.096	0.007	-0.094	-0.029	0.35*
VL	1.409	-0.009	0.001	-0.012	0.008	-0.89	0.016	0.078	0.067	0.063	-0.001	0.73**
PL	0.228	0.01	-0.021	-0.007	0.001	0.119	0.009	0.011	0.002	-0.072	0.001	0.29
DM	0.336	-0.006	0.001	0.001	0.001	-0.125	0.011	-0.002	0.031	0.017	-0.001	0.25
NOI	0.005	0.002	0.001	-0.007	0.001	0.012	0.001	0.002	-0.003	0.001	-0.001	0.02
NVPH	0.001	0.005	-0.031	-0.027	-0.001	0.050	0.009	0.001	-0.02	-0.02	-0.011	-0.05
IL	0.225	-0.001	0.001	0.002	-0.001	0.458	0.094	0.003	-0.005	0.211	0.001	0.99**
TFW	0.699	0.019	-0.001	-0.002	-0.001	-0.395	0.024	0.002	0.008	0.001	0.001	0.36*
TDW	0.505	0.001	0.001	-0.007	0.001	-0.473	0.007	0.002	-0.004	0.002	-0.001	0.04
HI	-1.254	-0.001	0.001	0.006	-0.001	1.503	0.004	0.006	-0.016	-0.013	0.001	0.25

LL=Leaf length(cm); LW= Leaf width(cm); VL= Vine length(cm); PL= Petiole length (cm); DM= Days to maturity, NIPV=Number of internodes/vine;NVPH= Number of vine per hill; IL=Internodes length (cm);TFW=Tuber fresh weight (t/ha); TDW=Tuber dry weight (t/ha) and HI= Harvest Index (%).

4. DISCUSSION:

Knowledge on the nature of associations among economically important traits is essential for direct or indirect selection and could consequently improve the efficiency of selection gains in plant breeding programs. In this study, correlation and path coefficient analyses were used to determine associations between qualitative and quantitative traits and also to determine the best selection criteria. Simple correlations among qualitative traits showed positive associations between leaf color and petiole color (Table 6). Positive correlations were also noted between leaf shape and twinge direction. Further, highly signiﬁcant and negative correlations were observed between leaf size and density, leaf density and twinge direction as well as twinge direction and tuber shape. Further, petiole color, twinge direction, tuber surface texture and tuber flesh color were negatively correlated with tuber fresh weight and length (Table 8) signifying that these traits may negatively limit yield gains.

Correlations between qualitative traits have not been reported in yam. Therefore, this study could be very useful in commercial breeding of yam. Further, the lack of correlation between leaf density and tuber shape with other qualitative traits may be of interest for further investigation. Simple correlation analysis among quantitative traits revealed highly signiﬁcant (p< 0.001) and positive correlations between leaf length with leaf width (r=0.77), internodes length (r=0.45) and TuL (r =0.45) (Table 7). This suggests that targeted selection to improve these traits would increase TuL (Mulualem and Dagne, 2013). Results in this study are in agreement with the report of Pandey et al. (1996), Alam et al. (2014) and Mashilo et al. (2016) who independently worked on taro, greater yam and bottle gourd landrace collections. The authors indicated that several quantitative traits including plant height, number of female flowers and number of branches per plant were important yield components inﬂuencing storage TFW in yam. Tuber length signiﬁcantly and positively correlated with leaf length (r=0.45) and leaf width (r=0.35), vine color (r=0.73) and internodes length (r=0.99), respectively (Table 7). So far, there are no studies that reported correlation of storage TFW with related traits in yam.

Associations among traits are often expressed using simple correlation coefficients. This may limit prediction on the success of selection. Therefore, correlation coefficients between various characters were partitioned into direct and indirect effects using path coefficient analysis. Path coefficient analysis is increasingly being used in plant breeding to improve selection efficiency through pinpointing traits with signiﬁcant effects on yield or yield components (Dominic et al., 2014). In this study, positive direct effect was exhibited between TFW and TuL. High direct effects were recorded between leaf shape followed by leaf density, tuber branching, leaf color and leaf size with TFW. Furthermore, these traits were well-correlated with TuL (Table 9). The poor correlation of leaf size and tuber shape with tuber yield suggests that direct selection for these traits may not provide storage tuber yield improvement. High positive direct effect was exhibited between days to maturity and storage TFW (Table 10) suggesting simultaneous selection of the two traits may improve genetic gain of tuber yield in yam breeding. This finding is in agreement with Mulualem and Weldemichael (2013) who reported that TuL is an important character in making selection in aerial yam. Further, Tsegaye et al. (2006) reported that storage TuL and dry matter contents are the best character to select Ethiopian sweet potato accessions. Hence, selection on the bases of early days to maturity and increased TuL may maximize storage tuber yield in yam. The negative direct effect of internodes length on TFW may be explained by the fact that selection based on internodes length might reduce TFW. Similar finding was reported by Monkola (2013) who indicated that the direct effect of internodes length on TFW of greater yam was small and negative. Whereas these are in contradiction to the result of Kifle (2006) and Dominic et al. (2014) who reported that number of verticals contributed more for tuber yield on taro and cassava, in that order. However, TuL, leaf width, number of internodes per vine and tuber dry weight had positive direct effects (Table 10). Vine length also had positive indirect effect on TFW through most of the traits except leaf width and number of internodes per vine where the direct effect of vine length was negative (Table 10). Vine length components seem to have less competitive effect with TFW at path coefficient analysis level and have not been selected against. LW had negative indirect effect through days to maturity, number of internodes per vine and tuber dry weight. It is interesting to note that days to maturity itself had positive direct effect on TFW and positive indirect effect through all characters except number of vines per hill. Mulualem and Dagne (2013) described the similar result. These authors reported that days to maturity being the novel character and had higher direct effect on fresh root yield on cassava.

The low negative association of tuber dry weight with TFW, which is not as such important on the basis of correlation estimates, revealed positive direct and indirect supplier to TFW via path analysis. Thus, selecting accessions based on this character would contribute for rapid yam tuber yield enhancement programmes. Leaf width had comparatively high positive direct effect (0.114) on TFW. Besides, it had positive and non-significant association with TFW. This is in agreement with Norman et al. (2011) who reported that taro accessions with wider leaves had high TFW. It had high negative indirect effects via days to maturity and tuber dry weight and negligible indirect negative effect through number of internodes per vine. Therefore, it is important to consider accessions with wider leaves in improving tuber yield in yam, as it was positively associated with TFW and its direct and indirect positive effect through yield contributing traits to TFW. The direct effect of TuL on TFW was positive and high (0.213). Positive direct effect of TuL on TFW was also reported by Mulualem and Weldemichael (2013) on aerial yam. Besides, the indirect effect of TuL on TFW through petiole length, days to maturity, tuber dry weight, number of vines per hill, internodes length and number of internodes per vine was high and positive. In contrast, the indirect influence through leaf length was higher and negative. Therefore, selection based on TuL is important to maximize TFW. Internodes length had significant and positive association with TFW. This positive association did not contribute to fresh tuber yield directly but indirectly through petiole length, days to maturity, vine length, tuber dry weight and TuL. However, harvest index and vine length had contributed indirectly to TFW. The value of harvest index had small and negative direct (-0.080) effect on TFW. Therefore, efforts required in breeding cultivars with a higher length of storage tuber can be achieved through selection of landraces. Landraces which produced the highest TFW were 10/002, 56/76, 17/02, 27/02, 7/83, 08/02, 59/02, 45/03 and 6/02. A high path coefficient value indicates that the change will result in a proportional (or inversely proportional) change in another correlated trait, whereas a strong correlation coefficient indicates that the change will have high effect on the second trait (Cramer and Wehner, 2000). Landraces which produced the highest TuL included: 27/02, 68/01, 08/02, 6/02, 75/02, 56/76, 13/87, 116, 39/87 and 07/03. These are useful genetic resources for yam breeding emphasizing storage tuber yield.

Knowledge of the associations among qualitative traits is important for simultaneous selections especially for yam which shows considerable variability in tuber qualitative parameters. The present study further showed that selection for qualitative traits such as LS may inﬂuence TuL whereas selection for leaf size and leaf density may not inﬂuence TFW and TuL in yam.

5. CONCLUSION:

Overall, the correlation and path analyses allowed selection of the yam landraces such as: 27/02, 56/76, 08/02, 10/002, 39/87, 45/03, 6/02, 116 and 7/83 which can be very useful landraces in breeding and conservation of this valuable horticulture crop in Ethiopia.

6. ACKNOWLEDGEMENTS:

Ethiopian Institute of Agricultural Research (EIAR) and Jimma Agricultural Research Center (JARC) are acknowledged for ﬁnancial support of this study.

7. REFERENCES:

1. Alam, S, Euphemia, S, Bora, P., Saud, BK., 2014. Genetic variation in different cultivars of greater yam (Dioscorea alata). Journal of root crops, 40(1): 1-5.

2. Asiedu, R, Alieu, S., 2010. Crops that feed the world 1. Yams, Journal of food Science, 2:305-315.

3. Bhatt, GM., 1973. Signiﬁcance of path coefﬁcient analysis in determining the nature of character association. Euphytica, 22:338–343.

4. Christopher, SC., 2000. Path Analysis of the correlation between fruit number and plant traits of cucumber populations. HortScience, (4):708–711.

5. Coursey, DG., 1967. Yams: an account of the nature, origins, cultivation and utilization of the Useful members of the Dioscoreaceae.Longmans. Green and Co Ltd (London), p 230.

6. Cramer, CS., Wehnerm TC., 2000. Path analysis of the correlation between fruit number and plant traits of cucumber populations. HortScience. 35:708–711.

7. Dansi, A, Dantseyand, BH., Vodouhè, R., 2013. Production constraints and farmers’ cultivar preference criteria of cultivated yams (DioscoreacayenensisDioscorearotundata complex) in Togo.International J. Biol. Chem. 9(1): 388-408.

8. Dewey, DR, Lu, KH., 1959. A correlation and path coeffic ients analysis of components of crested wheat grass seed production. Agronomy Journal, 51:515-518.Dominic, AO., Ukaobasi, E., Ogon T., 2014. Association and Path Coefficients Analysis of Fresh Root Yield of High and Low Cyanide Cassava (Manihot esculenta Crantz) Genotypes in the Humid Tropics. Journal of Crop Science and Biotechnology,17(2): 103-109

9. Dominic AO., Ukaobasi E, Ogon T. 2014. Association and path coefficients analysis of fresh root yield of high and low cyanide cassava (Manihot esculenta Crantz) genotypes in the humid tropics. Journal of Crop Science and Biotechnology,17(2): 103-109

10. Falconer, DS., Mackey, FC., 1996. Introduction to quantitative genetics. 4th ed. London: Longman Genres, 2008. Genres statistical software User’s Guide Version 16, Genres Inc. USA.

11. Girma, G., Korie, S., Dumet, D., Franco, J., 2012. Improvement of accession distinctiveness as an added value to the global worth of the yam (Dioscoreaspp) genebank. International Journal of Conservation Sciences, 3(3):199-206.

12. Gomez, KA., Gomez, AA., 1984. Statistical Procedures for Agricultural Research. 2nd ed. John Wiley and Sons, inc., New York.

13. Hildebrand, EA., 2003. Motives and opportunities for domestication: an ethno- archaeological study in southwest Ethiopia. Journal of Anthropological Archeology, 22:358-375.

14. IPGRI/IITA., 1997. Descriptors for yam (Dioscorea spp.). International Institute for Tropical Agriculture, Ibadan, Nigeria/International Plant Genetic Resources Institute, Rome, Italy, p.66.

15. Kifle, A., 2006. Characterization and divergence analysis of some Ethiopian taro [Colocasiaesculenta (L.)] accessions M.Sc. Thesis, Haramaya University, Haramaya, Ethiopia, p.84.

16. Lebot, V., 2009. Tropical tuber and tuber crops: cassava, sweet potato, yams, aroids. CABI: Wallingford, Oxfordshire.

17. Li, CC., 1956. The concept of path coefﬁcient and its impact on population genetics. Biometrics, 12:190–210.

18. Mashilo, J., Hussein, S., Alfred, O., 2016. Correlation and path coefficient analyses of qualitative and quantitative traits in selected bottle gourd landraces. Acta Agriculturae Scandinavica, Section B. Soil and Plant Science, 1-13.

19. Mohammadi, SA., Prasanna, BM., Singh, NN., 2003. Sequential path model for determining interrelationships among grain yield and related characters in maize. Crop Sci., 43:1690–1697.

20. Monkola, M., 2013. Genetic variability and association among yield and yield related traits in cassava (Manihotesculenta Crantz) in southwest Ethiopia. MSc. Thesis presented at School of Graduate studies, Jimma University, Jimma, Ethiopia. p.85.

21. Mulualem, T., Dagne, Y., 2013.Studies on correlation and path analysis for root yield and related traits of Cassava (ManihotesculentaCrantz) in South Ethiopia.Journal of Plant Sciences,1(3):33-38.

22. Mulualem, T., Weldemichel, G., 2013. Agronomical evaluation of aerial yam (Dioscorea bulbifera) accessions collected from South and Southwest Ethiopia. Greener Journal of Agricultural Sciences, 3(9): 693-704.

23. Mulualem, T., Weldemichael, G., Benti, T., Walle, T., 2013.Genetic Diversity of Cassava (Manihot esculenta Crantz) Genotypes in Ethiopia. Greener Journal of Agricultural Sciences, 3(9): 636-642

24. Norman, PE., Tongoona, P., Shanahan, P.E., 2011. Determination of interrelationships among agro morphological traits of yams (Dioscorea spp.) usingcorrelation and factor analyses. Journal of Applied Biosciences, 45: 3059– 3070.

25. Norman, P. E., Tongoona P., Danson J., Shanahan, PE., 2012. Molecular characterization of some cultivated yam (Dioscorea spp.) genotypes in Sierra Leone using simple sequence repeats. International Journal of Agronomy Plant Production, 3(8): 265-273.

26. Pandey, G., Dhobal, V, Sapra, RL., 1996. Genetic variability, correlation and path analysis in Taro [(Colocasia esculenta (L.)]. Journal ofhill. Res.,9: 299-302.

27. Paul, KK., Bari, MA., Debnath, SC., 2013. Correlation and path coefficient studies of yield and yield attributing characters in taro [(Colocasia esculenta (L.)]. Journal of Bangladesh Academy of Sciences, 37(2): 131-137

28. SAS, Institute. 2000. Statistical Analytical Systems SAS/STAT user’s guide version, 8(2) cary NC: SAS institute inc.

29. Sesay, L, Norman, P.E., Massaquo, i A., Gboku, M.L., Fomba, SN., 2013. Assessment of farmers’ indigenous knowledge and selection criteria of yam in Sierra Leone. Sky Journal of Agricultural Research, 2(1):1–6.

30. Statistical Package for Social Sciences., 1996. SPSS for windows. User’s guide: Statistics version 16. Inc. Cary, NC.

31. Tamiru, M., Heiko, C.B, Brigitte, LM., 2011. Comparative analysis of Morphological and farmers cognitive diversity in yam landraces (Dioscorea spp.) from Southern Ethiopia. Tropical Agriculture development, 55(1): 28-43.

32. Tsegaye, E., Devakara, FV., Nigussie, D., 2006. Correlation and path analysis in sweet potato and their implication for clonal selection. Journal of Agronomy, 5(3):391-395.

33. Williams, WA., Jones, MB., Demment, MW., 1990. A concise table for path analysis statistics. Agronomy Journal, 82:1022–1024.

34. Wright, S. ,1921. Theory of path coefficients. Genetics. 8:239–255

35. Wright, S., 1934. The method of path coefficients. Annal Mat Stat. 5:161–215.

36. Zeven, A.C., De Wet, J.M., 1982. Dictionary of cultivated plants and their region of diversity: Center for Agricultural Publication and Documentation, (Wageningen), p.259.

Received on 05.08.2020 Modified on 25.08.2020

Res. J. Pharmacognosy and Phytochem. 2020; 12(4):207-218.

DOI: 10.5958/0975-4385.2020.00035.7