EVALUATION OF WATER YAM ( DIOSCOREA ALATA L ) GENOTYPES FOR YIELD AND YIELD COMPONENTS IN ABAKALIKI AGRO-ECOLOGICAL ZONE OF NIGERIA

Six Dioscorea alata genotypes were evaluated alongside a popular cultivar in the locality for yield and yield components in 2004 and 2005 at the Department of Crop Production & Landscape Management, Ebonyi State University, Abakaliki. The experiment was laid out in a Randomized Complete Block Design (RCBD) with three replications. Results indicated significant effects of genotype, cropping season and their interaction on most of the traits studied. Genotypes with potentials for multiple-tubering were identified. The genotype, TDa 98/01176 produced the highest total number of tubers, number of ware tubers and number of seed tubers per plot. Similarly, TDa 98/01176 had the highest total fresh tuber yield. The lowest values for these traits were obtained for Okwalenkata, a landrace genotype. The genotypes performed better in 2004 than in the 2005 cropping season, and three genotypes TDa 9801176, TDa 98/01169 and TDa 98/01166 had the most stable yield across the two cropping seasons. The tuber yields of these genotypes were higher than the mean yield of all the genotypes. Key word: Dioscorea alata, genotypes, cropping season, evaluation, yield components INTRODUCTION Yams (Dioscorea spp.) are important tuber crop and staple food for millions of people in many tropical and subtropical countries. It has become a cash crop in countries along the coast of West Africa (Egesi et al., 2005). They produce edible starchy storage tubers, which are of cultural, economic, and nutritional importance in the tropical and subtropical regions of the world (Coursey, 1967). Generally, yam tubers are boiled, roasted, baked, or fried. However, in some regions, the tubers are boiled and then pounded to glutinous dough called "fufu." Cooked yam tubers or their products are usually eaten in association with protein-rich sauces. It has also gained importance in pharmaceutical industries (Hahn, 1995). Nigeria accounts for about 70% of the world’s production of yam, generating a global annual output of over 33 million metric tones (FAO, 2000). D. alata is popular and prevalent in Abakaliki agro ecological zone of Ebonyi State, Nigeria where it is called “Mbala or Nvula” (Igbo names). This popularity is attributed to its rapid multiplication; high yields and extended shelf life. The slicing of the tuber for planting is also very simple and it multiplies in quadruplicates. Yam yields are influenced by numerous environmental factors such as water (soil moisture), temperature, light and photoperiod (Orkwor and Asadu, 1998). Other constraints to yam production include biotic factors such as pests and disease organisms in the field and in the store. These factors have led to decrease in production over the years and have prompted breeding activities to generate high yielding varieties with some tolerance to environmental stresses. Therefore, the objective of the experiment was to evaluate seven D. alata genotypes for their yield and performances in Abakaliki agro-ecological zone of southern Nigeria.


INTRODUCTION
Yams (Dioscorea spp.) are important tuber crop and staple food for millions of people in many tropical and subtropical countries.It has become a cash crop in countries along the coast of West Africa (Egesi et al., 2005).They produce edible starchy storage tubers, which are of cultural, economic, and nutritional importance in the tropical and subtropical regions of the world (Coursey, 1967).Generally, yam tubers are boiled, roasted, baked, or fried.However, in some regions, the tubers are boiled and then pounded to glutinous dough called "fufu."Cooked yam tubers or their products are usually eaten in association with protein-rich sauces.It has also gained importance in pharmaceutical industries (Hahn, 1995).
Nigeria accounts for about 70% of the world's production of yam, generating a global annual output of over 33 million metric tones (FAO, 2000).D. alata is popular and prevalent in Abakaliki agro ecological zone of Ebonyi State, Nigeria where it is called "Mbala or Nvula" (Igbo names).This popularity is attributed to its rapid multiplication; high yields and extended shelf life.The slicing of the tuber for planting is also very simple and it multiplies in quadruplicates.
Yam yields are influenced by numerous environmental factors such as water (soil moisture), temperature, light and photoperiod (Orkwor and Asadu, 1998).Other constraints to yam production include biotic factors such as pests and disease organisms in the field and in the store.These factors have led to decrease in production over the years and have prompted breeding activities to generate high yielding varieties with some tolerance to environmental stresses.Therefore, the objective of the experiment was to evaluate seven D. alata genotypes for their yield and performances in Abakaliki agro-ecological zone of southern Nigeria.Experimental Design: Setts of about 200 g of D. alata genotypes treated with Emofarm dust (mixed with water) were planted in a Randomized Complete Block Design (RCBD) with three replications, under rainfed conditions.Plant spacing was 1 m x 1 m in ridges.The yam vines were staked with about 2m stakes one month after emergence.Fertilizer application (N.P.K 12-12-17, 400 kg/ha) was by broadcast ten weeks after planting, while weeding was done three times before harvest.

MATERIALS AND METHODS
Data Collection and Analysis: Data were collected on the following yield parameters at harvest: total number of tubers harvested from the sample plots, number of ware tubers (tubers above 1 kg) harvested per plot, number of seed tubers (tubers below 1 kg) harvested per plot, yield of ware tubers (t/ha), yield of seed yams (t/ha), total fresh tuber yield (t/ha).A combined analysis of variance for the two-year data (2004 and 2005) was done using the GLM procedure in SAS (SAS Institute, 1999), following the model below: Where, Y ijk = the observed performance of the i th genotype in the j th cropping season in the k th replication, µ = the overall mean of the trait, α i = the effect of the i th genotype (i=1,2,..,7), β j = the effect of the j th cropping season, (j=1 and 2) (αβ) ij = the genotype x cropping season interaction γ k = the effect associated with the k th replication (k=1,2,3) ε ijk = the residual associated with each observation.
Thereafter, reduced models, which excluded cropping season effects (separate analysis for each cropping season), were used to estimate performance of the genotypes for each cropping season.Thus, comparisons were made between estimates based on combined analysis and those using data subsets for cropping season.This would indicate genotypes that were consistent in performance across cropping seasons.
Genotype by trait analysis using GGE biplot analysis (Yan and Kang, 2003) was used to determine which variety was best and for what trait.This would aid selection of genotypes for the agro-ecological zone.

RESULTS AND DISCUSSION
Variance estimates of D. alata genotypes for yield and yield components: Significant genotype effect (P < 0.05) on total number of tubers harvested was recorded in this experiment, while number of ware tubers harvested per plot, yield of ware tubers and total fresh tuber yield (t/ha) was very highly significant (P < 0.001).The effect of genotype on number of seed tubers harvested per plot and yield of seed tubers was not significant (Table 1).Cropping season effect on the number of seed tubers harvested per plot and yield of seed tubers (t/ha) was highly significant (P < 0.01), and very highly significant (P < 0.001) for number of ware tubers harvested per plot, yield of ware tubers (t/ha) and total fresh tuber yield (t/ha).
Genotype x cropping season interaction was significant for yield of ware tubers (t/ha) and total fresh tuber yield (t/ha) only.

Mean performance of D. alata genotypes for yield and yield components:
The highest number of tubers harvested in this experiment was obtained for TDa 98/01176 (126.7).This was not different from value obtained for TDa 98/01169 but differed significantly from the values obtained for the other genotypes which were statistically the same (Table 2).Similarly, the highest mean number of ware tubers harvested per plot was obtained for TDa 98/01176 (43.3), followed by TDa 98/01166 (36.5).The values obtained for TDa 98/01169, TDa 98/01168 and TDa 240 did not differ significantly.The lowest mean number of ware tubers was obtained for Okwalenkata, a landrace genotype.
The highest mean yield of ware tubers (15.8 t/ha) was obtained for TDa 98/01176.This was not statistically different from values obtained for TDa 240 (15.0 t/ha), TDa 98/01166 (13.0 t/ha) and TDa 98/01168 (12.7 t/ha), but differed significantly from the value obtained for TDa 98/01169 (Table 2).The lowest value (5.5 t/ha) was obtained for 'Okwalenkata', the popular landrace in the locality.A similar trend was observed for total fresh uber yield.The highest value was obtained for TDa 98/01176 (23.8 t/ha), followed by TDa 240 (19.3 t/ha) and TDa 98/01169 (19.0 t/ha), while Okwalenkata had the lowest value for the trait.The new genotypes out yielded both the Institutional check (TDa 297) and the popular landrace (Okwalenkata) indicating a possible adoption of the new genotypes instead of the landraces especially if their functional and physicochemical properties are desirable.This result is expected because hybrids have been known to perform better than their unimproved counterparts.According to Obi (1991), hybrids are products of two or more parents of good agronomic characteristics and in most cases should perform better than either of their parents.
TDa 9801176 ranked the best among the genotypes studied, clustering with most of the traits based on Genotype by Trait plot (Fig. 1).Similarly, the means versus stability plot indicated TDa 9801176, TDa 98/01169 and TDa 98/01166 to be more stable and better adapted in Abakaliki agroecological zone (Fig 2).Future preference of these genotypes over the popular cultivar in the locality may be strengthened by the findings of Udensi et al., (2006) that two of these genotypes, TDa 99/00169 and TDa 98/01176 have high crude protein level and could be selected for intensive cultivation in Nigeria and other D. alata growing regions.3) for number of ware tubers harvested per plot, yield of ware tubers (t/ha) and total fresh tuber yield (t/ha), while the reverse was the case for number of seed tubers harvested per plot and yield of seed tubers (t/ha).In terms of total fresh tuber yield, the most responsive genotypes in 2004 in descending order were TDa 98/01176, TDa 240, TDa 98/01169 and TDa 98/01168, while in 2005 the most responsive genotypes were TDa 98/01176, TDa 98/01166, TDa 98/01169 and TDa 240 (Table 4).In the two cropping seasons, TDa 98/01176 was stable in performance giving the best yield.Multiple tubering is a desirable characteristic necessary for commercial yam seed production.The ability of some genotypes, especially TDa 98/01176 to consistently produce multiple tubers profusely in 2004 and 2005 indicates their potential value to commercial seed production.This characteristics would ensure availability of planting materials and would reduce cost of materials for field production on a per hectare bases as suggested by Okoli and Akoroda (1995) and Orkwor and Asadu (1998).In addition, low sett multiplication ratio of yam and dormancy impede breeding and selection programmes.For example, it will take several generations and seasons to obtain enough planting materials to evaluate a few clones of yam (Okoli and Akoroda, 1995).This provided an explanation for the selection of TDa 9801176 for multiple tuberisation in the environment.
The difference in the yield performance of the genotypes in the two cropping seasons was attributed to the level of adaptability of the genotypes to different prevailing biotic and abiotic factors.There is a strong indication that selecting genotypes based on mean yield of one cropping season alone would be inappropriate.Genotype yield and stability of performance over two or more cropping seasons would be necessary for evaluation of genotype performance.In both years, the most outstanding genotype was TDa 98/01176 having values that were above the mean for all the traits.Practical integration of yield and stability performance as a parameter in selection programs is valuable and greater emphasis on it would benefit breeders and farmers, as it will indicate the best environmental treatment for good performance of crop varieties (Egesi et al., 2005).
Based on genotype-by-cropping season interaction effects on mean fresh tuber yields, the recommended genotypes for Abakaliki agroecological zone were TDa 9801176, TDa 98/01169 and TDa 98/01166.Generally, farmers would prefer to use a high yielding genotype that performs consistently across different years and environments (Kang et al., 1991;Kang, 1993).Another approach could be to develop genotypes with specific adaptations, which are better suited to a particular environment.

Table 1 . Variance estimates for yield and yield components of D. alata genotypes after two cropping seasons in Abakaliki agro-ecological zone
Evaluation of Water Yam (Dioscorea Alata L) Genotypes For Yield And Yield Components in Abakaliki

Table 3 . Cropping Season effects on the performance of D.alata genotypes grown in 2004 and 2005 in Abakaliki agro-ecological zone
Oselebe, H.O. and Okporie E. O.

Table 4 . Mean yield of ware tubers and total fresh tuber yield (t/ha) of D.alata genotypes in 2004 and 2005 cropping seasons in Abakaliki agro-ecological zone.
Evaluation of Water Yam (Dioscorea Alata L) Genotypes For Yield And Yield Components in Abakaliki