Cassava (Manihot esculenta Crantz) is a major staple food in Nigeria, consumed daily by more than 100 million people. The global production of cassava in 2014 was 278.7 million tons with an estimate of 281 million tons  for 2015 and 288.4 million tons for 2016 (FAO, 2016). From available records, Nigeria still stands out as the world’s largest producer of cassava with a progressive production pattern   that   increased  from  42.5   million  metric   tons in 2010 to 54 million metric tons in 2012 with average production output of 12.2 t/ha in 2010 increased to 14.03 t/ha in 2012 (FAOSTAT, 2013). Total area harvested of the crop in 2012 was 3.85 million ha (FAOSTAT, 2013).  The diverse uses of cassava largely explain its popularity in the tropics. In Africa, most cassava produced is used for food consumption with 50% in processed form and 38% in the fresh and/or boil form; and 12% is used for animal feed. The diverse uses of cassava largely explain its popularity in the tropics. However, many pathogens and pests reduce cassava yields, especially in Africa including Nigeria. Diseases such as cassava mosaic disease (CMD), transmitted by a whitefly (Bemisia tabaci) vector and spread by infected cuttings, cassava brown streak virus disease (CBSD), cassava bacterial blight (CBB; Xanthomonas axonopodis pv. manihotis), and anthracnose (CA; Colletotrichum gloeosporoides) are among the most important diseases. Pests with a wide African spread are the cassava mealybug (CM; Phenacoccus manihoti), African root and tuber scale (Stictococcus vayssierei), cassava green mite (CGM; Mononychellus tanajoa) and nematodes (particularly Meloidogyne species) (Abaca et al., 2014).

Over-dependence on cassava-based diets may result in poor health, stunted growth, reduced capacity for physical activity, and in extreme cases, a high incidence of anaemia, corneal blindness, and compromised immunity (Saltzman et al., 2013). However, while the commonly available white cassava can provide most of the body’s daily energy requirements, it does not provide sufficient proteins, essential micronutrients and vitamin A, required for a healthy and productive life. Vitamin A deficiency can impair the body’s immunity to infectious diseases and cause eye defect that can lead to partial or complete blindness. Nearly one in three Nigerian children under five and one-quarter of all pregnant women in the country are vitamin A deficient (FAO, 2014). Billions of people around the world suffer from hunger and ‘hidden hunger’ or micronutrient malnutrition. Around 805 million people were considered chronically undernourished over the 2012 to 2014 period (FAO, 2014).

Those that do not get enough vitamin A and micronutrients (zinc and iron) from the foods they eat face severe health complications and even death. Micronutrient malnutrition can lower intelligence quotient (IQ), cause stunting and blindness in children, lower resistance to disease in both children and adults, and increase risks for both mothers and infants during childbirth. Malnutrition is the underlying cause of 45% of child deaths under the age of 5 (WHO, 2015).  In 2013, an estimated 161 children under the age of 5 were stunted (below median height for age) and another 51 million were wasted (below median weight for height) (Thompson et al., 2013).

This is especially true in regions with prolonged dry seasons that limit production and access to alternative sources of micronutrients such as fresh vegetables  (Von Grebmer et al., 2014).

Pro-vitamin A varieties that are presently available provide up to 40% of the daily recommended vitamin A intake for children less than 5 years old (De Moura et al., 2015). Nevertheless, new crosses to select varieties with an even higher content of β-carotene varieties are being generated through recurrent selection breeding scheme (Sánchez et al., 2014). This paper reports on the agronomic performance and suitability for quality garri production of adapted beta-carotene rich (Pro-vitamin A) cassava clones in diverse locations (Ibadan, Mokwa, and Onne) in Nigeria.

The fields were plowed, harrowed and ridged at 1 m apart. Mature stem cuttings (0.25 m long) of 25 genotypes including three checks were planted on plots of four ridges (Table 6). The ridges were about 50 cm high, each 10 m long and spaced 1 m apart. The plot size was 4 m × 10 m (40 m²). The experimental design was a randomized complete block with four replications. Blocking was done according to the topo-sequence of the field. The plots were weeded six times after planting and no fertilizers were applied. The experiment was conducted in three locations (Mokwa, Ibadan, and Onne) in Nigeria for two seasons (Figure 1). Mokwa (Niger State) is located in the southern Guinea savanna zone with latitude 9°18’N and longitude 5°04’E at about 457 m altitude about sea level (masl) and has a unimodal rainfall pattern with an annual total of 1069 mm, falling between June and October. Radiation is about 450 MJ m^-2 year^-1. The soil is alfisols and ultisols. The second environment was Ibadan (Oyo State) with latitude 7°31’N and longitude 3°54’E and is located in the forest savanna transition zone at about 150 masl. It is characterized by a bimodal rainfall also averaging 1300 mm annually, most of which falls between May and October. Radiation is about 5285 MJ m^-2 year^-1. The soil is slightly acidic alfisols. The third test environment, Onne (Rivers State), latitude 4° 43’N, longitude 7° 01’E, and 10 masl is in the rainforest zone, has a unimodal rainfall pattern with an annual average of 2400 mm falling between February and December. Relative humidity remains high throughout the year, with average values ranging from 78% in February to 89% in July and September. This site receives an average 4 h of direct sunshine daily, reaching 5060 MJ m^-2 year^-1. The soil is representative of highly leached acid ultisols.

Data collection

One month after planting, data was collected on Cassava Mosaic Diseases (CMD). Cassava Bacterial Blight (CBB) was scored monthly until 6 months after planting. Cassava anthracnose disease (CAD) was scored for 6 months after planting and monthly till 9 months after planting. Cassava green mite (CGM) was scored for between January and February. That was when it normally appeared and reached its peak period. The scale used for scoring was 1 to 5 (1= zero attack or resistance; 2 = little attack or little resistance; 3 = medium or moderate resistance; 4 = high attack or susceptible; and 5 = very high attack or highly susceptible).

Number of cassava plants sprouted at 1 MAP was counted and scored as number sprouted or germinated over total number planted.

The plant growth vigor at one month after planting was rated visually, per plot basis, using 3 for low vigor, 5 for intermediate vigor, and 7 for highly vigor cassava plants. Root size was categorized   into  small,  moderate  and  large  with  the  scale   3 = small, 5 = moderate, and 7 = large. For B-carotene, Provitamin A carotenoids represent precursors to vitamin A in humans. It was scored at harvest with the use of color chart: 1= white, 2= light cream, 3= cream, 4= light yellow, 5= yellow, 6=yellow deep, 7= orange and 8= pink. The root cyanide content was estimated by picrate acid method. It was rated on a 1 to 9 scale based on intensity of red color (higher intensity of higher HCN content of root sample): 1= <10 HCN; 2= 10-15 HCN; 3 = 15-25 HCN; 4 = 25-40 HCN; 5 = 40-60 HCN; 6 = 60-85 HCN; 7 = 85-115 HCN; 8 = 115-150 HCN; and 9 = >150 (Intense red). Taste of boiled roots was examined by panel of five people and the conclusion was recorded. The scale used was 1: sweet, 2: bland, and 3: bitter.

Dry matter content of the tuberous root is an important character for the acceptance of cassava roots by consumers who boil or process them. Fresh sample of 100 g were taken from each clone in each replicate and dried at 70°C in oven and they were re-weighed after 72 h of drying and have attained constant drying. The dried sample was weighed and root dry matter content percentage was calculated as the ratio between fresh weight (FW) and dry weight (DW), that is, DM (%) = (DW/FW) × 100.

For fresh root yield, all the underground roots per plot were weighed in kilogram (kg) and converted to tones per hectare (t/ha).

Garri yield is the weight of garri (a dried granule gotten from 10 kg of fresh cassava roots of each clone, after being peeled, grated, fermented, dewatered, fried and sieved) measured in kilogram.

Harvest index was calculated by (root weight / root weight + shoot weight) × 100. 

Statistical analysis

The data collected were subjected to analysis of variance (ANOVA) using General Linear Model (GLM) procedure for randomized complete  block  design in statistical analysis system (SAS, 1996) to test for the treatment of effect and significant interaction of the variables considered. The results of the different experiments were subjected to combined analysis of variance to examine genotype by environment interaction (G × E) effect and standard errors were calculated for each trait.

Plant height

Plant height was measured and the mean was calculated. The analysis of variance (Table 4) of the plant height indicated significant differences among the clones at probability level of 0.05 in Onne with CV of 17%, while in Ibadan and Mokwa a non-significant difference was observed.

Vigor

The result of vigor in G × E (Table 5) was very significant with CV of 17.90%. In Ibadan, the CV was 10.66% and not significant. In Onne, the CV was 21% and showed significance and in Mokwa it was 18% and significant. This result shows that vigor is a stable trait and not affected by the environment (Table 4).

Fresh tuber yield

This was very significant among the clones. From G × E analysis, the CV (31.28%) from combined analysis (Table 3) indicated a wide spread of the difference in the mean of the yield among the genotypes and across the locations. The yield could be considered stable at Onne with a high yield of about 24 t/ha. While Ibadan and Mokwa had yield of 15 t/ha (Table 4).

Dry matter

Dry matter content is a very important trait for acceptability of the cassava by consumers. It was significant in Ibadan at 0.05 (Table 4), with CV of 11.17%. In Onne, it was not significant. However, in combined analysis it was significant at a probability level of 0.001 with a CV of 12.88% and has a mean ranged from 28 to 35 across the locations (Table 5). This shows high dry matter percentage. Edoh et al. (2018) also reported higher dry matter percentage of 33.5% in one of her findings.

Cultivating cassava with yellow-pigmented root flesh is a valid strategy to solve the problem of how to improve the nutritional value of the diet in the regions where cassava is a staple food.

From the analysis of the 25 accessions, a significant (p=0.01)   carotene   content  was  determined  within  the range of variability for carotene concentration in the roots. A few genotypes were high up to 7 when a color chart that ranges from 1 to 8 was used. Olapeju et al. (2013) also reported higher carotene concentration in some of the varieties.

Locations accounted for most of the G × E interaction significance (p=0.001) which reflects the differences in soil types in which the clones were grown. This suggested that for the evaluation of cassava clones, it might be more appropriate to test genotypes over space rather than over time. This experiment was able to identify stable clones across locations such as 01/1413, 01/1442, 01/1663, 98/2132, 01/1277, and 01/1235.

Tuber yield, dry matter content, root size, fiber content, harvest index, sprouting and vigor of the varieties evaluated were all significant among the clones in the combined analysis (Table 5). Taste color of unexpanded leaves, height at branching, leaf area, and internode length were not significant. Overall, dry matter content showed that clone 94/0330 (yellow root) had the highest dry matter (38%), which was better than the best check 30572 (37%), followed by the clones 90/01554, 01/1335, 01/1368 and 01/1371 with values ranged from 30 to 35% (Table 3). However, in Ibadan (Table 4), plant height, vigor, mealiness and root size were not significant. In Mokwa (Table 4), sprouting and plant height were the only traits that were not significant. In Onne (Table 4), dry matter, sprouting, mealiness and taste were not significant.

In terms of disease resistance, all the clones evaluated were resistant to CMD and CBB vector infection and to the spread of the pathogen within the plant and across the locations (Tables 1 and 2).

The results showed that all clones were resistant to CMD, CGM and CAD vector infection and to the spread of the pathogen within the plant and across the locations. In terms of beta-carotene, clones 01/1115, 01/1413, 01/1663, and 01/1335 had high beta-carotene content, indicating the best genotypes for recommendation. In terms of dry matter content, clone 94/0330 (yellow root, 38%) was better than the best check 30572 (37%), clones 90/01554, 01/1335, 01/1368 and 01/1371 had dry matter values ranged from 30 to 35%, which were acceptable values. In terms of garri yield; yield per 50 kg of fresh root submitted for garri showed that clones 01/1649 (25%), 94/0330 (23%), and 90/01554 (23%) were better than the check 30572 (22%). Most of the clones evaluated were sweet, some were bland and non-was bitter in terms of taste.

In terms of yield, the best clones were 01/1368 (26 t/ha), 98/2132 and 01/1663 (25 t/ha). Clones 01/1412, 01/1115, 01/1235, 01/1610, 01/1649, and 95/0379 gave between 21 and 22 t/ha. In terms of the cyanide level, clones  01/1442,  01/1413,  01/1115,  and  01/1663  were very low, clones 01/1224, 01/1235, 01/1371, 95/0379, 98/2132, 94/0006, 01/1662, and 01/1412 were moderate.

None of them was high in cyanide level. Most of the root sizes were large and some were moderate while none were small among the clones evaluated.

In terms of harvest index, clone 01/1115 had the highest index of about 120% of the total yield. Clones 98/2132   (69%),   01/1235   (59%),  01/1412  (58%),  and 95/0379 (55%) were acceptable. Clones 01/1115 and 98/2132 were better than the best check (91/02324) with a harvest index of 60%. Clones 01/1235, 01/1412, 98/2132, 01/1115, 94/0006, and 01/1649 were better than the most popular check 30572 (55%). 

Four of the clones used in this experiment were already released varieties in Nigeria. Three of them (01/1368, 01/1412 and 01/1371) were  released  in  the  year  2011, while one (98/2132) was released in the year 2012.

The authors have not declared any conflict of interests.