Unfolding History of Local Guinea Sorghum Enhanced for Food Quality through Potassium Hydroxide and Bleach Test: Farmer Alternative to Climate Variability Resilience

Sorghum (Sorghum bicolor (L). Moench) is an important staple food crop produced essentially under rained conditions in Mali. To preserve and exploit sorghum genetic diversity, germplasm collection was undertaken in 1978 across Mali. During this germplasm evaluation in 1982, a particular guinea sorghum cultivar from Chegue village in north western Mali, recorded as CSM63, was identified for its earliness and photoperiod insensitivity. This cultivar was unfortunately segregating not only for a testa presence, which is detrimental to To color and nutritional quality, but also for flowering time and plant height. Field screening coupled with the potassium hydroxide and bleach (KOH: B) tests were performed to eliminate testa while maintaining CSM63 agronomic values. Overall objectives were to trace history of CSM63 native population, screen for testa absence and define its favorable cropping zone. Field experiments were conducted at the Cinzana Agronomic Research Station (SRAC) while laboratory tests for testa absence and cooking quality were conducted at Sotuba Food Technology Laboratory (LTA) in Mali. Two thousand (2000) plants from original CSM63 population were grown in a complete isolation from other sorghum varieties to select 288 panicles for testa absence. For yield trial, a completely randomized block design was used with 3 replications. After series of selfing, a pure new line (CSM63-18-1AF) was identified without a testa with similar maturity, grain yield, better To color and nutritional values than the original CSM63. The new line was named CSM63E and its cropping zones were delimited beyond its center of origin. CSM63E was released for large scale production by Institut d'Economie Rurale (IER) and registered in National and West African Countries Regional Sorghum Catalogs. The adopted line is named Jakunbe (prevent drought) in Bambara language) by farmers. Jakunbe is the number one certified sorghum seed produced in Mali in 2016 in term of quantity. Cropping both Jakunbe and late maturing varieties beyond isohyets 700 mm exposes farmers to yield losses caused by non-proper environment.


INTRODUCTION
Sorghum (Sorghum bicolor (L.) Moench) is a main staple cereal crop in Mali along with millet, rice and maize. Sorghum is cultivated in rain-fed conditions, or in river receding conditions in the northern regions. In 2016, Malian sorghum production accounted for 1 637 912 t ranking fourth in term of production behind rice (Oriza sativa) (2 811 049 t), maize (Zea mays) (2 533 999 t) and millet (Pennisetum glaucum) (1 927 106 t). Sorghum stayed in the same position in 2015 with 1 527 456 t, following the same ranking [1]. Among the five cultivated races of sorghum, the guinea race composed of the guineense and margaritiferum working groups, accounts for more than 70% of sorghum cultivated in West and Central Africa and may account for more than 50% of the all sorghum produced in Africa [2]. Guinea race dominates southern, central and Northern Mali zones in areas and production. There is a large maturity variability within this race following south-north rainfall gradient.
Most sorghums grown by farmers are local or improved guinea cultivars very well adapted to the local environments rather than introductions or improved sorghums without guinea background. Consumers prefer the cooking traits of guinea sorghums and their derivatives excluding those with a testa. Testa occurrence is under genetic control and poor seed renewal from farmers.
Testa is maternally inherited. The genetic constitution of the testa, aleurone and embryo are: testa is a diploid and derives entirely from maternal tissue. Aleurone is triploid with one sperm of the male combining with the two polar bodies of the female, and the embryo is a diploid, receiving one genome from each parent. After the double fertilization (formation of the embryo (2n) and endosperm (3n), unfertilized diploid cells from the ovule divide mitotically to form part of the seed coat (testa) which is not affected by paternal genes [3,4,5].
Genetically, two dominant genes (B 1_ B 2_ ) control the presence or absence of testa. If either B 1 or B 2 are homozygous recessive, the testa cells either disintegrate and are absorbed or do not develop. When the genes controlling the testa, B 1 and B 2, and the spreader, S, are dominant, tannins from the testa will diffuse into the pericarp. When S is recessive with dominant B 1 (most guinea types) and B 2 (USA germplasm types ), tannins reside only in the testa layer [6,7].
When B 1_ B 2 are dominant, a pigmented testa is present. When either or both genes are homozygous recessive (B 1_ B 2, b 1 b 1 B 2, or b 1 b 1 b 2 b 2 ) pigmentation is absent. The color of the pigmented testa is controlled by another gene (Tp) in which brown is dominant to purple. The spreader gene (S) allows the brown color of a pigmented testa to be present in the epicarp (S-). Authors [7] argued that several genes are involved in pigmentation control and pericarp thickness: When mesocarp is thin, Z gene is dominant (Z-) and thick when the gene is recessive (zz). R, Y, I, B 1_ B 2, and S genes are associated with the presence of phenolic compounds that could be antifungal, conferring grain mold resistance, while thin mesocarp, determined by the amount of starch granules present, is also thought to confer grain mold resistance. These mold resistance phenolic compounds are present in some guinea sorghum varieties, thus conferring them poor purple Tô color not well culturally appreciated by consumers [7,8].
Pericarp color (white, red, yellow) depends on thickness of the seed coat and presence or absence of the testa. Three genes (R_Y_Z_) control the color and intensity of the pericarp and variation in the appearance of the testa [6]. Authors [7] claimed that seven loci are known to be responsible for the different characteristics affecting caryopsis traits (grain): R, Y, I, Z, B1, B2, and S genes. The R and Y genes determine pericarp color. If both genes are dominant (R-Y), then the pericarp is red. When the Y gene in homozygous recessive (R-yy or rryy), the pericarp is colorless or white regardless of the R gene. A lemon-yellow pericarp is found when the R gene is homozygous recessive and the Y gene is dominant (rrY-). The intensifier gene (I) modifies the color of the pericarp to appear bright when dominant (I-) and dull when recessive (ii).
Testa [9] affects adversely nutritional quality of sorghum grain. Studies in rats, chicks and livestock have shown that high tannin in the diet adversely affects digestibility of protein and carbohydrates and reduces growth, feeding efficiency, metabolizable energy and bioavailability of amino acids. There is no direct evidence regarding anti-nutritional effects of dietary tannins in human subjects, although high dietary tannin may have some carcinogenic effect. Iron absorption in Indian women was lower when they were fed porridge prepared from bird-resistant high-tannin sorghum in place of porridge prepared from tannin-negative sorghum [10]. In a study elucidating the relationship of the testa to agronomic and nutritional traits in sorghum, [11] reported that among the nutritional and mineral traits the testa types had more protein, lysine percent sample, potassium, and tannin; and less lysine percent protein, oil, carbohydrate, phosphorus, gross energy, enzyme (α-amylase) activity, in vitro dry matter digestibility, and metabolizable energy than the non-testa types. However, the results were highly influenced by environment. They concluded that sorghums with testa were inferior nutritionally to non-testa ones and tended to be inferior agronomically. The only advantage of the testa was higher protein and lysine percent sample [9].
Tannin from sorghum was advocated as having strong antioxidant activity [12,13]. These authors [12,13] found that tannin sorghums had antioxidant activities higher than most non-tannin sorghums. Meantime [12,14] argued that high molecular weight (MW) tannins have the greatest antioxidant activity in vitro among natural antioxidants. This was attributed to the proximity of many aromatic rings and hydroxyl groups and the fact that tannins were not able to act as prooxidants [12]. A major concern about tannins though is that they may not be biologically effective antioxidants due to their large molecular size and their tendency to bind food molecules into insoluble complexes. However, [13] demonstrated that even when complexed with proteins, sorghum tannins retained at least 50% of their antioxidant activity. Such proteincomplexed tannins may serve as free radical sinks in the digestive system thus sparing other antioxidants.
Traditional methods to inactivate and detoxify tannins include moisturizing the grain with alkali several hours prior to utilization, formaldehyde, polyethylene and glycol and gelatin treatments [10]. Another method of determining condensed tannins in sorghum is the vanillin-HCl method when the blanks are subtracted to eliminate background non-tannin materials. However, it requires significant time and is not readily applied in routine grading of sorghum [15].
Grain hardness is the most important and consistent grain characteristic that affects Tô quality [16]. Sorghums with corneous (hard) endosperm generally produce good quality Tô whereas very hard like "kende" margariteferum type and very soft endosperm like "Gadiaba" durra type sorghums produce poor quality Tô [17]. Authors [17] claimed that Tô of poor quality resulted from fine flour particles (<250 µm) while medium size flour particles (<425 µm) produced intermediate sorghum Tô quality. Guinea sorghums, including CSM63, have hard and corneous grain predisposing them to good Tô quality.
CSM63 was identified for its early maturing among guinea race cultivars in the 1984 preliminary agronomic trait evaluation of the 1978 Malian sorghum germplasm collection [18]. The collections were aimed at preserving Malian germplams from genetic erosion and identifying racial diversity and maturing group of cultivars in the Sahel region to secure production due to recurrent droughts. Despite its earliness, CSM63 showed limited food and market values due to its testa and high tannin content by consumers. Different types of foods, including Tô and couscous from guinea sorghums, constitute the major daily food sources consumed by most rural and low income urban Malians. However, in Tominian, Koutiala and Bandiagara districts of Mali, tannin content sorghum grains are well preferred by women for local beer.
The objectives of this study were to: (1) screen CSM63 for the absence of testa, (2) identify nontesta CSM63 derivatives for agronomic and cooking qualities and (3) define its optimal cropping area.
Thus, the need for improving the cooking quality of CSM63 by eliminating the testa while maintaining its original agronomic values was imperative for Tô, Malian traditional food. Pure line selection (PLS) (single plant selection from a segregating landrace variety population with progeny testing) in addition to the potassium hydroxide (KOH) and bleach (B) test were used to improve CSM63.

Plant Material
A team composed of IER and CIRAD scientists collected CSM63 (PI8939) in 1978 [18]. CSM63 local name is Keninke from Chegue village in Nara District (37 km from Mauritania). Chegue is located in longitude 06°.57' West and latitude 15°.11' North. Farmers indicated that CSM63 matures in 90 days. The original CSM63 segregates for plant height, maturity, pericarp thickness, glumes color and testa.
Plant material was composed of original CSM63 population, selected derived pure lines from original CSM63, CSM219E, CSM417E and Framida as checks.

Field screening
Approximately 1/14 of a hectare was planted in 1986-1987 cropping season with the CSM63 original population in a complete isolation from other sorghum fields at the Cinzana Agronomic Research Station (SRAC) in well 10 (F10). The final population was 2000 plants. Two hundred and eighty-eight (288) selfed panicles were retained for testa absence at maturity. Seed from 288 panicles were planted along with bulk seeds of the original population at the SRAC to determine flowering and testa absence in 1987-1988 cropping season. Ten panicles in each progeny row, except those from the original population, were selfed per plot before flowering. Forty-five lines (45) from 15 families were visually selected for testa absence and planted in 1989-1990 season at SRAC for flowering, testa absence, plant height, maturity and flour quality. A field test for testa presence was performed by scraping 10-20 seeds per panicle with pocket knife and by flour appreciation before the KOH: B test on 45 families.

Agronomic yield test
For testa screening experiment, one row observation plot of 5 m was used per entry from 1987 to 1989.
For yield test of 1990-1991, randomized complete block design (RCBD) was used with three replications. Experimental unit was three rows per entry. Spacing between rows was 0.75 m and 0.50 m between hills on the row for preliminary and yield tests. Row length was 5 m long. Net plot length was 4.5 m (9 hills /row) with an expected 27 hills at harvest by plot for grain yield. Planting was done on 19/07/1991. Thinning was performed at two plants per hill with a population density of 53 333 plants per hectare. Experiments were under rain-fed conditions. Hundred (100 kg/ha) of diammonium phosphate (DAP = 18N-46P 205 ) were applied two weeks at thinning time and 50 kg/ha of urea (46N) were applied as side dressing 40 days after emergence. Furrows were dug alongside each ridge close to seedlings to bury fertilizers. The first weeding was done on 2 July during thinning time; the second was on 30 July and the third was on 15 September during ridging time. Harvest was achieved after physiological maturity when grains were completely dried. Panicles from each plot was dried and weighted twice to confirm weight stability before threshing. Threshing was done by women using mortar and pestle.

Selected agronomic data
Data on number of days to 50% flowering (FD), plant height (PH) in cm, number of hills harvested (HH), number of panicles harvested (PAH), percent threshing (%PT), panicle weight (PWT) and grain yield GY (kg/ha) were collected.

Bleach, potassium hydroxide (KOH: B) and tô cooking tests
The bleach test method [15,19]  For the Tô cooking quality, ten (10 gr) grams of potassium hydroxide were dissolved in 2.5 liters of water to create an alkali solution in 1989. Tô was then cooked in 20 g of flour and 80 ml of alkali solution. The pH of the solution was 11 and that of the Tô was 9 on average. Keeping (consistency) quality of Tô was measured 24 hours after cooking.

Laboratory data collected
For cooking properties, testa presence, percentage of decortication (PD), consistency of Tô in millimeter (TCmm) and manual (touching) (MTC) after 24 hours for keeping quality (overnight keeping) and Tô color (TC) were recorded.
Tô consistency after overnight keeping was measured in millimeter (mm) (using penetrometer and manually (Tô consistency commonly used by households) was scored using a scale of 1(very good), 2 (good), 3 (soft), 4 (clammy, sticky) and 5 (very sticky). Tô color (TC) was scored on a scale of 1 (cream or white pale, good Tô color), 2 (gray to slightly maroon), 3 (maroon to dark gray), 4 (brown or light dark) and 5 (dark or dark brown, very bad TC).
Samples with a score of 3 and above are classified as unacceptable for TC but the grains can be acceptable for couscous, another Malian food, according to the Sotuba Food Technology Laboratory (LTA) standards.

Data analysis
Descriptive statistics and analyses of variance for parametric and none parametric data using MINITAB 14 software were performed. Percent decortication was not transformed because all samples had the same denominator values. Tô manual consistency, color, and conservation were analyzed using Kruskal-Wallis, χ 2 type test for means rank separation. All significances were determined at the probability threshold of P = 0.05.

Environmental Conditions at the Experimental Site
The Cinzana Agricultural Research Station (SRAC) is five km from the Cinzana village. SRAC is located in the Sahelian zone of Mali ( Fig. 3) with geographic coordinates of 13°15' N; 5°58 W and 265 m of altitude. SRAC is 45 km from its regional capital, Ségou. It has monomodal rainfall pattern with distinct rainy season in summer accounting on average for 690 mm. Mean annual maximal temperature is 35.6°C and minimal is 19.3°C (Fig. 1) with a maximum day length of 12h47. Maximum temperatures are recorded from April to May and the minimum from December to February. The climate is characterized by an inter-annual variability of rainfall, turbulent high intensity, and irregular intra-season rains. The soil is classified as a leached tropical ferruginous soil with spots and concretions for the Station and loamy-sand type with pH varying from 5.5 to 6 (slightly acid) at the experiment site.

Rainfall (mm) Temp. (°C)
Monthly average rainf all Minimum temp. Maximum temp. Eight lines were retained (Table 1) after the KOH:B test. There was no difference among the 9 lines for grain yield, while there were highly significant differences among lines for FD, PH and % decortication. Selected agronomic and cooking qualities are shown (Tables 1, 3, 4).

Laboratory Data
The average percentage of testa of the 45 random individuals was 7.8%, while that of the original population was 76% after KOH: B test (Table 4). Three lines showed no testa presence (CSM63E-13-1AF, CSM63E-13-2AFand SM63E-18-1AF). Fig. 2 shows some false testa presence in CSM 219E caused by insects bites and weathering.
For the 9 lines, there were differences among treatments due to decortication (Table 1). Tô color and consistency were the same for non testa samples. Kruskal-Wallis rank test (

CSM63E Adaptation Zone
Currently CSM63E has been intensively grown by framers in Mali from 400 to 900 mm rainfall zones. Delayed planting is applied in late July for maturity to occur end of September in rainier areas (900 mm). Each gradient planting period was applied to delineate CSM63E optimal yield potential from 700 mm (Kondogola village), 900 mm rainfall (Kafara village) and 800 mm (Finkoloni village) (Fig. 3). Study on the impact of the Cinzana Agronomic Research Station indicated that CSM63E is the most cultivated improved sorghum at farm level in Ségou (81%) (600-700 mm) and Mopti (99%) (400-500 mm), regions. Large scale adoption of CSM63E was subsequent to its early maturing and grain yield (229-371 kg/ha) over local checks by farmers [20]. Malian Seed Certification Central Laboratory report of 2016-2017 cropping season data showed that CSM63E is the most produced certified sorghum seed among 39 varieties. Its certified see produced in 2016-2017 cropping season was 365.750 t or 62% of the total certified sorghum seeds (589.953 t) [21].

DISCUSSION
Elimination of testa from the original population (CSM63) leading to CSM63E has resulted in a rapid acceptance of the improved cultivar by farmers and housewives subsequent to its improved nutritional values. Tô with dark or red color has no visual attraction and labeled for negative quality for human consumption. By analogy to [9] findings, foods processed from tannin cereals have poor nutritional values. Tannins in the grain impart an astringent taste which affects palatability, reducing food intake and consequently body growth. Tannins bind to both exogenous and endogenous proteins including enzymes of the digestive tract, affecting the utilization and digestibility of protein and carbohydrates, feeding efficiency, metabolizable energy and bioavailability of amino acids [9]. There have been no Tô nutrients composition analysis from this study, but based on findings [9] on non testa sorghum poor food nutrients availability, inference can be made that nutritional values of CSM63 was improved.
False testa presence observed in this present study was reported in previous laboratory tests results [22]  Maturity alleles presence are very important when making decisions on which lines to choose where spatial rain distribution is a major limiting factor for good harvest and food security. CMS63E registered and released was not statistically earlier than the original population. Thus agronomic integrity of the original population in term of maturity was preserved.
Good Tô color and nutritional values of CSM63 derived line have been achieved with potential added market value. Author [23] reported that the textural quality of traditional sorghum porridges determines their acceptability to consumers, while [16] claimed that sorghum Tô that is too soft and sticky cannot be molded between the fingers, sticks to the teeth, sticks to the palate during consumption, and sticks to the cooking utensils. Consumers prefer sorghum cultivars that consistently produce relatively firm and nonsticky Tô. Data from cooking tests of CSM63E correspond to good Tô criteria [16]. A good Tô has the ability to be a good dinner or breakfast. It is important to notice that original CSM63 had good Tô keeping quality opposed to its poor nutritional value and food attractive appreciation.
Adaptation zone of CSM63E was defined based on different planting periods from 700 mm to 900 mm gradients. As this variety originates in Nara, where rain fall is less than 500 mm, it has proven to be very productive in the 600-700 mm zones where it is very well sold. Beyond this area (600-700mm), CSM63E has shown low yield potential in July planting time in 800-900 mm of Koutiala and Kafara. Main raisons for its low yield due to late planting are increasing pest damages (crop establishment from damages on seedlings), soil nitrogen depletion after rain onset, seedlings logging. For normal planting period in its area of adaptation, there are no major constraints. While in rainier area (>700 mm) with early planting, this variety does poorly due to pollen wash from August down pour. Grains are attacked by mold and insect complexes, with anthocyanin leaking into the endosperm, leading to no yield or very poor grain quality. The second most important negative impact is the spread of midges in later varieties. [24] argued that early maturing varieties are not advised where midges (Stenodiplosis sorghicola) attacks on sorghum are serious issue. Authors [24] argued that midge spread occurs from early maturing varieties to late ones, thus resulting in major yield loss for farmers. Beyond 700 mm rainfall, cropping CSM63E is too risky for farmer. CSM63E, based on its certified seed volume, is well accepted by producers, thus implicitly by consumers. This acceptance is due to its early maturing and food quality, thus alleviating climate variability burden of the very volatile Sahelian environment of Mali.

CONCLUSION
Combined field screening and laboratory tests were efficient to eliminate testa in a sorghum cultivar in Mali. A pure line, CSM63E-18-1AF (CSM63E), was identified and released. The derived line and its original population are within same early maturity group. Derived line expresses better Tô color and nutritional values and is cropped beyond its center of origin with 400 mm of rainfall to 700 mm. This pure line has become an alternative solution for farmer to overcome drought and be more resilient. CSM63E is the number one certified sorghum seed produced in Mali in 2016. Cropping zone of CSM63E was defined, beyond which farmers are exposed to poor harvest, grain quality and specially damage due to midge for late maturing sorghum varieties. This variety is grown outside Mali in West Africa Sahelian zone for its earliness and improved processed food qualities, thus overcoming hunger. Strong contribution from extension agents and NGOs are recommended to help farmers to maintain their seed genetic purity. Good agronomic practices and value added products coming from the improved cultivar can contribute to poverty alleviation in rural zones in West Africa.