Hybrid maize varieties have not yet been adopted by many farmers in Mali. This is partly due to the lack of good hybrid varieties that are tolerant to drought. Recurrent drought is one of the major constraints to maize production in West Africa. It can be devastating if it occurs for a long period and especially during flowering.

The general objective of this study was to increase maize production in Mali to ensure food security and income generation. The specific objectives were (i) to determine farmers’ criteria and preferences for open pollinated varieties (OPV) and drought tolerant hybrid varieties in Mali, (ii) to evaluate maize inbreds and hybrids for their tolerance to drought, (iii) to study the gene effects conditioning performance under both drought and no drought conditions, (iv) to identify hybrids tolerant to drought stress that could be released, (v) to measure the association of inbred root characteristics and tolerance to drought, and (vi) to determine grain yield stability of the hybrids across varying environments.

A Participatory Rural Appraisal (PRA) was conducted at Siramana and Kita, two villages in Mali where maize is the main starchy staple. Siramana geographic coordinates are 11.3° North and 5.7° West. Kita geographic coordinates are 13°02’06’’ North, 9°29’22’’ west. Rainfall in these two sites is around 1000 mm per year. Group and individual discussions with farmers were conducted in these two localities. In addition a questionnaire was administered to maize farmers of these two localities, using a PRA approach.The knowledge on farmers’ criteria and preferences for open pollinated (OPV) and for hybrid varieties could help the research program, to focus on and release acceptable technologies. In addition the number of varieties used by farmers, and the collection of nineteen available varieties in the two locations used for PRA will be useful for conservation and future exploitation. Other related information on the farming systems was also documented. Results indicated that farmers used three main criteria for selecting varieties. These were high yield potential, drought tolerance, and early maturity. Farmers preferred hybrids compared to OPVs, but had concerns about access to hybrid seed and its cost. Farmers used tractors for land preparation for field size greater than 20 hectares. The main consumed crops in these two sites were maize, sorghum and pearl millet. The major source of income was from crop production which accounted for more than 85%. Cotton was the major cash crop. For seed supply thirty nine point five percent (39.5%) of farmers used recycled seed. Only 12.5% of farmers used seed from Research Institute. However there is interaction between farmers and seed companies. Thirty one point five percent (31.5%) of farmers used seed from seed companies

Since drought was a major concern for farmers, two experiments were conducted to evaluate drought tolerant hybrids generated from crosses between thirteen inbred lines. Seventy eight hybrids (from a half diallel cross of 13 inbreds) and two checks were evaluated under water stress and no water stress across four environments during the off season of 2010. The test sites included: Farako (10°50’00’’ North 6°51’0’’ West), Sotuba (12°39’47’’North 7°54’50’’ West) and Yanfolila (11°11’0’’ North 8°9’0’’). Two planting dates were used to create two different environments at Farako. The experiment consisted of a split-plot design with water regime as the main plot and hybrids as the sub plot. Water stress increased anthesis-silking interval resulting in reduced grain yields. Hybrid V841-73/9071 had the lowest grain yield reduction (35.3%) due to water stress, followed by CML 505/1368 (41.9%). In general, hybrids having 9071 as one the parents performed well. This line showed high root weight and root number under stress. The following Genotypes 9071/CML442; CML442/TZCOMP3-C2-S2; 87036/CML442; CML505/1368; V481-73/CML442; C11O-5/9071; J-16-1/TZ COMP3-C2-S2 and CML444/87036 performed well under both well watered and water stress conditions. The inbreds with best positive GCA were 9071 and CML442. These 2 lines also exhibited the highest frequency of appearance as parents among the 20% top hybrid both under well watered and water stress conditions.

During the rainy season of 2011, the same 80 genotypes (hybrids and checks) were tested across nine environments. These included: Sotuba (12°39’47’’North 7°54’50’’ West), Kafara (12° 07’19’’North and 7°48’40’’West), Siramana (11°47’00’’ North, -8°07’00’’ West), Koulikoro (39°12’21’’ North 80°44’99’’), Bema (14°29’40’’ North 9°8’34’’West), Kebila (11°17’0’’ North, 7°2’0’ West), and Tamala (11°25’00’’ North 7°29’00’’ West). Two planting dates were used in Sotuba and Siramana to create additional environments. The results revealed that Sotuba 1, Sotuba 2, Kebila and Kafara represented high yielding environments characterized by better performance and very high yield of the tested genotypes. Siramana 1, Siramana 2, Bema, Tamala, and Koulikoro constituted low yielding environments. Nine hybrids, namely P43SRC9FS100-1-1-8/CML442, C-110/J-16-1, EXP1 24/TZL COMP3-C2-S2, CML444/ TZL COMP3-C2-S2, P43SRC9FS100-1-1-8/V481-73, TZL COMP3-C2-S2/C-110, 87036/V481-73, CML442/TZL COMP3-C2-S2, and CML44/67036 were stable in the high yielding environments. Across the low yielding sites, none of the hybrids exhibited stable performance. Inbred lines V481-73, P43SRC9FS100-1-1-8, 9071, J-16-1, CML444, KU1403x1368, and CML442 showed positive GCA on high yielding environments indicating a predominance of additive gene action. The three best SCAs were obtained with P43SRC9FS100-1-1-8/V481-73, CML442/87036 and (KU1403x1368)/CML505.