search
for
 About Bioline  All Journals  Testimonials  Membership  News


African Crop Science Journal
African Crop Science Society
ISSN: 1021-9730 EISSN: 2072-6589
Vol. 4, Num. 1, 1996, pp. 127-138
African Crop Science Journal
Vol.5. No.2, pp.127-138 1997

Variation for adaptability to dryland conditions in sorghum

G.O. OMANYA, P.O. AYIECHO and J.O. NYABUNDI

Department of Crop Science, University of Nairobi, P.O. Box 30197, Nairobi, Kenya (Received 16 January, 1995; accepted 1 February, 1996)


Code Number: CS96049
Sizes of Files:
    Text:
    Graphics: Tables (gif) - 105.3  

ABSTRACT

Twenty sorghum genotypes [Sorghum bicolor (L.) Moench] were screened for adaptation to dryland conditions in Kenya during the short rains of 1992/93 and the long rains of 1993. The study examined changes in the phenological and physiological characters of these genotypes as a result of drought stress under rainfed field conditions. Parameters measured were days to 50% flowering, plant height, seed yield, seed yield reduction, seed weight, panicle weight, panicles m^2, harvest index, biomass yield, root length density, abaxial stomatal conductance, leaf relative water content, leaf rolling and firing. Significant variation was observed among the tested genotypes. Grain yield under the drought conditions of the second season was positively and significantly correlated with harvest index, panicles m^2, panicle weight per plant, seed weight, root length density, leaf relative water content and stomatal conductance. In contrast, drought susceptibility index and yield reduction were negatively and significantly related to seed yield in the drier second season. The latter season received only 16mm of rain and additional 190mm through supplemental irrigation as compared to 869mm in the wetter first season. The strong correlation of seed yield and these traits suggest that these traits may be advantageous for yield and yield stability under the dryland conditions. The most adapted accessions to dryland conditions were Epson1 029, Epson1 065, E 525HR, Sskn 0003, Kat 369 and Serena. These genotypes achieved high absolute grain production in both seasons and exhibited relatively low yield reduction in comparison to susceptible lines such as Namonimbiri, Sskn 0001 and Local Sorghum Brown. They also exhibited relatively higher values of root length density, leaf relative water content and stomatal conductance under drought. Hence drought resistance was associated with maintenance of high plant water status.

Key Words: Drought stress, correlation, Sorghum bicolor, susceptibility index

RESUME

Vingt genotypes de sorgho [Sorghum bicolor (L.) Moench] ont ete cribles pour lptation aux conditions des terres seches au Kenya durant la courte periode des pluies en 1992/93 et la longue duree de pluies en 1993. L'etude a examine les chagements dans les caractres phenologiques et physiologiques des genotypes soumis au stress de secheresse sous les conditions de champ arroses par les pluies naturelles. Une variation significative a ete observee dans les genotypes testes. Les paramtres mesures etaient: le nombre de jours 50% floraison, la hauteur de la plante, rendement en grains, reduction de rendement en grains, poids de grains, poids panicule par m^2, indice de recolte, rendement en biomasse, la densite de la longueur de racine, la conductance stomatale abaxiale, la teneur relative en eau dans les feuilles, loulement et brunissement des feuilles. Le rendement en grains sous les conditions de secheresse de la deuxine saison etait positivement et significativement en correlation avec lice de recolte, les panicules/m^2, poids de panicules par plante, poids de grains, densite et poids de racines, la teneur relative d dans les feuilles, et la conductance stomatale. Au contraire, lice de susceptibilite la secheresse et la reduction en rendement etaient negativement et significativement en relation avec le rendement en grains dans la plus sche seconde saison. La dernire saison recue que 16 mm d de pluie et 190 mm supplementaires grace ligation supplementaire telle que comparee avec 869 mm recu dans la premire saison humide. La forte correlation du rendement en grains et ces traits suggrent que ces traits pourraient etre avantageux au point de vue de rendement et de stabilite de rendement dans les conditions de terre sche. Les accessions les plus adaptees aux conditions de terre scheetaient Epson 1 029, Epson 1 065, E 525 HR, Sskn 0003, Kat 369 et Serena. Ces genotypes ont realisees de production de grains absolument plus eleve dans les deux saisons et ont montre une reduction de rendement plus basse par rapport aux lignes susceptibles telles que Namonimbiri, Sskn 0001 et le sorgho local brun. Elles ont egalement montre des valeurs elevees en densite de longueur des racines, teneur relative en eau dans les feuilles et en conductance stomatale sous secheresse. Ainsi, la resistance la secheresse etait associee au maintien d'un haut niveau d'eau dans la plante.

Mots Cles: Stress de secheresse, correlation, Sorghum bicolor, indice de susceptibilite

INTRODUCTION

The rains of East Africa in most semi-arid areas are normally short and erratic. Extending crop cultivation into these sub-optimum environments is becoming increasingly important to overcome food deficits. The problem of low yields in drought-prone areas is further aggravated by a rapidly increasing population. In Kenya, there is an accelerated migration of farmers to the drylands due to lack of agricultural land in high potential areas (Mwa and Kanyenji, 1987). This has created the need for the development of suitable crops for the drylands. In such situations the most practical method for the farmer is to adopt drought-resistant cultivars along with conservation and use of soil and water resources to optimise returns. Sorghum [Sorghum bicolor (L.) Moench] has long been recognised as a relatively drought-resistant crop. The gap between genetic yield potential and the realised yield in sorghum is primarily related to environmental stress factors, such as water deficit, which are the major contributors to yield reduction in semi-arid environments. Genetic variability for drought resistance is available in sorghum, with some genotypes being better adapted to dry conditions than others (Gebrekidan, 1987). Sorghum varietal improvement for drought conditions could be achieved more efficiently if traits associated with drought resistance could be identified and utilised as selection criteria.

The objective of the study was to assess the variability in drought resistance among sorghum genotypes and identify suitable morphological and physiological parameters for selecting germplasm adapted to limited moisture conditions.

MATERIALS AND METHODS

Twenty sorghum genotypes (Table 1) acquired from the German-Israel Agricultural Research Agreement (GIARA) Sorghum Improvement Project in the Department of Crop Science, University of Nairobi, were used in this study. The field experiments were conducted at the University of Nairobi Dryland Research Field Station located at Kibwezi, on latitude 2 degrees 17'00''S and longitude 38 degrees 36'36''E in Eastern Kenya. Soil type at this site is a deep eutric Luvisol with good drainage and sandy clay to clay texture. The experiments were conducted in the November 1992 to February 1993 and April to August 1993 seasons. The former was fairly wet with 869 mm of well distributed rain as opposed to the latter which had only 16 mm of rain. It was necessary to apply supplementary irrigation (190 mm) in the latter season, three weeks after sowing, to ensure crop establishment. A randomised complete block design with two replicates was employed. Plots consisted of three rows, 5 m long, spaced at 0.75 m. Plants were thinned to 0.20 m within the rows. The centre row in each plot was used for data collection.

Days to 50% flowering was determined when 50% of the heads in each plot had at least one anther exposed. The extent of leaf rolling and firing were visually scored on a scale of 0 to 5 (0=no rolling/firing, 5 = extreme rolling /firing) at weekly intervals for 6 weeks, beginning 30 days after emergence (DAE). Root length density (RLD) was quantified using the line intercept method suggested by Tennant (1975) and based on the equation:

RLD = (L x 103)/176.625 in cm dm^3,

where 176.625 was the volume of soil sampled using an auger, hammered to pick soil of its volume without compaction.

L = (3.14 ND)/4

where L = length of roots in cm, N = no. of intersections between roots and grid lines, D = grid size, in this case 0.5 cm Stomatal conductance and leaf relative water content were monitored at midday at weekly intervals starting 30 DAE, on the abaxial leaf surfaces of the youngest fully expanded leaves using a steady state porometer (LI-1600). Leaf relative water content (RWC) was determined weekly, beginning 30 DAE, as the water content at sampling time relative to that at full turgor. To obtain this, leaf disc sections from youngest fully expanded leaf were obtained and weighed to determine fresh weight (FW) and then floated in deionized water for 24 hours. The leaf discs were then gently blotted to remove water from leaf surface then weighed to determine their turgid weights (TW). The discs were then oven-dried to determine the dry weight (DW), and leaf relative water content was calculated as; RWC = {(FW - DW)/(TW - DW)} x 100.

Plant height was recorded on ten mature plants in the central row. The average for the ten plants gave the mean for each genotype in each replicate. Ten plants in the middle row of each plot were harvested. Panicles were counted, sun-dried and weighed before and after threshing. Seed yield was then computed. Seed weight was determined from1000 grains for each plot. The remaining sorghum stover was harvested at ground surface level and oven-dried to a constant weight. Total above-ground biomass which included stover and whole panicles was used to obtain biomass yield. Harvest index was also calculated as the proportion of grain in the total biomass.The weight of harvested panicles was divided by ten to get the panicle weight per plant. The number of harvested panicles was expressed as a fraction of the harvested area for each plot. The drought susceptibility index (DSI) based on Fischer and Maurer (1978) was calculated for each field plot as follows:

DSI = [1-Y/Yp]/[1-X/Xp]

Where Y = yield under stress (second season);
Yp = yield under no stress (first season);
X = average yield over all genotypes under stressed conditions (second season);
Xp = average yield over all genotypes under non-stressed conditions (first season).

The reduction in seed yield in the second season was expressed as a percentage of the first season yield.

The mean values of the sorghum genotypes were compared using Least Significant Difference (LSD) values (Steel and Torrie, 1981). Correlation analysis was conducted to identify the significant relationships between variables during the two seasons.

Table 1

RESULTS

Average duration to anthesis were 54 and 57 days after emergence (DAE) during the first and second seasons, respectively. The dates in Tables 2 and 3 show that days to 50% flowering ranged from 35 (in Epson1 029) to 82 (in Namonimbiri) in the first season (November 1992-February 1993) and 43 (in Epson1 015) to 86 DAE (in Namonimbiri) in the second season (April-August 1993).

A comparison between mean seed yield in the first and second seasons revealed that yields in the drier second season were on average 33% less than yields in the wetter first season (Table 3). Mean seed yield among all genotypes were 204 and 137 g m^2 in the first and second seasons, respectively. First season yields ranged from 117 (in Sskn 0001) to 293 g m^2 (in Sskn 0003). In the second season the highest seed yield was recorded by Kat 369, which gave 208 g m^2 while Sskn 0001 produced only 65 g m^2. The low yields in the second season reflected the water deficit conditions that season. However, Epson1 029, Epson1 065, Sskn 0003 and Kat 369 maintained high yields in both seasons. The highest 1000 seed weight was recorded for Epson1 065 which weighed 32.04 and 30.41 g during the first and second seasons, respectively. In contrast, the lightest seeds were obtained from genotype Sskn 0001 which gave mean 1000 seed weights of 14.03 and 10.59 g, during the first and second seasons, respectively. Mean panicle weights per plant were 49.32 and 47.42 g during the first and second seasons. The heaviest panicles in the two seasons were obtained from Epson1 065, Sskn 0003 and Kat 369. The number of panicles per m2 ranged from 12 to 21 in the first season with Sskn 0003, Epson1 129, Epson1 029 and Kat 369 having the highest panicle number. These genotypes also had the highest number of panicles in the second season. Mean harvest indices for the genotypes in both seasons was 0.11 with Epson1 028 and Epson1 029 having the highest indices in the first season and Epson1 029 having the highest index in the second season. Harvest index ranged from 0.06 to 0.19 and 0.05 to 0.19 in the first and second seasons, respectively. The mean biomass production among the genotypes was 2099 g m^2 in the first season and 1333 g m^2 in the second season (Tables 2 and 3). Biomass yields ranged widely from 948 to 2983 and 651 to 1843 g m^2 during the first and second seasons, respectively. Evidently, moisture stress hampered biomass productio during the second season, reducing it by about 36% as compared to the first season. The DSI values in Table 3 suggest that Makueni local, Serena, IS 8193, Sskn 0023, Kat 369, Epson1 029, Epson1 065, E525HR, Sskn 0012, Sskn 0023 and Sskn 0003 were the most tolerant to drought.

Average root length density (RLD) were 47.3, 25.7 and 4.7 cm dm^3 at depths of 15, 30 and 45 cm, respectively, during the first season (Table 4). Root length density in the second planting was higher for all genotypes, averaging 67.9, 29.9 and 13.4 cm dm^3 at 15, 30 and 45 cm depths, respectively. In the first planting the landraces Makueni local, Local Sorghum Brown and Namonimbiri had relatively high RLDs at 45 cm, averaging 10.0, 10.1 and 10.0 cm dm^3, respectively, while the ICRISAT germplasm generally had low RLDs at 45cm. The dwarf Epson1 129 was extreme with no roots at the 45 cm soil depth. Cultivars Epson1 015, Epson1 018, Epson1 028, Epson1 029 and Epson1 065 had on average 0.9, 0.6, 0.7, 0.3 and 0.9 cm dm^3, respectively, of root length density at 45cm. At 30 cm, Sskn 0012 and Namonimbiri had comparatively high mean RLD values of 45.6 and 52.3 cm dm^3, respectively. At 15 cm, these landraces shifted ranks, with Sskn 0012 attaining the highest RLD of 88.1 cm dm^3 followed by Namonimbiri with 68.5 cm dm^3. The lowest RLD at 15 and 30 cm was recorded in Epson1 028. During the second planting, Kat 369 had the highest root length density values of 43.9 and 21.2 cm dm^3 at 30 and 45 cm depths, respectively. On the other hand Epson1 129 had the least mean RLD values of 19.1 and 4.5 cm dm^3 at 30 and 45 cms, respectively. At 15 cm depth Sskn 0003 and Epson1 003 produced the highest and lowest density of roots, respectively. Average leaf relative water content values were 73.4 and 67.8% during the first and second season, respectively. A narrow range between 70.9 and 76.4% was observed for this trait in the first season, while in the second season, it displayed a wider range from 56.4 to 73.9%. Mean stomatal conductance values during the two seasons varied significantly among genotypes with ranges from 0.63 to 0.80 and 0.42 to 0.75 cm in the first and second seasons, respectively (Table 5). On average, variation for stomatal conductance was higher among genotypes during the first season, with the highest value being obtained for Sskn 0023, Kat 369 and Epson1 118. In the second season the highest stomatal conductance values were obtained for Epson1 065, Kat 369 and ICSV 112.

In the first season, most cultivars had low leaf rolling (LRS) and leaf firing scores (LFS), ranging from 0.0 to 2.0 and 0.0 to 1.0, respectively (Table 5). Indeed most genotypes had green leaves at harvest. However, high LRS and LFS were noted, ranging from 1.5 to 4.0 and 1.0 to 3.0, respectively, in the drier second season. The local landraces, for example, Namonimbiri and Local Sorghum Brown had mean LRS and LFS values of 4.0 and 3.0, respectively, in the second season. On the other hand, ICRISAT germplasm, such as Epson1 065, Epson1 029 and Epson1 028, exhibited fairly low LRS values of 1.5 each. Low leaf firing scores were also observed in entries such as Kat 369, Epson1 028, Epson1 118, IS 8193, Serena and ICSV 112 which had mean LFS values of 1.5 or less.

Correlation analysis involving the data obtained during the first season suggest that significant correlations occurred between seed yield and panicle weight (r=0.73**) and that harvest index was strongly associated with the number of panicles per m2 (Table 6). Increased biomass yield was also related to days to flowering, RLDs 15-30 and 45cm. It was also observed that RLD-30 and RLD-45 were positively correlated with days to flowering. However, panicle weight was negatively associated with leaf firing (r=-0.66**).

During the second season, grain yield had significant positive correlations with harvest index, panicles per m2, seed weight, RLD-30, RLD-45, panicle weight, leaf relative water content and stomatal conductance (Table 7). Harvest index was also positively correlated with panicles per m2, seed weight, leaf relative water content and stomatal conductance. On the other hand, seed yield correlated negatively with leaf rolling, leaf firing, drought susceptibility index and seed yield reduction. Strong negative correlations were also observed between harvest index and days to flowering, leaf rolling and firing. The number of panicles per m2 increased with stomatal conductance (r=0.52*) but decreased with leaf rolling (r=-0.61**). Plant height increased significantly with biomass (r=0.71**). Seed weight also correlated positively with RLD-30, RLD-45, panicle weight, leaf relative water content and stomatal conductance, while it was negatively correlated with leaf rolling.

Root length density at 45cm was positively associated with panicle weight (r=0.68**) and stomatal conductance (r=0.56**). Similarly, RLD-30 was positively correlated with panicle weight. Days to flowering was significantly and negatively related to leaf relative water content (r=-0.80**) while it was positively associated with leaf rolling (r=0.65**) and leaf firing (r=0.46*). Panicle weight also correlated positively with leaf relative water content and stomatal conductance and was negatively associated with leaf rolling and firing. As in the previous season, leaf relative water content was positively correlated with stomatal conductance (r=0.56**) while it was negatively related to leaf rolling and firing. Leaf rolling also tended to increase with leaf firing (r=0.62**).

DISCUSSION

The sorghum genotypes included in this study were quite variable in phenology, seed yield and physiological responses to water deficit. Mean seed yield during the first season was high (204 g m^2) and this was probably due to the favourable growing conditions as a result of heavy rain (869 mm). The second season was very dry with 16 mm of rainfall and 190 mm supplied by sprinkler irrigation to enable crop establishment. Therefore, the conditions in the second season were taken to represent severe drought stress, and yield in this season was assumed to reflect the performance of these cultivars under drought stress. The resulting average yield reduction of about 33% in the second season was taken here as reflective of the effect of drought stress on seed yield. On the average, the genotypes which may be considered as the most drought resistant based on seed yield and yield reduction under drought, Kat 369, Sskn 0003, Epson1 065, Epson1 029, E 525HR and Serena, showed a relatively higher plant water status as reflected by leaf relative water content values compared to the most drought susceptible genotypes such as Local Sorghum Brown, Namonimbiri and Sskn 0001. The drought resistant varieties also exhibited higher stomatal conductance, possibly as a result of the better tissue water status. These drought-resistant cultivars had low drought susceptibility indices, also suggesting low sensitivity to dryland conditions. Drought resistance in these genotypes could also have been conditioned by relatively shorter growth duration. Earliness assists plants to escape terminal drought by creating asynchrony between plant development and the occurrence of stress. This phenomenon has also been reported by Rao et al. (1979) in tropical sorghum hybrids and varieties. Earliness is effective when stress becomes important during later stages of growth, enabling plants to escape a late-season stress, as observed in the second planting. The drought stress in the second season hastened panicle exsertion in Epson1 028, Sskn 0001, Ssn 0023, IS 8193, Makueni local and Serena as compared to the first season. Flowering was delayed in all other genotypes except for E 525HR and Kat 369 which took the same number of days to flower in both seasons. Studies by Stout et al. (1978) have also indicated that water stress extended the vegetative growth of some sorghum cultivars while it shortened it in others. The present investigation suggests a relative yield advantage of early flowering genotypes, such as Sskn 0003, Epson1 029, Epson1 065 and E 525HR. Willey and Basiime (1973) have reported that low harvest index and reduced assimilate partitioning to the panicle also appear to characterise the tall and late endemic sorghums of Africa. These deficiencies seem to persist, at least partially, in the landraces Local Sorghum Brown and Namonimbiri, irrespective of the water regime.

The more drought-susceptible genotypes had lighter seeds and lower grain yield. Indeed seed weight was negatively associated with yield reduction in the second season. Thus, it appears that yield reduction was enhanced by a decrease in seed weight. This parameter also showed strong positive correlations with seed yield of the cultivars under the dryland conditions. In terms of both seed yield and seed yield reduction in the second season, genotypes Kat 369, Sskn 0003, Epson1 029, Epson1 065, E 525HR, and Serena appeared to be most adapted to the dryland conditions. Seed yield reduction was negatively and significantly correlated to seed yield during the drier second season. This suggests that high yielding cultivars such as Kat 369, Serena, Sskn 0003, Epson1 029, Epson1 065 and E525HR suffered relatively less yield reduction. These results also support the conclusion of Arnon (1972) that high yielding cultivars under favourable conditions will also yield better under unfavourable environments. On the other hand the greatest seed yield reduction was observed in the lines Epson1 118, Epson1 018, Sskn 0001, Local Sorghum Brown and Namonimbiri which, except for Epson1 018, also exhibited the lowest absolute yields in the two seasons. Epson1 018 had high yield in the wet first season, but was greatly affected by the second season drought conditions, displaying relatively unstable grain production. It was observed that although some lines showed low absolute yields, they also experienced low yield reduction under drought. Such results were evident in Makueni local, Is 8193 and Sskn OO23.

This study revealed negative relationships between leaf relative water content and stomatal conductance with seed yield reduction. The lines Epson1 029, Epson1 065, Kat 369, Sskn 003 and Serena had higher leaf relative water content and stomatal conductance values and suffered less yield reduction. Blum et al. (1989) have also associated drought resistance in sorghum varieties with higher stomatal conductance. High leaf relative water content allows the plant to maintain turgidity and hence high stomatal conductance. The latter would lead to continued photosynthesis under drought. In addition, expansive growth is highly sensitive to plant water status in such a way that decreasing leaf relative content rapidly depresses growth. It is, therefore, expected that genotypes which maintain high leaf relative water would exhibit relatively less reduction in biomass and yield.

During the first season, late flowering landraces such as Namonimbiri, Local Sorghum Brown, Sskn 0012 and Sskn 0001 exhibited relatively higher root length densities. It is believed that genes that control maturity of sorghum also affect growth rate and root length density (Blum et al., 1977; Blum and Arkin; 1985). This might partially explain the high root length density in late flowering during the first season. Root length density was strongly correlated with seed weight, panicle weight and stomatal conductance under water deficit conditions of the second season. These associations further illustrate that under water deficient conditions, root length density provides drought resistant cultivars with superior yield by positively influencing components such as seed weight and panicle weight. The positive association between root length density and stomatal conductance suggests that when cultivars are better able to explore and utilise soil moisture, they will maintain turgor and higher stomatal conductance. This in turn probably enhances photosynthesis, culminating in increased seed yield.

Late flowering genotypes such as Namonimbiri and Local Sorghum Brown had high biomass yields especially in the first season. In addition late maturing genotypes were also relatively taller as compared to early genotypes. However, plant height at maturity was negatively correlated with seed yield during the first season while it was not significantly correlated with seed yield and seed weight in the second season. In this study, high leaf rolling and firing scores were indicative of low plant water status. Previous results obtained in sorghum by Blum (1987) have shown that leaf rolling was delayed in genotypes of greater osmotic adjustment. In this study, less leaf rolling was observed in high yielding genotypes such as Kat 369, Epson1 065, Epson1 029 and ICSV 112. In contrast, landraces Local Sorghum Brown and Namonimbiri exhibited intense leaf rolling and firing. Less leaf rolling and firing together with higher stomatal conductance may have enabled greater carbon dioxide assimilation in the more drought-resistant genotypes. This finding agrees with those of Flower et al.(1990). Thus, these genotypes were probably able to photosynthesise over a longer time, hence the yield advantage. Leaf rolling is induced by loss of turgor (Blum, 1987). This was confirmed by the negative relationship between leaf rolling and leaf relative water content. Genotypes such as Local Sorghum Brown, Namonimbiri and Sskn 0001 had higher leaf rolling and firing scores and low leaf relative water content in contrast to Kat 369, Epson1 029, Epson1 065 and ICSV 112. Late genotypes, including the landraces Local Sorghum Brown, Namonimbiri and Sskn 0001, had high firing scores, and also displayed low stomatal conductance values. This may have resulted in insufficient transpirational cooling and the subsequent heating of all or part of the leaf. Thus, leaf firing may serve as an additional criteria, together with other traits. Peacock et al. (1989) also supported the use of leaf firing in identifying drought-resistant cultivars.

The improvement of yield performance under drought stress conditions requires the combination of high yield potential with both morphological and physiological attributes of drought resistance. The results of this study suggest that sorghum genotypes differ markedly in drought response. Important strong positive correlations were revealed between grain yield and traits such as harvest index, panicles per m2, panicle weight per plant, seed weight, root length density, leaf relative water content and abaxial stomatal conductance. Drought susceptibility index and seed yield reduction were negatively and significantly related to yield. The strong correlations of seed yield with these parameters suggest their importance in water deficit situations and, therefore, they may assist in selection for drought resistance in sorghum. However, the lines used in the current study were not consistently classified by the individual variables measured. This suggests that assessment of genotypic drought response should be based on observations of a combination of both morphological and physiological characteristics in a multiple selection criteria. Among the germplasm studied, genotypes Epson1 029, Epson1 065, E 525HR, Sskn 0003, Kat 369, and Serena were best adapted to dryland conditions, exhibiting superior absolute grain yields in both seasons and low yield reductions under the severe drought conditions in the second season.

ACKNOWLEDGEMENTS

The authors gratefully thank the German-Israel Agricultural Research Agreement (GIARA) Programme and the University of Nairobi for supporting this study.

REFERENCES

Arnon, I. 1972. Crop Production in Dry Regions. Vol. II. Leonard Hill Books, London. pp.92-135.

Blum, A. 1987. Genetic and environmental considerations in the improvement of drought stress avoidance in sorghum. In: Food Grain Production in Semi-arid Africa. Menyonga, J.M., Bezuneh, T. and Youdeowei, A. (Eds.), pp.91-99. OAU/STRC-SAFGRAD.

Blum, A., Arkin, G.F. and Jordan, W.R. 1977. Sorghum root morphogenesis and growth. I. Effect of maturity genes. Crop Science 17: 149-150.

Blum, A. and Arkin, G.F. 1985. Sorghum root growth and water-use as affected by water supply and growth duration. Field Crops Research 9:131-142.

Blum, A., Mayer, J. and Golan, G. 1989. Agronomic and physiological assessment of genotypic variation for drought resistance in sorghum. Australian Journal of Agricultural Research 40:49-61.

Fischer, R.A. and Maurer, R. 1978. Drought resistance in spring wheat cultivars. I. Grain yield responses. Australian Journal of Agricultural Research 29:897-907.

Flower, D.J., Rani, A.V. and Peacock, J.M. 1990. Influence of osmotic adjustment on the growth, stomatal conductance and light interception of contrasting sorghum lines in a harsh environment. Australian Journal of Plant Physiology 17:91-105.

Gebrekidan, B. 1987. Sorghum improvement and production in Eastern Africa. In: Food Grain Production in Semi-arid Africa. Menyonga, J.M., Bezuneh, T. and Youdeowei, A. (Eds),pp.141-154. OAU/STRC-SAFGRAD.

M'ragwa, L.P. and Kanyenji, B.M. 1987. Strategies for the improvement of sorghum and millet in semi-arid Kenya. In: Proceedings of EARSAM 6th Regional Workshop on Sorghum and Millet Improvement, 20-27th July 1988. pp 173-189.

Peacock, J.M., Azam-Ali, Publisher, S.N. and Matthews, R.B. 1985. An approach to screening for resistance to water and heat stress in sorghum [Sorghum bicolor (L.) Moench]. In: Arid Lands, Today and Tomorrow, Whitehead, E.E, Hutchinson, C.F., Timmermann, B.N. and Varady, R.G. (Eds.). Office of Arid Lands Studies, University of Arizona, Tucson, Arizona.

Rao, N.G.P., Rao, J.V., Rana, B.S. and Rao, V.J.M. 1979. Responses to water availability and modifications for water use efficiency in tropical dryland sorghums. In: Proceedings of Symposium on Plant Responses to Water Availability, 22-24 Feb. 1979, IARI, New Dehli, India.

Sanchez-Diaz, M.F. and Kramer, P. J. 1971. Behaviour of corn and sorghum under water stress and during recovery. Plant Physiology 48:613-616.

Steel, R.G.D. and Torrie, J.H. 1981. Principles and Procedures of Statistics. A Biometric Approach. McGraw-Hill Inc.

Stout D.G., Kannangara, T. and Simpson, G.M. 1978. Drought resistance of Sorghum bicolor. 2. Water stress effects on growth. Canadian Journal of Plant Science 58:225.

Sullivan, C.Y. and Blum, A. 1970. Drought and heat resistance of sorghum and corn. In: Proceedings of the 25th Annual Corn and Sorghum Research Conference of the American Seed Trade Association. Kansas, U.S.A, pp. 55-66.

Tennant, D. 1975. A test of a modified line intersect method of estimating root length. Journal of Ecology 63:995-1001.

Willey, R.W. and Basiime, D.R. 1973. Studies on the physiological determinants of grain yield in five varieties of sorghum. Journal of Agriculture, Cambridge 81:537-548.

Copyright 1996 The African Crop Science Society


The following images related to this document are available:

Line drawing images

[cs96049e.gif] [cs96049b.gif] [cs96049c.gif] [cs96049f.gif] [cs96049g.gif] [cs96049d.gif] [cs96049a.gif]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2024, Site last up-dated on 01-Sep-2022.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Google Cloud Platform, GCP, Brazil