Performance of durum wheat landraces in a Mediterranean environment (eastern Sicily)

Summary The evaluation aimed at identifying landrace genotypes adapted to the rather unfavourable growing conditions of durum wheat in Sicily, to be used as parental material in a breeding programme. The trial was carried out in three seasons experiencing varying climatic conditions, and included 75 landraces, 25 of which were selected under severe drought in Syria. Wide differences were observed for most traits among genotypes and seasons of evaluation. Yields of the best performing entries identified in each season never significantly differed from that of the best check variety. The top-yielding landraces were consistently better than the remaining entries for the three yield components, viz. number of spikes per plant, number of kernels per spike and mean kernel weight. In the driest season they were also significantly earlier in heading, confirming the importance of earliness under drought. An analytical breeding approach relying on an array of morpho-physiological traits as selection criteria did not seem appropriate for the given environment, as the variable stress level enhanced the importance of specific traits under specific situations. The genotypic response was largely season-specific. Nonetheless, five landraces were in the best group in all seasons. The selection made in Syria also appeared somewhat effective in Sicily, particularly in the less favourable seasons.     

Variation in utilization efficiency and tolerance to reduced water and nitrogen supply among wild and cultivated barleys

Broad genotypic variation in the response to low soilmoisture and reduced nitrogen supply was found amongthe wild Hordeum spontaneum accessions and thelandraces and modern cultivars of H. vulgare ofdifferent geographic origin. Measurements at the endof vegetative growth in plants grown in soil culturesrevealed genotypically specific responses to the usedenvironmental factors. Cultivars and breeding linesfrom Syria and Ethiopian landraces combined bothdrought resistance and tolerance to low nitrogen. TheSyrian barleys were also distinguished by a highnitrogen utilization efficiency (NUE) under low Nnutrition. European cultivars indicated a pooradaptation to N shortage, but some of them wereresistant to soil drought. No stress resistant barleyswere found among the wild accessions and Sardinianlandraces. Genotypic differences in the relativevalues of NUE and water use efficiency were associatedwith low N-tolerance. Some Syrian selections,Ethiopian landraces and the modern German cv. Maresiwere found to be most drought resistant. Maintenanceof a relatively high photosynthetic activity of theuppermost leaves was associated with droughtresistance. As far as concerning with the vegetativegrowth phase, the modern Syrian germplasm andEthiopian landraces may be recommended as donors ofadaptative characters for local barley breeding.     

The relationship between the regeneration system and genetic variability in the cucumber (Cucumis sativus L.)

Durum wheat has been subjected to intense breeding in Italy due to its local economic importance. Four groups of Italian cultivars representative of different breeding eras were compared in northern Syria for yield potential and morphophysiological features at a moderately favourable site, and drought tolerance at a stressful site. Group 1 included indigenous landraces; Group 2 comprised genotypes selected from exotic landraces (released in 1920's–1930's); Group 3 included genotypes selected from crosses or mutagenesis involving Group 2 materials (1950's–1960's); Group 4 comprised genotypes selected from crosses between CIMMYT and Group 2 materials (from 1970's). Under moderately favourable conditions, a yield increase of 1.03 t ha–¹ was observed from Group 1 through Group 4, corresponding to a genetic gain of about 0.017 t ha–¹ per year. Such increase was only partly accounted for by a parallel increase in spike fertility and seed weight. Plant stature decreased dramatically from Group 1 to Group 4; a remarkable reduction of height was already attained in Group 3, before the introduction of dwarfing genes from bread wheat. Earliness of heading and grain filling duration increased consistently across breeding phases, the length of crop cycle remaining almost unaltered. No significant improvement of drought tolerance resulted from the breeding activity, suggesting the need to put more emphasis on selection for real stress tolerance rather than escape. Overall variation for morphophysiological traits, assessed by a principal components analysis, highlighted the great diversity among the Group 1 cultivars, while variation within Groups 3 and 4 was lower. Indigenous landraces, little used in the breeding history, as well as novel, unexploited exotic germplasm sources could contribute to broaden the crop genetic base in the region.     

The drought response of landraces of wheat from the northern Negev Desert in Israel

Summary Diverse landraces of wheat, collected from the semi-arid (150 to 250 mm of total annual rainfall) Northern Negev desert in Israel were considered as a potential genetic resource of drought resistance for wheat breeding. These materials were therefore evaluated for their reponses to drought stress in agronomical and physiological terms. Up to 68 landraces, comprising of Triticum durum, T. aestivum, and T. compactum were tested in two field drought environments, in one favourable field environment, under post-anthesis chemical plant desiccation which revealed the capacity for grain filling from mobilized stem reserves, under a controlled drought stress in a rainout shelter and in the growth chamber under polyethylene glycol (PEG)-induced water stress. Biomass, grain yield and its components, harvest index, plant phenology, canopy temperatures, kernel weight loss by chemical plant desiccation, growth reduction by PEG-induced drought stress and osmotic adjustment were evaluated in the various experiments.Landraces varied significantly for all parameters of drought response as measured in the different experiments, which was in accordance to their documented large morphological diversity. Variation in grain yield among landraces under an increasing drought stress after tillering was largely affected by spike number per unit area. Kernel weight contributed very little to yield variation among landraces under stress, probably because these tall (average of 131 cm) landraces generally excelled in their capacity to support kernel growth by stem reserve mobilization under stress. Yield under stress was reduced with a longer growth duration of landraces only under early planting but not under late planting. Landraces were generally late flowering but they were still considered well adapted phenologically to their native region where they were always planted late.Landraces differed significantly in canopy temperature under drought stress. Canopy temperature under stress in the rainout shelter was negatively correlated across landraces with grain yield (r=0.67^**) and biomass (r=0.64^**) under stress. Canopy temperature under stress in the rainout shelter was also positively correlated across landraces (r=0.50^**) with canopy temperature in one stress field environment. Osmotic adjustment in PEG-stressed plants was negatively correlated (r=–0.60^**) with percent growth reduction by PEG-induced water stress. It was not correlated with yield under stress in any of the experiments. In terms of yield under stress, canopy temperatures and stem reserve utilization for grain filling, the most drought resistant landrace was the Juljuli population of T.durum.     

Rhodes grass breeding in Kenya I. Intra-variety variation and character relationships

Summary Spaced swards of 3000 plants and a selected subpopulation of 750 plants of 3 commericial Rhodes varieties, Mbarara, Masaba and Pokot, were studied between 1969 and 1975. Intra-variety variation was large in heading date, vigour of growth, seed (PGS) yield, persistence, drought tolerance and in vitro digestibility.There were on average 6 weeks between the heading date of the first and the last 5% of the plants within a variety. Variation in heading date was largely accounted for by variation in vigour and vice versa. Vigorous plants tended to be early heading and only few plants were found to possess a combination of high vigour and late heading.Among plants equally selected for high vigour from within a variety, early-heading plants were higher in PGS yield, persistence, drought tolerance and digestibility.     

Geographical patterns of phenotypic diversity in <Emphasis Type="Italic">Oryza sativa</Emphasis> landraces of Côte d’Ivoire

The knowledge of agro-morphological diversity and the distribution pattern of variation among conserved accessions could be an invaluable aid in germplasm management and crop improvement strategies. In this study, the geographical pattern of morphological variation of 880 landrace rice (Oryza sativa (L.)) accessions in Côte d’Ivoire was evaluated for 13 agro-morphological characters. Shannon –weaver diversity index indicated an overall mean of 0.47 ± 0.07 in the collection, with the greatest diversity from derived Savanna and north-west (H′ = 0.52 and 0.50) while the west-central had the lowest diversity (H′ = 0.41). Canonical discriminant analysis showed that traits such as panicle length, grain size (weight and length), tillering ability, number of days to heading and maturity were the main discriminatory characteristics. Result of the phenotypic frequency shows that, the landraces from the north and north-western zones were mostly tall, early heading and maturity, compared to those from the west and west-central which were mainly dwarf to medium height, late heading and maturing. This differential distribution of landraces with height, heading and maturity period reflected the distribution pattern of different Oryza sativa landraces in Côte d’Ivoire, which could be useful in germplasm management and breeding programs.     

Reaction of wild species of the geneus Lens to drought

Summary Susceptibility to drought stress is a key factor in dry land lentil (Lens culinaris Medikus) production in the Mediterranean region of West Asia and North Africa. This study examined the response to drought stress of 121 accessions representing all subspecies of the genus Lens; cultivated, and the wild L. culinaris ssp. orientalis (Boiss.) Ponert, L. culinaris ssp. odemensis (Ladiz.), L. nigricans M.B. Godr. ssp. nigricans Godr. and L. nigricans ssp. ervoides (Brign.) Ladiz. for their potential use in breeding for dry land conditions. Accessions were grown under two moisture regimes (dry land and dry land plus supplemental irrigation) at Breda, Syria during the 1990–91 and 1991–92 seasons. The cultivated lentil had markedly superior seed and straw production than did the wild Lens species. Time to flowering accounted for less than 10% of the variation in yield of wild accessions under rainfed conditions in the two seasons, showing that, in contrast to the cultivated germplasm, drought escape was relatively unimportant in wild lentil. Performance under drought in wild lentil, measured in terms of dry land seed yield or drought susceptibility index (S), was randomly distributed among collection locations with little relation to collection site aridity. Direct selection of wild lentil germplasm for biomass yield under dry conditions is of little value and an evaluation of wild accessions in hybrid combination is needed.     

Levels of phenotypic variation for developmental traits in landrace genotypes of durum wheat (Triticum turgidum ssp. Turgidum L. conv. durum (Desf.) MK.) from Jordan

Summary A germplasm collection comprising 132 landrace genotypes of durum wheat (Triticum turgidum) ssp. turgidum L. conv. durum (Desf.) MK., from 10 districts in Jordan was evaluated for days to booting (DB), days to heading (DH), days to anthesis (DA), days to maturity (DM), filling period (FP) and grain yield (GY). In spite of the relatively strong genetic (Gr) and phenotypic (Pr) correlation coefficients between the developmental traits, the variation among genotypes was significant. GY was only positively and significantly correlated with DA and FP, while it was negatively correlated with DH. Genotypes with different combinations of early, medium and late developmental traits have been identified. Different combinations of DH and FP may or may not lead to differences in DM and GY. However, genotypes with long FP and medium-late DH gave the highest grain yield. Multiple linear regression of the duration of maturity (M) stage on the duration of booting (B), heading (H) and anthesis (A) stages (M = –260.87 + 2.09 B + 1.23 H + 0.9 A, R ² = 0.92) indicates that it is possible to manipulate the duration of these developmental stages and develop genotypes with high yield potential and early maturity, a highly desirable in drought-prone environments.     

CIMMYT's approach to breed for drought tolerance

Summary About 32% of the 99 million ha wheat grown in developing countries experiences varying levels of drought stress. Three major drought types have been identified: Late drought (LD) is common in the Mediterranean region, early drought (ED) is found in Latin America and wheat is produced on residual soil moisture (RM) in the Indian subcontinent and part of Australia. Until 1983, CIMMYT selected all germplasm under near optimum conditions for its yield potential and tested only advanced lines under drought. In spite of many critics, this approach proved to be successful, since in the mid 80's CIMMYT germplasm was grown on 45% of the wheat area in LC with annual rainfall from 300–500 mm and on 21% in areas with less than 300 mm. Since 1983, CIMMYT's drought breeding methodology is to alternate segregating populations between drought stressed and fully irrigated conditions (FI) and to test advanced lines under a line source irrigation system. To compare the efficiency of these approach, yield of four, mostly leading varieties, from each of the regions with LD, ED, RM, and FI and twelve recent CIMMYT cultivars selected for high yield under FI and RM conditions (ALT) were compared under four different moisture regimes (FI, LD, ED, and RM) in 89–90 and 90–91 in Yaqui Valley, Mexico. Genotypic correlation between yield and days to flowering, days to maturity, height, grains m^-2, TKW, test weight and grain fill period were calculated.Mean grain yield of the four best lines in the ALT group was highest under all moisture stress regimes, followed by the FI-group. However, the highest yielding cultivar within each moisture regime was from the FI-group under FI, from the LD-group under LD, and from the ALT-group under ED and RM conditions. Estimates for genetic advance suggest that FI is the best environment for increasing grain yield even in all three drought environments. This indicates that yield potential per se is beneficial also in drought environments. The highest yield in drought environments was realized by the CIM cultivars selected under FI and RM. Simultaneous evaluation of the germplasm under near optimum conditions, to utilize high heritabilities and identify lines with high yield potential, and under stress conditions to preserve alleles for drought tolerance seem at present the best strategy.     

Adaptation of barley (Hordeum vulgare L.) to harsh Mediterranean environments

Summary Differences in development of the apex may be the reason for cultivar differences in adaptation of barley to terminal drought in Mediterranean environments. The present study was conducted to identify apical development patterns of barley adapted to terminal drought stressed Mediterranean environments and to determine plant characteristics which can be used as criteria to select for an adapted development. Thirty-five two-rowed barley (Hordeum vulgare L.) entries were grown at two sites in northern Syria (Tel Hadya and Breda) in 1988/89. Four apical development patterns were observed: a slow or fast vegetative development, depending on the vernalization requirement, combined with a slow or fast generative development, depending on the daylength response of the crop. Early heading was related to fast generative development. Leaf appearance rates on the main shoot were constant during a major part of the pre-anthesis period, but significant differences were observed among development patterns. Genotypic differences in main shoot tiller number were associated with differences in the onset of tiller appearance and not with differences in tiller appearance rate or final leaf number on the main shoot. Since vernalization requirements and daylength responses are largely independent of terminal drought stress, selection for an adapted phenology can be done in favourable environments. Morphological traits related to these responses (winter growth habit, cold tolerance, plant colour, growth vigour, heading date) can be used as criteria for selection for adaptation to low-rainfall Mediterranean environments.     

Geneticanalysis of drought stress response in rapeseed (Brassica campestris and B. Napus). II. Yield improvement and the application of selection indices

Summary The expected improvement in grain yield if selection was made for plant characters measured between flowering and maturity was evaluated in populations of rapeseed (Brassica campestris and B. napus) grown in a droughted environment. Drought was commenced at flowering in each species and measurements were made on plants at the commencement of the drought stress, during the stress treatment and at crop maturity.Substantial genetic and phenotypic variation was observed in yield as well as the different morphophysiological determinants of yield. In B. campestris no single parameter was found to be a suitable alternative selection criterion to yield since the correlated responses in yield if selection was for another character was lower than if selection was for yield alone. By the use of a selection index however, joint selection for yield, as well as harvest index, 1000 seed weight and seeds per pod, was expected to be 20% more effective than direct selection for yield under drought. In the B. napus population direct selection for flowering time or for harvest index was predicted to result in a genetic advance in yield equal to or greater than that obtained by direct selection for yield, whereas joint selection for flowering time and yield should result in a 16% greater yield increase. Selection for increased yield in these populations grown in droughted conditions is discussed.     

Using host plant resistance to manage biotic stresses in cool season food legumes

Summary Kabuli chickpea (Cicer arietinum L.) is the common cultivated type of chickpea in arid and semi-arid environments in the Mediterranean region. Ascochyta blight, (Ascochyta rabiei (Pass.) Labr.), leaf miner (Liriomyza cicerina, Rond.) and cold, are the three most important stresses on chickpea grown under semi-arid conditions in this region. Phenotypic frequencies for responses to these stresses in the eight major chickpeagrowing regions of the world were estimated from 5,672 kabuli chickpea accessions assembled from these regions. In addition, the accessions were evaluated for 12 morpho-physiological and three phenological characters under semi-arid Mediterranean conditions at the International Center for Agricultural Research in the Dry Areas (ICARDA), Aleppo, Syria. Considerable regional differences in frequency distributions for response to the three stresses were observed. Average phenotypic diversity for responses to the three stresses was lower (H_o=0.474) than for morpho-physiological (H_o=0.754) and phenological (H_o=0.812) characters. The highest frequencies of accessions resistance to Ascochyta-blight and leaf-miner were found in South Asia and South Central Asia, respectively. A small number of chickpea breeding materials of ICARDA showed a moderate level of tolerance to cold. A group of four characters showing the strongest bivariate association with each of the three stresses was identified from the latter group. Then, a discrete multivariate log-linear analysis of the five-way frequency table was performed for each group. The simplest log-linear model for each group included both two- and three-factor association terms, but no independent factors. This suggested the potential for indirect selection for stress tolerance using one or more of these associated characters. The roles of these characters in ideotype breeding of kabuli chickpea for arid and semi-arid Mediterranean conditions deserves careful assessment.     

Identification of levels of resistance to cassava root rot disease (<Emphasis Type="Italic">Botryodiplodia theobromae</Emphasis>) in African landraces and improved germplasm using <Emphasis Type="Italic">in vitro</Emphasis> inoculation method

Cassava root rot disease is an increasing problem in Africa where yield losses of about 80% have been recorded. We evaluated 290 African landraces and 306 improved genotypes from the germplasm collections of the International Institute of Tropical Agriculture (IITA), for sources of resistance using root slice laboratory assay. Disease severity was assessed quantitatively by direct percentage estimation (PS) and by use of a rating scale (RS). Both methods of assessment were compared for identification of variability in the germplasm, and genotypes were classified into response groups using an enlarged rank-sum method that combined the PS and RS assessments. The two scoring methods revealed continuous variation (P < 0.001) for resistance in the sets of germplasm. Disease assessments based on PS and RS were highly correlated in both the improved germplasm (r = 0.75) and the landraces (r = 0.72). Based on PS assessment, 50 improved genotypes (16.3%) and 53 landraces (18.3%) showed significantly lower disease scores than the resistant control. The rank-sum method separated each set of collections into highly resistant, resistant, moderately resistant, moderately susceptible, susceptible and highly susceptible groups. Fifty-nine improved genotypes (16.4%) and 61 African landraces (16.9%) were identified as either highly resistant or resistant. Generally, these genotypes exhibited resistance by limiting the growth of the pathogen (reduced amount of invaded surface area). This type of rate-reducing resistance is highly heritable and a quantitative trait which can be harnessed in breeding. Genotypes subsets were identified for further studies into the genetic basis of resistance to root rot disease.     

Exotic barley germplasm: variation and effects on agronomic traits in complex crosses

Summary Utilization of exotic germplasm offers an approach to broaden genetic variability in breeding populations. This study was conducted in order to 1) compare germplasm of exotic origin with adapted Swedish barleys with respect to genetic differences and 2) to evaluate how exotic material affected agronomic performance in complex crosses. Allozyme studies showed the following Nei's gene diversities among parents: 0.13 (adapted parents), 0.16 (landraces) and 0.25 (H. spontaneum). Cluster analysis indicated that parental groups were genetically divergent. Earliness, straw length, number of ears per plant and thousand kernel weight (TKW) were studied. Variation in agronomic traits showed the following pattern: landraces > H. spontaneum > adapted lines. The best sources for earliness were adapted parents and landraces. Mean straw length was greatest in H. spontaneum lines. Number of ears per plant was quite similar in all groups. The highest TKW was among landraces and adapted parents. Hybrids from the complex crossing programme exceeded parents in earliness and TKW. An index composed from the four traits showed the most favorable frequency distributions for adapted parents and hybrids. Both genetic and agronomic studies indicate that new valuable variation from exotic germplasm may be introduced into barley breeding material.     

White lupin (Lupinus albus) variation for adaptation to severe drought stress

This study aimed to reduce the gap of knowledge on white lupin drought tolerance variation, by assessing the grain yield of 21 landraces from major historical cropping regions, one variety and two breeding lines in a large phenotyping platform that imposed controlled severely stressed or moisture‐favourable conditions after an initial stage of favourable growth. Drought stress reduced grain yield by 79%. Genetic correlation coefficients indicated moderate consistency of genotype responses across conditions for grain yield (r_g = 0.76), fairly high consistency for straw yield (r_g = 0.85) and harvest index (r_g = 0.91), and high consistency for flowering time (r_g = 0.99). However, low genetic correlation for yield (r_g = 0.31) occurred among a subset of genotypes with early phenology. Specific adaptation to either condition implied significant (p = 0.05) genotype × condition interaction of crossover type between well‐performing genotypes. Early flowering was an important stress escape mechanism, but intrinsic drought tolerance could be inferred from responses of a few genotypes. Various landraces out‐yielded the improved germplasm under stressed or favourable conditions.     

Resistance to Helminthosporium leaf blight and agronomic performance of spring wheat genotypes of diverse origins

Helminthosporium leaf blight (HLB), caused by a complex of Cochliobolus sativus (Ito & Kurib.) Drechsler ex Dastur and Pyrenophora tritici-repentis Died, is a serious disease of wheat (Triticum aestivum L.) in the warm lowlands of South Asia. Wheat cultivars grown in the area are either susceptible to HLB or possess low levels of resistance to it. A replicated field study was conducted in 1999 and 2000 at two sites in Nepal to determine the level of HLB resistance and other desirable traits in 60 wheat genotypes of diverse origin. The test genotypes were planted in main strips divided into two strips one of which was sprayed four times with Tilt (a.i. propiconazole) @ 125 g of a.i. ha^–1. Four readings of HLB were recorded to calculate the area under the disease progress curve (AUDPC). Other traits under investigation included biomass yield (BY), grain yield (GY), 1000-kernel weight (TKW), harvest index (HI), days to heading (DH) and maturity (DM), plant height (PHT), and effective tiller number (ETN). Wheat genotypes differed significantly for all traits. Mean AUDPC values ranged from 45 to 1268. A few exotic genotypes were highly resistant to HLB. Losses in GY due to HLB ranged from 2 to 26%, and TKW was reduced by up to 33%. A few genotypes showed HLB tolerance, i.e., relatively smaller GY and TKW reductions despite high levels of HLB. In general, medium to late maturity and higher levels of HLB resistance and low to high GY and TKW characterized genotypes exotic to South Asia. Biplot analysis identified several genotypes that were HLB-resistant and agronomically superior. Results suggest it is possible to improve HLB resistance of local wheat cultivars based on selective breeding using this pool of germplasm.     

Associations of some characters with seed yield in chickpea collections

Summary Three thousand two hundred and sixty-seven kabuli chickpea (Cicer arietinum L.) germplasm accessions were grown during the spring season of 1980 at Tel Hadya, the main research station of ICARDA, Syria to determine the components of seed yield. Observations were recorded on seed yield and 14 other characters. Correlation and path coefficient analyses were done to find out associations among characters and to assess the direct and indirect contribution of each character to seed yield.Large variation was observed for all the characters studied except days to flowering, days to maturity and protein content. Correlation and path coefficient analyses showed that biological yield and harvest index were the major direct contributors to seed yield. The 100-seed weight, plant height, days to flowering and maturity, canopy width, and protein content contributed to seed yield mainly through indirect effect via biological yield and harvest index. The 100-seed weight and seed yield were major contributors to biological yield. Major contributor to protein content was days to maturity. Results indicated that selection for high biological yield and harvest index would lead to high seed yield; and selection for large seed size would lead to high biological yield. Therefore, these characters should receive the highest priority in selecting high yielding plants in chickpea breeding.     

Transferability of Chickpea Selection Indices from Normal to Drought-prone Growing Conditions in a Mediterranean Environment*

Chickpea (Cicer arietinum L.) is normally grown as a spring-sown rainfed crop in the Mediterranean region in areas receiving more than 400 mm of rainfall annually. Because of large fluctuations in the total amount and distribution of the rainfall, the crop occasionally suffers from extreme terminal drought. Breeders face a dilemma in selecting materials for both normal moisture regimes and drought conditions because it is often difficult to conduct two parallel breeding programmes. This study examines the transferability of the selection indices (developed as a multiple regression of yield on most influential traits) from one environment to another with the objective of selecting an appropriate environment for the development of cultivars adapted to both environments. Experiments were conducted for 3 years (1989-1991) during the spring at two locations with normal moisture regime (hereafter referred to as drought-free) in Syria and Lebanon and one drought-prone location in Syria. Each year, 192-240 newly bred lines were evaluated in replicated trials for seed yield, days to flowering and maturity, plant height and 100-seed weight. Correlation analyses showed that increased seed size, early maturity and reduced plant height at the drought-prone location and early maturity at the drought-free location were of prime importance in increasing seed yield. Regression equations developed to predict seed yield showed that days to flower and maturity accounted from 67-80 % variation in seed yield at the dry location, whereas at the drought-free locations the contribution of days to maturity was smaller except in 1991 in Lebanon. The percent success in the transferability of the selection indices from the drought-free environments to the drought-prone environment was higher than that from the drought-prone environment to the drought-free environment. These results indicate that chickpea breeding material developed under drought-free environments could be useful under drought-prone conditions. Early-maturing lines selected under normal environments would also perform well under drier conditions.     

Diversity and adaptation in rice varieties under static (ex situ) and dynamic (in situ) management

This study compares genebank-conserved and farmer managed populations of the same farmers' varieties of rice. Seven varieties that had been collected twice, in the early 1980s and in 1991, were recollected in 1997 after having been grown continuously in farmers' fields. Since the first genebank collection, rice cultivation in the Meking delta has been intensified with a rather abrupt switch from single to double cropping, more use of chemical fertilisers, improved water management, and more market oriented production. Many farmers' varieties have been maintained as a second crop but with a considerably delayed planting time compared to previous practice. In this experiment, the ex situ materials represent adaptation to pre-intensification conditions while the in situ populations have been exposed to the intensive production system for a number of years. The materials were tested in the wet season of 1997 under current farmers' management practices in an on-farm field experiment within the area where the varieties originated. Agronomic, stress resistance and morphological traits and variation at 7 isozyme loci were observed in the field or laboratory. Analysis of variance (ANOVA) and Principal Component Analyses (PCA) were used to study differences in agronomic and morphological traits between ex and in situ populations. Isozyme variation was analyzed by Nei's diversity indices and Wright's F-statistics. Farmer-managed populations showed a general trend of later flowering and maturity time, more uniformity of grain quality, lower frequency of undesired off-types, and reduced drought stress tolerance compared with corresponding ex situ populations. There were no significant differences in grain yield or tolerance to biotic stresses. Allelic frequencies of isozymes showed no consistent differences that could be related to changes of the farming system. These results are interpreted as an adaptation to the changed farming system and include natural and farmers' selection for maturity time (all varieties are photoperiodic)and market standards. The poorer drought tolerance may reflect the fact that such stress was common before intensification but is not normally a factor under the current water management regime. For in situconservation strategies this case sheds some light on the survival of allelic diversity vs. adaptedness. Isozyme data indicate maintenance of allelic diversity. Adaptedness, however, is at risk under on-farm conservation. Natural and intentional selection will normally not remain constant over time. Consequent genetic changes include loss of adaptation to past conditions and building up of adaptation to new. In this case such changes have happened surprisingly fast. However, changes are limited to adaptation to factors of the environment and to market-relevant quality traits. Yield seems to be unaffected. Considering needs for crop improvement this case has kept the materials `updated' with respect to adaptation and unchanged with respect to yield potential. This revised version was published online in August 2006 with corrections to the Cover Date.     

Environment of selection and type of germplasm in barley breeding for low-yielding conditions

Summary Thee groups consisting of 332, 243 and 280 barley breeding lines (entries) of known selection history were evaluated in 10, 9 and 8 environments, respectively, to determine the relationship between grain yield in low yielding (LYE) or high yielding (HYE) environment, and selection history and type of germplasm. One cycle of selection in LYE produced on average five times more entries outyielding the best check in LYE than selection in HYE. A retrospective analysis indicated that the highest yielding lines in LYE were lower yielding (15%–28%) in HYE when compared with the best check, and by 20% and 38% compared with the best entries in HYE. In contrast, the highest yielding lines in HYE were lower yielding (4%–33%) in LYE when compared with the best check, and by 33% and 40% when compared with the best entries in LYE. The highest yielding lines in LYE did not differ consistently from the highest yielding lines in HYE for a number of morphological and developmental traits including days to heading. This suggests there are many paths to high yield in LYE and that analytical breeding based on individual traits may not be appropriate for variable environments. Only 0.07% of the highest yielding entries in LYE was selected for high yield in HYE conditions confirming previous results indicating that selection for high yield in HYE is an inefficient strategy for improving yield in low yielding conditions. This frequency is 28 times lower than the frequency of high yielding entries in LYE selected from landraces or crosses with landraces in low yielding conditions. The results imply that the most cost-effective strategy for barley breeding in low yielding conditions is to select repeatedly in low yielding conditions and to include adapted germplasm (landraces) in the breeding material.