Lupinus Research Articles

Information costs of reduced-effort habitat monitoring in a butterfly recovery program

Habitat monitoring typically requires a large amount of effort and resources. Project managers are likely to consider cost-cutting options but they may not critically review the information costs of implementing those options. An effort recently began in New York State to monitor critical habitat and restoration progress aimed at recovering the federally endangered Karner blue butterfly (Lycaeides melissa samuelis). Specific strategies were proposed to increase efficiency of fieldwork: (1) estimate larval host plant (wild blue lupine, Lupinus perennis) abundance from cover data, (2) area-based standardization of sample size for nectar sampling, (3) use a minimum cover threshold to trigger nectar species counts, (4) use multiple observers. I quantitatively reviewed these time savers for effects on raw data quality, and for potential effects on interpreting habitat quality as part of recovery criteria. Cover-based models of lupine abundance differed between metapopulation recovery areas, area-based sampling was sufficient to detect over 80% of priority nectar species in most sites, nectar information loss was high due to the minimum cover threshold, and overall results from different observers were similar with one exception. Direct lupine stem counting is recommended over cover estimation, statistical interpolation (rarefaction) and extrapolation (asymptotic estimators) is recommended over area-based nectar sampling, use of minimum cover criteria is strongly discouraged, and field crews need botanical expertise and careful instructions. This example highlights why strategies to save time and money with monitoring fieldwork should be ‘put to the test’.

Determination of the relationship among Old World Lupin (Lupinus sp.) species using RAPD and ISSR markers.

Randomly Amplified Polymorphic DNA (RAPD) and Inter Simple Sequence Repeat (ISSR) markers were used to determine the genetic relationships among 20 Old World lupin genotypes from 3 lupin species (Lupinus albus, L. angustifolius, L. luteus) using 15 primers in each case. Polymorphism was observed in reactions with 9 out of 15 tested RAPD primers, and 11 out of 15 tested ISSR primers. Relationship was determined on the basis of polymorphic products analysis which was shown in the form of dendrograms by UPGMA method. Both RAPD and ISSR markers worked well to identify distinct grouping among three species and among genotypes within a species. The parallel use of both data set seems to allow precise estimation of cultivars relationship and diminishing mistakes connected with method’s technical limitations. Such marker systems may potentially be used in the development of new lupins cultivars. Key words: White lupin, genetic diversity, RAPD, ISSR.

Winter survival of pea, faba bean and white lupin cultivars in contrasting Italian locations and sowing times, and implications for selection

SUMMARYIncreasing grain legume yields via autumn sowing requires winter-hardy material. Forty-nine pea, 24 faba bean and 11 white lupin cultivars recently released by 29 breeding institutions worldwide were sown in autumn 2002 at Lodi (northern Italy) and Foggia (southern Italy) on two different dates at each location, with the objective of assessing the winter survival of species and cultivars, its consistency across locations and sowing times, its relationship with grain yield and various morphophysiological traits and the implications for selection. The winter seasons were representative for the sites and had 54 frost days and −7·3°C absolute minimum temperature in Lodi, and 8 frost days and −3·4°C absolute minimum temperature in Foggia. The species differed in optimal sowing time at each location. Mild winter temperatures, preventing a sufficient hardening against late frosts, led to higher proportions of plants killed over winter in Foggia than in Lodi (0·17 v. 0·12 for pea; 0·20 v. 0·16 for faba bean; 0·34 v. 0·11 for lupin). Variation in winter mortality was much larger within species than among species. Winter mortality and grain yield of cultivars were inversely correlated (r=−0·49, −0·43 and −0·74 for entry means over locations of pea, faba bean and lupin, respectively). The consistency across environments of genotype winter mortality, estimated by genetic correlations, was high across sowing times and low across locations for faba bean, moderate across sowing times and fairly low across locations for pea, and always fairly high for lupin. A visual cold tolerance score was always related to better winter survival. Winter survival tended to correlate with larger seeds in faba bean and lupin, and was associated with later flowering in lupin. A rosette-like winter growth habit (defined by lower height/number of leaves ratio of seedlings in January) was correlated with winter survival in all species (after partialling out the influence of seed size in faba bean and lupin). The optimal index of indirect selection for low winter mortality included the cold tolerance score, the seedling height/number of leaves ratio and the seed size for faba bean, the first two of these variables for pea, but only the first variable for lupin. The gain in predicted efficiency over direct selection was high for faba bean (29%) and modest for pea and lupin (⩽10%).

Resource availability, matrix quality, microclimate, and spatial pattern as predictors of patch use by the Karner blue butterfly

Determination of which aspects of habitat quality and habitat spatial arrangement best account for variation in a species’ distribution can guide management for organisms such as the Karner blue butterfly ( Lycaeides melissa samuelis), a federally endangered subspecies inhabiting savannas of Midwest and Eastern United States. We examined the extent to which three sets of predictors, (1) larval host plant ( Lupinus perennis, wild lupine) availability, (2) characteristics of the matrix surrounding host plant patches, and (3) factors affecting a patch’s thermal environment, accounted for variation in lupine patch use by Karner blues at Indiana Dunes National Lakeshore, Indiana and Fort McCoy, Wisconsin, USA. Each predictor set accounted for 7–13% of variation in patch occupancy by Karner blues at both sites and in larval feeding activity among patches at Indiana Dunes. Patch area, an indicator of host plant availability, was an exception, accounting for 30% of variation in patch occupancy at Indiana Dunes. Spatially structured patterns of patch use across the landscape accounted for 9–16% of variation in patch use and explained more variation in larval feeding activity than did spatial autocorrelation between neighboring patches. Because of this broader spatial trend across sites, a given management action may be more effective in promoting patch use in some portions of the landscape than in others. Spatial trend, resource availability, matrix quality, and microclimate, in general, accounted for similar amounts of variation in patch use and each should be incorporated into habitat management planning for the Karner blue butterfly.

Genetically Engineered Plants, Endangered Species, and Risk: A Temporal and Spatial Exposure Assessment for Karner Blue Butterfly Larvae and Bt Maize Pollen

Genetically engineered maize (Zea mays) containing insecticidal endotoxin proteins from Bacillus thuringiensis (Bt) delta-endotoxin proteins has been adopted widely in the Midwestern United States. The proteins are toxic to several lepidopteran species and because a variety of maize tissues, including pollen, may express the endotoxins, the probability of exposure to nontarget species, including endangered species, needs to be understood. The objective of this study was to assess the potential temporal and spatial exposure of endangered Karner blue butterfly larvae (Lycaeides melissa samuelis) to Bt maize pollen in Wisconsin using probabilistic exposure techniques and geographic information systems analysis. Based on degree-day modeling of butterfly phenology and maize pollen shed, there is some potential for temporal exposure of larvae to maize pollen. However, in the majority of years and locations, maize pollen shed most likely will occur after the majority of larval feeding on wild lupine (Lupinus perennis). The spatial analysis indicates that some Karner blue butterfly populations occur in close proximity to maize fields, but in the vast majority of cases the butterfly's host plant and maize fields are separated by more than 500 m. A small number of potential or existing Karner blue butterfly sites are located near maize fields, including sites in two of the four counties where temporal overlap is most likely. The exposure assessment indicates that these two counties should receive the highest priority to determine if Karner blue butterfly larvae are actually at risk and then, if needed, to reduce or prevent exposure.

Effects of self-pollination and maternal resources on reproduction and offspring performance in the wild lupine, Lupinus perennis (Fabaceae)

We examined the effects of self-pollination and resource addition to maternal plants of Wild Lupine on seed production in the field, and on offspring performance in the greenhouse. Although 24% of flowers set fruits when open-pollinated, only 11% of flowers set fruits when self-pollinated. Self-pollination significantly reduced fruit and seed production per inflorescence and increased aborted seeds per fruit. Resource addition to maternal plants significantly increased fruit and seed number in the field. Moreover, selfed plants exhibited greater variability in seed production in the absence of resource addition to the maternal plant. We planted a total of 1,306 of the seeds from this experiment in the greenhouse. While self-pollination did not affect the proportion of seeds emerging, it slowed seedling emergence by 5–10%, and reduced offspring biomass by 25–35%. Interestingly, resource addition to the maternal plants significantly decreased proportion of seedlings surviving after 5 months. Moreover, offspring from maternal plants with resource addition expressed more inbreeding depression in the seedling stage compared to offspring from maternal plants without resource addition, for which more inbreeding depression occurred during seed maturation and emergence. These results indicate that conservation efforts using benign environments to increase number of seeds or offspring may face compensating reductions in survivorship at other life stages.

Exudation of low molecular weight organic acids byLupinus albusL.,Lupinus angustifoliusL. andLupinus luteusL. as affected by phosphorus supply

Effet de l'apport de phosphore sur l'exudation d'acides organiques de faible poids moleculaire par Lupinus albus L., Lupinus angustifolius L. et Lupinus luteus L. L'objectif de cette recherche etait d'etudier l'influence d'une application de fertilisant phosphate sur la qualite et la quantite d'acides organiques exudes par six varietes de trois especes de lupins: Lupinus albus L. (varietes Minori et Nelly), Lupinus angustifolius L. (varietes Borweta et Bordako) et Lupinus luteus L. (varietes Borsaja et Borselfa). Nous avons egalement etudie l'influence du milieu de collecte d'exudat (eau deionisee, solution a 0.05 mM CaCl 2 ) sur la composition et l'intensite des acides origaniques exudes par L. albus (variete Minori). Les plantes ont ete cultivees dans un sable quartzeux avec (+ P) et sans (- P) fertilisation pendant 21 jours.[...]

Effects of Culture Filtrates of Rhizobacteria Isolated from Wild Lupine on Germination, Growth, and Biological Nitrogen Fixation of Lupine Seedlings

In order to select potential Plant Growth Promoting Rhizobacteria (PGPRs), a selection of strains from the predominant genera in the rhizosphere of four lupine species, based on genetic divergence criteria, carried out in a previous study, yielded 11 Aureobacterium (Aur), four Cellulomonas (Cell), two Arthrobacter (Arth), two Pseudomonas (Ps), and six Bacillus (Bc) strains. Cell‐free culture filtrates of each bacterium were assayed for effects on germination, growth, and biological nitrogen fixation (BNF) of Lupinus albus L. cv. Multolupa seeds or seedlings. Four (Aur 6, Aur 9, Aur 11, and Cell 1) of the twenty‐five strains assayed promoted germination. Aureobacterium 6 and Aur 9 also increased root surface, total nitrogen content, and BNF. As a result of the screening, and considering all the variables studied, Aur 6 can be considered a plant growth promoting rhizobacterium and is suitable for further field trials in other plants and in different production systems.

The gene encoding the PSST subunit of respiratory chain complex I is present in more than one copy in yellow lupine

Three copies of the PSST gene were identified in the lupine genomic root library, however, only two transcripts were found in the lupine root cDNA library. The transcript of the third PSST gene was identified in RNA from lupine flowers. The genes are 92% identical in the coding region, while the 5′ parts of the reading frames specifying the N-terminal presequences for mitochondrial import show about 87% sequence identity. The differences between genes concern mostly the third nucleotide of the codons and the length of the intron. Transcripts of three PSST genes differ in abundance in flowers and leaves.

COMPARISON OF DIFFERENT METHODS TO DETECT PHYTOPHTHORA SPP. IN RECYCLING WATER FROM NURSERIES

Five methods for the detection of Phytophthora spp. in water recirculation systems of four commercial nurseries were compared: direct plating, three bait tests (apple, lupin seedling, Rhododendron leaf test) and DASELISA with two Phytophthora specific polyclonal antisera. Water and sediment samples were taken from different places in the recirculation systems over one year. With all methods except direct plating Phytophthora spp. could be detected in the four nurseries. In most cases the detection rates were influenced more by the method itself than by different conditions on particular nurseries. The Rhododendron leaf test trapped the widest range of Phytophthora species and was the most successful method followed by ELISA with antiserum against P. cinnamomi. Although the chemical analysis of the water samples showed a good water quality for nursery cultivation, there was a close relationship between the detection rates obtained with ELISA and the total salt content of the water. The results are compared with those of in vitro experiments and their relevance for commercial nurseries is discussed.

Influence of day length and temperature on number of main stem leaves and time to flowering in lupin

SummaryA growth chamber experiment was carried out to investigate the influence of day length and temperature on the development of flowering in eight varieties of the three grain lupin species Lupinus albus (Wat and C3396), L. angustifolius (Gungurru, Polonez and W26) and L. luteus, (Juno, Radames and Teo). The plants were grown at two temperatures, 10°C and 18°C, in combination with five daylength regimes: 10, 14, 18, 24 h day at full light intensity and 10 h full light extended with 8 h low intensity light. Increased daylength decreased days from sowing to flowering in all varieties, but had little effect on thermal time to flowering in most varieties. However, C3396, W26 and Radames had a significantly longer thermal time to flowering at high, non‐vernalising temperature (18°C) at short daylengths. Low light intensity daylength extension did not significantly influence thermal time to flowering. For flower initiation, measured as number of leaves on the main stem three types of response were found. All varieties formed fewer leaves on the main stem at 10°C than at 18°C, although the two thermo‐neutral varieties of L. luteus, Juno and Teo, gave only a small response to temperature and daylength. In Polonez, Gungurru and Wat, low temperature decreased leaf number, but there was only a small response to changes in daylength. Three varieties, C3396, W26 and Radames, showed longer thermal time to flowering at 18°C with short daylengths. This could be explained by a greater number of main stem leaves formed at short daylength at non‐vernalising temperatures. Increased daylength decreased leaf number in these varieties, but never to a smaller number than for plants grown at 10°C. In these varieties, low intensity extension of the daylength had a similar (W26, Radames) or decreased (C3396) effect compared to full light extension. The hastening of time to flowering by long days could be separated into two effects: a high light energy effect hastened development by increasing the rate of leaf appearance in all varieties, while low light energy in thermo‐sensitive varieties was able to substitute for vernalisation by decreasing leaf number.

The effect of military training activity on eastern lupine and the Karner blue butterfly at Fort McCoy, Wisconsin, USA.

The US Department of Defense (DOD) manages over 10.1 million ha of land, much of which is used for training military personnel. However, vast sections receive little or no use, and military lands have become refuges for many species. At Fort McCoy, Wisconsin, USA, populations of the endangered Karner blue butterfly (Lycaeides melissa samuelis) are found in oak and pine barren communities where wild lupine (Lupinus perennis), a perennial forb required by Karner blue butterfly larvae, still occurs. Oak and pine barren communities are disturbance-dependent, and the barrens ecosystems in the Midwest have declined in extent by 98% because of fire suppression, succession, and habitat fragmentation. We studied the effects of disturbance by military manuever training on the density of lupine and Karner blue butterfly at Fort McCoy. We also wanted to determine whether military training activity could enhance Karner blue butterfly habitat. At locations where tracked vehicles had driven through lupine patches, the abundance of lupine and nectar-producing plants was greater in the median strip between vehicle ruts than in vehicle ruts or 5 m outside the vehicle ruts. The proportion of lupine stems with Karner blue butterfly larvae feeding sign (the ratio of stems fed upon to stems examined) was greater in areas where military vehicles had traveled than where they had not. The proportion of lupine stems with feeding sign and lupine stem density was also positively related to the occurrence of prior bivouacs and fires caused by military munitions. Shrub and forest canopy abundance were lower in areas traveled by tracked vehicles. At the scale of the lupine patch, lupine abundance and the proportion of lupine stems with feeding sign were positively correlated with military training activities, suggesting that maintenance of lupine habitat can be achieved in concert with military training.

Effects of Temperature and Photoperiod on Days to Flowering, Yield and Yield Components of Lupinus albus (L.) under Field Conditions

A better understanding of the agronomic importance of planting date and the influence of cold temperatures and photoperiod during germination and plant growth may lead to better management strategies for cultivation of the sweet white lupin (Lupinus albus). The effects of planting date (temperature and photoperiod) were determined on the number of days to flowering, yield and yield components of four early to medium and one late sweet white lupin genotype in a field trial at Potchefstroom, South Africa, planted during February 1996 to January 1997. Moisture stress was avoided through regular irrigation. Duration of the developmental phases planting date to emergence, emergence to floral initiation, initiation to first flower, duration of flower and days to physiological and harvest maturity was related to field measurements of temperature and photoperiod. Differences in the main determinants of yield, i.e. seeds per pod, pods per plant, single seed mass (SSM), plant and pod height and yield, were measured. Results showed that both temperature and photoperiod influence the growth and development of the Lupinus albus genotypes ‘Esta’, ‘Hantie’, ‘Tifwhite’, ‘Kiev’ and ‘LAL 186’. Temperature influences include the effect of vernalization at seedling emergence. Minimum grass temperatures under 5 °C at emergence are effective for vernalization. However, after grass temperatures at emergence increased again from June to December, to gether with an increase in the photoperiod length, ‘Tifwhite’ as well as the other genotypes still flowered earlier, confirming that these cultivars are long‐day plants, which is in accordance with controlled‐environment data. Cool vernalizing temperatures thus not only influence obligate vernalization requiring genotypes such as ‘Tifwhite’, but also influence the non‐obligate genotypes studied. Plan‐ting date had a significant influence on pods per plant, single seed mass (SSM) and seed yield. In all trials laterplanting, from June to November, decreased SSM and seed yield. The highest seed yield of 1.5 t ha−1 was obtained for the 10 June planting date and the lowest average yield of 0.450 t ha−1 for the 5 November planting date.

Fish meal replacement by plant meals in extruded feeds for Atlantic salmon, Salmo salar L.

The replacement of fish meal protein with soybean meal (SB) or protein concentrates made from narrow-leafed lupin (LP) or field peas (PP) was investigated in extruded feeds for Atlantic salmon. Salmon (47 g) were fed for 63 days on extruded feeds containing each of the plant meals to replace 25% and 33% of the fish meal protein and performance compared against a nutritionally balanced control and a commercial salmon feed formulation (extruded under the same conditions). There were no significant differences in weight gain between the control and feeds containing the plant proteins. The commercial feed produced significantly higher weight gain than the control feed and LP at both replacement levels. Feed consumption was significantly higher for LP at 33%, but there were no other significant differences between the other feeds. Feed efficiency ratio (FER) and productive protein value (PPV) were highest for PP and SB and not affected by inclusion level, whereas they were significantly lower for LP at 33% inclusion. The weight gain and feed efficiency ratio data showed that soybean meal and pea protein concentrate had the best potential for replacing at least 33% of the fish meal protein in extruded salmon feeds and that lupin protein concentrate was less well utilised at the higher inclusion level. These results support the use of processed plant meals as important replacement protein sources for fish meal in extruded feeds for Atlantic salmon.

Genetic evidence supports the new anatolian lupine accession,Lupinus anatolicus, as an old world “rough-seeded” lupine (sectionScabrispermae) related toL. pilosus

A noteworthy wild lupine accession was recently discovered in southwestern Turkey and was proposed as a new separate Old World “smooth-seeded” species close toL. micranthus and namedL. anatolicus. Its species status was controversial with respect to cytological and crossability data. In order to examine the position and the evolutionary relationships of this Anatolian accession relative to the Old World lupines, we investigated new data from seed coat micromorphology, and from internal transcribed spacer (ITS) nucleotide sequences of the nuclear ribosomal DNA repeat. The micromorphological seed coat pattern ofL. anatolicus, as revealed by scanning electron microscopy, is characterized by pluricellular tubercles, which represent the typical and unique pattern of the Old World “rough-seeded” lupines (sect.Scabrispermae). In accordance with the micromorphological results, the genetic distances and phylogenetic relationships among the Old World lupines, estimated from ITS data, unambiguously support the new Anatolian lupine accession as part of theL. pilosus-L. palaestinus lineage within the strongly monophyletic group containing all theScabrispermae. The results provided in this study, together with other lines of data available from the literature, are thus hardly compatible with the hypothesis that this new Anatolian lupine accession could be related to Old World “smooth-seeded” lupines (includingL. micranthus); instead, it appears closely related toL. pilosus.

Composition of mycoflora and aflatoxins in lupine seeds from the Sudan.

Thirteen seed samples of lupines (Lupinus termis Forrsk.) were evaluated for their phytosanitary status by incubation on potato-dextrose agar (PDA) and moist sterile filter papers (blotter method) at 28±2oC. The seeds were also assayed for the presence of aflatoxins and toxigenic fungi. Forty-seven species and 10 varieties in 14 genera of fungi were recovered. Among these, 45 species and 10 varieties were new to lupine seeds, and 5 species and 2 varieties were new to the mycoflora of the Sudan. The genus Aspergillus (10 species, 8 varieties) was the most common, followed by Rhizopus (1 species), Fusarium (6 species) and Alternaria (5 species), while the remaining genera (Chaetomium, Cladosporium, Curvularia, Drechslera, Penicillium, Phoma, Emericella, Mucor, Sclerotium, Ulocladium) displayed lower levels of contamination. Of possible pathogens on lupine plants, Alternaria (5 species) (stem lesion) and F. oxysporum (root rot) were recovered from the seeds. Thin-layer chromatographic analysis of chloroform extracts of the 13 seed samples revealed that two samples contained very low concentrations of aflatoxins B1 and B2 (4.5-6.5 µg/kg).

Habitat use by the endangered Karner blue butterfly in oak woodlands: the influence of canopy cover

The Karner blue butterfly Lycaeides melissa samuelis is an endangered species residing in the Great Lakes and northeastern regions of the United States. Increased canopy cover is a major factor implicated in the decline of the Karner blue at many locales. Therefore, we examined how the butterfly's behavior varied with canopy cover. Adult males at Indiana Dunes National Lakeshore used habitat under canopy openings for nearly 90% of their activities; females used openings and shaded areas more equally. The frequency of oviposition on the sole host plant, wild lupine Lupinus perennis, was highest under 30–60% canopy cover even though lupine was more abundant in more open areas. Larvae fed preferentially on larger lupine plants and on lupines in denser patches. However, lupines were generally larger in the shade. Therefore, shade-related trade-offs existed between lupine abundance and distribution of larval feeding and oviposition. Also, heterogeneity of shading by sub-canopy woody vegetation was greater at oviposition sites than at sites where lupine did not grow. Given the importance of shade heterogeneity, a mixture of canopy openings and shade, on a scale similar to daily adult movement range, should be beneficial for this butterfly.

The effect of canopy cover and seasonal change on host plant quality for the endangered Karner blue butterfly (Lycaeidesmelissasamuelis).

Larvae of the Karner blue butterfly, Lycaeidesmelissasamuelis, feed solely on wild lupine, Lupinusperennis, from the emergence to summer senescence of the plant. Wild lupine is most abundant in open areas but Karner blue females oviposit more frequently on lupines growing in moderate shade. Can differences in lupine quality between open and shaded areas help explain this disparity in resource use? Furthermore, many lupines are senescent before the second larval brood completes development. How does lupine senescence affect larval growth? We addressed these questions by measuring growth rates of larvae fed lupines of different phenological stages and lupines growing under different shade conditions. The habitat conditions under which lupines grew and plant phenological stage did not generally affect final larval or pupal weight but did significantly affect duration of the larval period. Duration was shortest for larvae fed leaves from flowering lupines and was negatively correlated with leaf nitrogen concentration. Ovipositing in areas of moderate shade should increase␣second-brood larval exposure to flowering lupines. In addition, larval growth was significantly faster on shade-grown lupines that were in seed than on similar sun-grown lupines. These are possible advantages of the higher-than-expected oviposition rate on shade-grown lupines. Given the canopy-related trade-off between lupine␣abundance and quality, maintenance of canopy heterogeneity is an important conservation management goal. Larvae were also fed leaves growing in poor soil conditions and leaves with mildew infection. These and other feeding treatments that we anticipated would inhibit larval growth often did not. In particular, ant-tended larvae exhibited the highest weight gain per amount of lupine eaten and a relatively fast growth rate. This represents an advantage of ant tending to Karner blue larvae.

Food from lupins

Food from lupins

Genotypic variation of seed yield and architectural traits in dwarf autumn-sown white lupin

Variabilite genetique du rendement en graines et de l'architecture chez le lupin blanc d'hiver de type nain. Le nanisme est un moyen de reduire le developpement vegetatif exuberant du lupin blanc d'hiver. Trente-cinq genotypes de lupin blanc d'hiver de type nain ont ete evalues dans cinq milieux en France. Ces trente-cinq genotypes sont XA 100, le geniteur de nanisme et trente-quatre autres genotypes nains selectionnes dans la descendance de croisements entre XA100 et des genotypes non-nains. Le rendement en grains, le poids de 1 000 grains et le nombre de grains par metre carre presentent une large variabilite genetique et une forte heritabilite. Le rendement moyen de la population naine est egal a 37,8 q/ha et celui du geniteur de nanisme a 34,9 q/ha. Sur les cinq milieux, l'interaction genotype milieu n'est pas significative pour le rendement en grains. Les ramifications primaires constituent le niveau de ramification ayant la plus forte contribution au rendement. Des caracteres d'architecture ont ete notes dans quatre lieux. La date de floraison et le nombre de feuilles sont genetiquement tres variables. La longueur des entre-noeuds varie selon les genotypes. Peu de relations ont ete observees entre le rendement en grains et les caracteres d'architecture. La correlation genetique entre la longueur des tiges, tige principale ou ramifications primaires, et le rendement est positive. La date de floraison n'est pas correlee au rendement.