| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Texas Agric. Exp. Stn., 2301 Experiment Station Rd, Bushland, TX 79012
b Montana State Univ., Cent. Agric. Res. Cent., HC 90 Box 20, Moccasin, MT 59462
c Oregon State Univ., Columbia Basin Agric. Res. Cent., P.O. Box 370, Pendleton, OR 97801
* Corresponding author (w-payne{at}tamu.edu)
Received for publication February 2, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: PNW, Pacific Northwest
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
The inland PNW climate has been classified as midlatitude semiarid, or "Interior Mediterranean" using the Köppen method (Ramig, 1988, citing Critchfield, 1966) because 
70% of the 250 to 450 mm of annual precipitation is received during winter months. Spring weather rapidly transitions from cold and wet to hot and dry (Payne et al., 2001). Therefore, potential alternative spring crops need to tolerate cold temperatures during early growth stages, when there is sufficient moisture for germination, and to escape or tolerate high temperatures, high evaporative demand, and drought during later growth stages when rain is increasingly improbable.
Lupinus is a genus of self- or cross-pollinating, mostly indeterminate plant species indigenous to diverse geographic regions. Virgil noticed the benefits of rotating wheat with lupin more than 2000 yr ago (Gladstones, 1998). White lupin was domesticated in Germany during World War I in response to a need for high-protein pulse crops adapted to temperate conditions. More recently, narrow-leafed lupin was domesticated in semiarid southern Australia (see Cowling et al., 1998, for review) where it is grown during the winter and spring months. For both narrow-leafed and white lupin species, domestication efforts have centered on low alkaloid content, early flowering, reduced pod shattering, and soft seeds (Cowling et al., 1998). Cowling (2001) provides a list of alternative taxonomic nomenclature and common names for lupin.
Advantages of growing lupins in wheat-based cropping systems include increased yield of the following wheat crop (Cox, 1998; Gregory, 1998), contribution of N and organic matter to soils, and the ability to directly feed lupin grain on-farm to poultry, cattle, and sheep due to the absence of alkaloids or trypsin inhibitors (Edwards and van Barneveld, 1998).
White lupin was first introduced into the southeastern USA during the 1930s as a plowdown green manure (Meronuck et al., 1991). Production peaked in the 1950s and then declined due to reduction of seed supply, loss of government support, availability of N fertilizer, freeze damage, and disease pressure. Despite recent renewed interest in white lupin in North America (Putnam et al., 1992, 1993; Noffsinger and van Santen, 1995; Faluyi et al., 1997; Kearney, 1999; Reeves et al., 1999), it remains a very minor crop. We know of no published studies on white lupin in the inland PNW, but local farmers and researchers at Oregon State University and Washington State University have occasionally experimented with the crop. In most cases, efforts were abandoned due to yield instability and pod disease tentatively attributed to the bacteria genus Erwinia. However, newer cultivars in Australia have been selected for increased yield stability and disease resistance (Cowling et al., 1998).
To our knowledge, there have been no studies on the agronomic feasibility of narrow-leafed lupin in the PNW, but there exists a potential market. Regionally, there has been recent expansion of the dairy industry, which demands reliable sources of high quality protein. Currently, one-half of the lactating dairy cows in the region are fed imported canola (Brassica napus) meal from Canada, and approximately 60% are fed soybean [Glycine max (L.) Merr.] meal grown in the Midwest (Mowery and Spain, 1999). Comparisons for milk production and quality between cows consuming lupin-based diets and cows consuming soybean-based diets have been variable (May et al., 1993; Moss et al., 1999; Singh et al., 1991) but provide evidence that lupin might be substituted as a high quality protein source, with the added advantage that no extrusion or processing is required. There is also international market potential; approximately 75% of the 1000000 t of lupin annually produced in Australia is exported to markets in Asia, including Korea, Japan, and Indonesia, and to Europe and South Africa (Cox, 1998), mostly as a protein supplement to ruminant and monogastric animal feeds. Many of these countries already import soft white wheat from the inland PNW.
Growing conditions of the inland PNW differ in important ways from those of southern Australia. Average temperature and growing season rainfall of the two regions are similar, but lupin is sown under conditions of decreasing temperature and daylength in southern Australia, whereas in the inland PNW, they would be sown under conditions of increasing temperature and daylength.
The objective of this study was to evaluate agronomic potential of white and narrow-leafed lupin for wheat-based cropping systems of the inland PNW.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
Experiment 1
In 1998 and 1999, four public narrow-leafed lupin cultivars—‘Merrit’, ‘Yorrel’, ‘Chittick’, and ‘Danja’—and four public white lupin cultivars—‘Kiev Mutant’, ‘L2085’, ‘Ultra’, and ‘L1047’—were grown at Oregon State University Columbia Basin Agricultural Research Center stations near Pendleton (45°43'N, 118°W; elev. 454 m) and Moro (45°29'N, 120°30'W; elev. 561 m). Mean annual precipitation at Pendleton is approximately 420 mm, of which 70% is generally received between 1 September and 11 April. Mean annual precipitation at Moro is 280 mm with similar seasonal distribution.
The experimental design was a randomized complete block with four replications. In 1998, individual plot size at both sites was 1.5 by 5.2 m. In 1999, plot size was 1.5 by 5.8 m.
In 1998, plots were sown on 28 April at Pendleton and 27 April at Moro. In 1999, plots were sown on 30 March at Pendleton and 1 April at Moro. In both years, seeds were sown into tilled ground from which winter wheat had been harvested the previous summer. In both years and at both sites, sowing rate was 86 seeds m–2 using a Hege 551 (Hege Maschinen GmbH, Domäne Hohebuch, Waldenburg, Germany) plot drill with double-disc openers spaced 15 cm apart. Seeds were treated with benzene hexachloride (1,2,3,4,5,6-hexachloro cyclohexane) at a rate of 0.85 g kg–1 seed and carboxin (5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxamide) at a rate of 4.1 g kg–1 seed. Seeds were inoculated with Bradyrhizobium sp. (Lupinus) before sowing.
Granular fertilizer 16–20–0–14 (N–P–K–S) was applied before sowing at a rate of 112 kg ha–1. Ethalfluralin [N-ethyl-N-(2-methyl-2-propenyl)-2,6-dinitro-4- (trifluoromethyl) benzenamine] was applied preplant in liquid form at a rate of 0.84 kg a.i. ha–1 and incorporated with a cultivator into the upper 5 cm of soil. Metribuzin [4-amino-6-(1,1-dimethylethyl)-3-(methylthio)-1,2,4-triazin-5(4H)-one] was broadcast pre-emergence at a rate of 0.28 kg a.i. ha–1.
In 1998, soil samples were taken with an auger at 30-cm intervals to a depth of 1.2 m in three of the four replications for each cultivar at plant emergence and again after harvest to calculate change in profile water storage during the growing season. The change in water storage and rainfall data was used to calculate crop water use (crop transpiration plus evaporation from the soil surface) during the growing season. When calculating crop water use, drainage from the root zone and runoff were assumed to be negligible (Payne et al., 2001; Chen and Payne, 2001). Water use efficiency was calculated from yield divided by crop water use. Samples were hand-harvested at maturity from one randomly chosen 1-m2 area within the plot, including any grain on the ground.
Experiment 2
Five narrow-leafed lupin cultivars were sown under no-till conditions at three sites on three dates, using three sowing rates. A randomized complete block experimental design with four replications was used with a split-split plot restriction on treatment arrangement. Main plots were sowing date, subplots were sowing rates (112, 168, and 224 kg seeds ha–1), and sub-subplots were cultivar, including one public cultivar (Merrit) and four Plant Breeders Rights–protected cultivars (‘Belara’, ‘Kalya’, ‘Tallerack’, and ‘Tanjil’). The 100-seed weight for these cultivars was 14.8 g for Merrit, 14.4 g for Tanjil, 13.2 g for Kalya, 15.0 g for Belara, and 13.8 g for Tallerack. Sub-subplot dimensions were 3.1 by 31 m. A John Deere 1560 no-till drill, fitted with 18 single disc openers spaced 17.5 cm apart, was used. Seeds were sown to moisture at 2.5 to 3.5 cm.
Experimental fields were near Helix, Condon, and Lexington, OR. Mean annual precipitation is 400 mm at Helix, 300 at Condon, and 250 mm at Lexington. At the Helix site, lupin was sown on 2 April, 23 April, and 10 May 1999. At the Lexington site, sowing dates were 31 March, 20 April, and 8 May 1999. At the Condon site, seeding dates were 8 April, 29 April, and 14 May 1999. Lupin was sown into wheat stubble at the Helix and Lexington sites and into barley stubble at the Condon site.
Ethalfluralin was applied presowing, at a rate of 0.84 kg a.i. ha –1, with a GANDY spreader (GANDY Co., Owatonna, MN), using a 10% granular formulation to reduce adherence of herbicide to surface residue. Metribuzin was broadcast-applied postsowing, pre-emergence at a rate of 0.28 kg a.i. ha–1. Incorporation was not possible due to no-till conditions. Bradyrhizobium sp. (Lupinus) inoculant was mixed with seeds in the drill box. Starter liquid fertilizer was applied with seed as a 1:1 mix of thiosol and polyphos solution, equivalent to 16.8 kg N ha–1, 11.2 kg P ha–1, and 20.2 kg S ha–1. Plots were harvested using a Hege 140 plot combine (1.5-m header) and by hand from a randomly chosen area of 1 m2, including any seed on the ground. Harvest loss was calculated from the equation [(hand-harvest yield – combined yield)/hand-harvest yield] x 100, with yield expressed on a per unit area basis.
Experiment 3
In 2000, the narrow-leafed lupin cultivars Merrit and Kalya were sown on three dates into winter wheat stubble at the Columbia Basin Agricultural Research Center near Pendleton, using a John Deere 1560 no-till drill. A randomized complete block experimental design with four replications was used with a split-plot restriction on treatment arrangement, with sowing date (21 March, 31 March, and 17 April 2000) as main plots and cultivar as subplots. Subplot size was 3.1 by 22.9 m. A 10% granular ethalfluralin formulation was applied presowing at a rate of 0.84 kg a.i. ha–1 using the GANDY spreader. Lupin was sown at a rate of 168 kg seeds ha–1. Before sowing, lupin seeds were inoculated with Bradyrhizobium spp. (Lupinus) by mixing inoculant slurry with seeds in the drill box. Starter liquid fertilizer was applied at a distance of 13 to 25 mm from the seeds as a 1:1 mix of thiosol (ammonium sulfate) and polyphos (ammonium polyphosphate) solution equivalent to 16.8 kg N ha–1, 11.2 kg P ha–1, and 20.2 kg S ha–1. After sowing, metribuzin was broadcast-applied on the soil surface at a rate of 0.28 kg a.i. ha–1. Plots were harvested with a Hege 140 combine. Grain from combined samples was sampled from each plot and its N content determined using the automated combustion method described by Sheldrick (1986). The N content was multiplied by a factor of 6.25 to obtain crude protein content.
One neutron probe access tube was installed in the center of each subplot, and soil moisture content was measured weekly using a field-calibrated CPN model 503 neutron probe (Campbell Scientific, Nuclear Corp., Pacheco, CA). Probe calibration was described by Payne et al. (2001).
Statistics
All statistical analyses were made using the GLM module of SYSTAT version 10.2 (SYSTAT Software, Inc., Richmond, CA). For Exp. 1, a preliminary analysis of variance was made for the combined data set using site, year, and cultivar as fixed effects in a factorial model. Analysis of variance was then made for individual sites for each year using a randomized complete block design. For Exp. 2 and 3, analysis of variance was made as described by the SYSTAT manual for split-plot designs. This includes different error terms for "between-plot" and "within-plot" effects. Mean separations were made using Tukey's hsd test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
A combined ANOVA revealed that all main effects (site, year, and cultivar) were significant (P < 0.0001) (Table 1). Because all two-way interactive effects (P < 0.0001) and the three-way interactive effect (P < 0.05) were also significant, we present mean yields and mean separation among cultivars for individual sites and years (Table 2).
|
|
Among the narrow-leafed lupin cultivars, yield was greatest for Merrit and least for Chittick at both sites. Yield among narrow-leafed cultivars corresponded directly to earliness of flowering date. At Moro, Merrit flowered on 13 June, Yorrel and Danja on 16 June, and Chittick on 25 June. Flowering dates were nearly identical at Pendleton, indicating that narrow-leafed lupins required only 50 to 60 d to flower. This represents a much shorter vegetative growth period and earlier flowering compared with narrow-leafed lupins grown under Australian conditions. Dracup and Kirby (1996), for example, give a range of 75 to 98 d to flowering for the narrow-leafed lupin cultivar Gungurru when sown on 29 April. In general, this crop will develop and mature faster in the inland PNW than in Australia because of photoperiod response to the longer daylength experienced during the cropping season. Rate of development is hastened by increasing temperature (up to an optimum) and daylength (Dracup and Kirby, 1996). In 1999, when sowing dates were approximately 4 wk earlier than in 1998, yield of white lupin was comparable to that of narrow-leafed lupin (Table 2). At Moro, the white lupin cultivars L2085 and Kiev Mutant had similar yields to narrow-leafed lupin of 1400 kg ha–1, whereas L1047 and Ultra had yields of 1000 kg ha–1. There were no differences among narrow-leafed cultivars at Moro; however, at Pendleton, the narrow-leafed cultivars Chittick and Yorrel had greater yields than Merrit, which appeared to have a poorer stand than the other cultivars. Danja's yield was similar to Merrit's. Seed for cultivars other than Merrit were certified and purchased directly from seed dealers in Australia. We speculate that poorer stands of Merrit in 1999 were due to reduced seed quality caused by an unknown combination of seed damage from harvesting and suboptimal storage conditions (Perry et al., 1998). The white lupin cultivar Kiev Mutant had significantly greater yield than all white and narrow-leafed lupin cultivars in Pendleton except the latest-maturing narrow-leafed cultivar, Chittick.
Experiment 2
Yields from combining were extremely low in this experiment due to pronounced drought at the experimental sites (Fig. 1). Crops failed for the second and third seedings at Lexington and for the third seeding at Condon. We therefore only made ANOVAs for individual sites; seeding date was not used as a factor at Lexington.
|
300 kg ha–1. At the Moro station, which is 65 km from the Condon site, rainfall during this period was 206 mm, and wheat yields were 1200 kg ha–1. The Pendleton station, which normally receives 25 to 50 mm more rain than the Helix site due to its greater elevation, received 396 mm during this period, and wheat yields were about 3600 kg ha–1. There were no statistical interactions between the three experimental factors at any of the sites. Despite drought conditions that ranged from moderate at Helix to disastrous at Lexington, yields showed similar response to cultivar at the three sites and to sowing date at Helix and Condon (Fig. 1). A prominent trend was yield decline with delayed sowing. At the Helix site, for example, when yield of all cultivars was pooled and mean yield divided by the number of days between the first and second sowing, the mean rate of yield decrease was 5.8 kg ha–1 d–1. Additionally, at all sites, the cultivar Tallerack matured later and yielded less than the others. Finally, there was little to no response to increased sowing rate. Sowing rate response was statistically significant at Helix and Lexington, but the increase in yield was so small as to barely recover the additional seed used at the higher rate. It is possible that there would have been a greater yield response to sowing rate under more favorable rainfall conditions. In drought-prone areas of Australia, narrow-leafed lupin is typically sown at rates of 80 to 100 kg ha–1 (Gregory, 1998; Dracup et al., 1998).
Mean harvest loss was 32% at Helix, 24% at Condon, and 62% at Lexington; median harvest loss was 35% at Helix, 54% at Condon, and 66% at Lexington. As severity of drought increased, plant stature decreased, making harvest more difficult. But even for taller plants, the stiff, rigid structure resulted in pod loss and shattering due to shaking by the cutter bar. Additionally, much of the cut material fell off the combine platform. Control of combine harvest loss is an important aspect of lupin management in Australia (Riethmuller and Blanchard, 1995) and is achieved with double-density knife guards to reduce shaking of the plant, extending the harvester platform, and assisted movement of material to the auger using air reels, vibrating mats, or platform sweeps. Preliminary experiments at the Pendleton station suggested that the use of a stripper header (Siemens et al., 2002), which is often used to combine no-till wheat, can also greatly reduce combine harvest loss.
Experiment 3
Overwinter rainfall at Pendleton in 1999–2000 was 379 mm; spring rainfall was 110 mm. Since there was no effect of cultivar on yield, and no interaction between cultivar and sowing date, data for the two cultivars are pooled in Fig. 2. Narrow-leafed lupin yields from the earliest planting were near 1900 kg ha–1 in this experiment, which was the greatest yield we observed for the cultivar Merrit. We believe that the favorable yields were due to a combination of favorable overwinter and especially spring rainfall, which is critical to yield of spring crops such as field pea (Pisum sativum L.) (Payne et al., 2000, 2001), and early planting. Similar to results from Exp. 2, delayed sowing was associated with a strong decrease in yield. Mean yield decreases between dates when all cultivars were pooled were 60 kg ha–1 d–1 when sowing was delayed from 21 to 31 March and 27 kg ha–1 d–1 when delayed from March to 17 April. Decreased protein content was also associated with delayed sowing (Fig. 2). These results confirm the importance of early sowing to lupin yield in the inland PNW and suggest that it is an important determinant of quality as well.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
Based on the highest yields in Fig. 5, obtained with early planting in 2000, our data suggest a yield potential (Evans, 1993) of more than 2000 kg ha–1 and therefore good agronomic potential for narrow-leafed lupin. Our data point strongly to early sowing as the most important agronomic factor to obtaining acceptably high yield, protein content, and plant height because sowing delay is associated with diminishing rainfall, rapidly increasing temperature and evaporative demand, and declining soil moisture storage. Perry et al. (1998) state that time of sowing was a topic of extensive research in the early phase of the narrow-leafed lupin industry in Australia and that lupin yields declined more rapidly than cereals as the sowing date is delayed and the effective growing season reduced. Despite the opposite trends in temperature and daylength during the Australian and inland PNW growing seasons, our data also suggest that sowing at the first opportunity will be a critical management practice.
There are clearly important agronomic challenges to address before narrow-leafed lupin will be widely accepted in the inland PNW. In our view, the two greatest challenges are reducing harvest loss and the availability of an effective herbicide for weed control, particularly for no-till conditions.
While the technology to reduce combine harvest loss exists, an effective herbicide treatment for the inland PNW is still needed. Current environmental regulations in the USA restrict the use of herbicides applied to lupin in Australia and Canada, such as atrazine [6-chloro-N-ethyl-N'-(1-methylethyl)-1,3,5-triazine-2,4-diamine] or simazine (6-chloro-N,N'-diethyl-1,3,5-triazine-2,4-diamine). The lack of a label for an effective herbicide is a major obstacle to successful introduction of narrow-leafed lupin into the inland PNW. Yet its introduction as a replacement of fallow would have expected environmental benefits in terms of increased organic matter input, improved air quality, and reduced runoff.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. T.T. Phan, S. R. Ellwood, K. Adhikari, M. N. Nelson, and R. P. Oliver The First Genetic and Comparative Map of White Lupin (Lupinus albus L.): Identification of QTLs for Anthracnose Resistance and Flowering Time, and a Locus for Alkaloid Content DNA Res, May 26, 2007; (2007) dsm009v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
