Introduction
Edamame or vegetable soybean (Glycine max [L.] Merrill) is a specialty soybean that is harvested as a vegetable when the seeds are still immature (R6 stage). It is usually sold as pods or as whole harvested stems, but occasionally as shelled beans. The seeds are usually boiled in the pod, shelled out, and eaten as a snack, as a vegetable with meals, or added to soups or confections. When eaten as a vegetable, the seeds are added to salads, stir-fried, or combined into vegetable medleys. As a sweet, the beans are ground to a paste, sweetened, and used as a topping for sticky rice. Most production and consumption is in China, Japan, Taiwan, and Korea. Most production research, variety development, and production guides originate in Taiwan and Japan.
Washington State University has conducted edamame variety trials and is involved in a breeding program for improved varieties for the Pacific Northwest. Very little information on production techniques for the coastal areas of the Northwest is available. Information derived from the Asian literature is often inadequate to guide production in our area. For example, the Japanese literature recommends application of 35 to 55 lb N/acre in addition to 8 tons/acre of well-decomposed animal-waste compost, which provides an unknown amount of available nutrients. Recommendations for plant populations are not consistent. If this crop is to be grown in the Pacific Northwest, which appears to have a good climate for production of high yields and quality, we will need to develop recommendations based on the soil types and cultural methods likely to be used in this region.
In our trial in 1998, the cultivars 'Shironomai' and 'Butterbeans' had better yield and quality than did 'White Lion.' Yield of all three cultivars increased with increasing rate of N, with no indication that the highest rate of applied N, 72 lb/acre, was sufficient for maximum yields. Yield per unit area was higher at 20- than at 30-inch between-row spacing (Hemphill 1999).
The purpose of the 1999 trial was to investigate the effects of two between-row spacings and four rates of applied N on the two edamame cultivars that had the higher yields in 1998. In the 2000 trial the objectives were to investigate the interaction of N rate and rhizobium inoculum on yield of edamame and to conduct a yield trial of promising varieties.
Methods
1999
Non-inoculated 'Butterbeans' and 'Shironomai' edamame were seeded to a winter-fallowed Willamette silt loam, pH 5.8, at the NWREC on June 2. Plot preparation included plowing, disking, a broadcast and incorporated application of triple superphosphate and sulfate of potash, each at 200 lb/acre, and harrowing to form a seedbed. The two cultivars were seeded on 20 and 30-inch rows, with 3-inch within-row spacing. Metolachlor (1.5 lb/acre) was applied after planting for weed control. Escaped weeds, mostly lambsquarters, Canada thistle, and red-root pigweed, were controlled by hand-hoeing. Nitrogen rates applied were 0, 36, 72, and 108 lb/acre as ammonium nitrate, with half the N applied 1 week after seeding and the remainder on July 14. The experimental design was a randomized complete-block split plot, with cultivar x spacing combinations as main plots and N rates as subplots and four replications of each treatment combination. Subplot size was 15 x 20 feet. Plots were sprinkler irrigated as necessary, usually 1 inch/week. Yields were estimated by harvest of a 10-foot section of one of the centermost rows of each subplot. Pods were stripped by hand, separated into categories of two or more beans/pod, one bean/pod, and unmarketable (lacking developed beans). Cultivars differed slightly in days to optimal maturity, with 'Shironomai' harvested on September 10 and 'Butterbeans' on September 13.
2000
Both non-inoculated and inoculated 'Butterbeans' edamame were seeded on 20-inch rows to a winter-fallowed Willamette silt loam, pH 5.8, at the NWREC on June 1. Plot preparation included plowing, disking, a broadcast and incorporated application of triple superphosphate and sulfate of potash, each at 200 lb/acre, and harrowing to form a seedbed. Metolachlor (1.5 lb/acre) was applied immediately after planting for weed control. A few escaped weeds, mostly lambsquarters, subterranean clover, and red-root pigweed, were controlled by hand-hoeing. Nitrogen rates applied were 0, 36, 72, and 108 lb/acre as ammonium nitrate, with half the N applied 1 week after seeding and the remainder on July 21. The experimental design was a randomized complete-block split plot, with inoculum as main plot and N rates as subplots. Subplot size was 15 x 20 feet. Plots were sprinkler irrigated as necessary, usually approximately 1 inch/week. SPAD chlorophyll meter readings were taken on the leaves of 10 plants per plot at approximately weekly intervals. Yields were estimated by harvest of a 10-foot section of one of the centermost rows of each subplot on September 12. Pods were harvested as in 1999.
Results
1999 Cultivar
The two cultivars differed significantly in plant development (Table 1). Stand of 'Shironomai' was 27 percent less than that of 'Butterbeans,' very consistent with results from 1998. Canopy height and width did not vary between cultivars. 'Butterbeans' had slightly greener leaves, as measured by the Minolta SPAD meter, on August 12. The two cultivars did not differ in flower development (data not shown).
Despite its relatively poor early stand, 'Shironomai' produced a higher yield/acre, higher yield of marketable pods (2 or more beans), greater mean pod weight, larger pod weight/plant, and larger mean bean weight than did 'Butterbeans' (Table 2). The number of plants harvested did not vary by cultivar. Apparently, late-emerging plants of 'Shironomai' made up for the early deficiency in stand. The number of plants harvested for both cultivars represented about 60 percent of the intended stand.
Although there were significant interactions of cultivar and N rate affecting total yield/plot and weight of 100 pods (Table 3), 'Shironomai' was superior at each N rate. The mean weight of 100 pods and the total weight harvested/plot was highest at 72 lb applied N/acre for 'Butterbeans,' but at 108 lb N/acre for 'Shironomai.' Mean bean weight of 'Butterbeans' was more responsive to rate of applied N than was that of 'Shironomai' (Table 3). Mean bean weight of 'Shironomai' significantly exceeded that of 'Butterbeans' in the absence of applied N, but not in the presence of N fertilizer (Table 3). Cultivar and between-row spacing interacted on marketable pod weight/plant and the weight of 100 marketable pods (Table 4). For weight of 100 pods, 'Shironomai' was the better variety at 20-inch, but not at 30-inch spacing. Marketable pod weight was higher for 'Shironomai' at both spacings. Marketable pod weight/plant was higher at 30-inch spacing for 'Shironomai' but not for 'Butterbeans.'
1999 Spacing
Spacing did not affect canopy size when measured on July 29. Leaf chlorophyll content was slightly higher at the 20-inch spacing when measured on August 12. (Table 1).
The 20-inch spacing, as in 1998, produced a lower yield per foot of row, lower mean pod weight, and lower total and marketable yield/plant than did the 30-inch spacing. However, also as in 1998, the yield per unit area was higher with the 20-inch spacing as the larger number of plants more than offset the higher yield/plant obtained with the wider spacing (Table 2).
Spacing and rate of applied N interacted on yield of marketable and total pods/plant (Table 5). In each case, the yields increased between 72 and 108 lb N/acre for the 20-inch spacing, but not for the 30-inch spacing.
1999 N rate
As in 1998, rate of N had no effect on stand, but canopy height and width increased linearly with increasing N (Table 1). Increasing N had no effect on flowering in 1999 but leaf chlorophyll content increased linearly with increasing rate of applied N.
As in 1998, total weight harvested/plot, yield/unit area, weight of marketable pods/plot, mean pod yield/plant, and mean marketable pod weight/plant increased linearly with increasing rate of N (Table 2). In contrast to 1998, mean weight of the individual pod also increased linearly with increasing applied N. Also in contrast to 1998, the mean bean weight increased quadratically with increasing rate of applied N. The number of beans/pod, the yield of unfilled pods, and the number of plants harvested per plot did not vary with N rate (Table 2). Also as in 1998, there were no 3-way interactions of N rate, spacing, and cultivar affecting plant development or yield.
2000 N rate
As in 1998 and 1999, leaf chlorophyll content, as estimated by SPAD readings, increased linearly with increasing rate of applied N on each of six dates of measurement (Table 6). For the last two measurement dates, presence of inoculum greatly increased SPAD readings at the two lower rates of applied N, but this was not true at the two higher rates of N application (Table 7). This trend also occurred earlier in the season but the interaction of inoculum and N rate was significant only on the last two dates that SPAD readings were taken.
Rate of applied N did not affect the number of plants harvested. In a strong departure from results obtained in 1998 and 1999, yield of marketable pods, yield of pods containing two or more beans, and total weight, averaged over the presence and absence of inoculum, was greatest at only 36 lb applied N/acre (Table 8). However, this yield was not significantly greater than with no applied N. Yield of pods containing only one bean, yield of unmarketable pods, weight of 150 pods, total pod weight per plant, number of beans from 25 pods, and mean bean weight did not vary with N rate. Weight of beans from 25 pods declined linearly with increasing rate of N.
2000 Inoculum
Leaf chlorophyll content did not respond to the presence of inoculum with the seed during much of the growing season, but higher SPAD readings were obtained from inoculated plants late in the growing season. SPAD readings declined in the absence of inoculum (Table 6).
The presence of inoculum reduced the number of plants present at harvest by 15 percent (Table 8). Averaged over N rates, inoculum had no effect on yield of pods containing two or more beans, yield of pods containing one bean, unmarketable yield, total yield, weight of beans from 25 pods, number of beans from 25 pods, marketable yield per plant, and average bean weight. However, the weight of 150 pods and total pod yield per plant increased with inoculum. There were no significant interactions of N rate and inoculum affecting any component of yield.
Discussion and Conclusions
For the purposes of these experiments, marketable yield was defined as those pods acceptable for fresh marketing in-pod. However, pods with only one bean may be marketable as shelled green edamame or as seed for planting. Yields of both one-bean pods and pods with two or more beans responded similarly to treatment.
Given that complete canopy closure was not obtained with the 30-inch spacing, and that mean bean weight and weight of 100 marketable pods was not affected by spacing in 1999 and only slightly reduced at 20 inches in 1998, 20 inches appears to be a reasonable between-row spacing for these cultivars on this Willamette soil.
Although the number of beans/pod, mean bean weight, and total and marketable pod yield/plant did not increase between the two highest rates of N, yield/plant and per unit area did increase between 72 and 108 lb applied N/acre in 1999, indicating that the latter rate is needed for maximum production in a situation where the seed was not inoculated, the soil had not recently been used for production of legumes, and there was not a history of application of manures or composts. In 1998 we noticed a strong tendency for a high proportion of the plants to have nodules at the zero N rate, but not at other N rates. These plants, although dark green in color, did not exceed non-nodulated plants in size. This was not the case in 1999. Inoculated seed might have responded differently to the N rate.
We have no explanation for the failure of the crop yield to respond to N rate in 2000 as it had in 1998 and 1999. SPAD readings indicated a large response to applied N. Plant biomass was not measured but perhaps the extra N resulted in larger plants but did not increase the set of pods that reached marketable size before harvest. High rates of N also delay maturity and it may be that a later harvest date would have resulted in higher yields at the higher rates of N. Inoculum was expected to increase yield at low rates of applied N but not a the higher rates. However, this interaction did not occur as the best yields were obtained with zero or the low rate of applied N, regardless of the presence or absence of inoculum.
Literature Cited
Hemphill, D.D., Jr. 1999. Vegetable research at the North Willamette Research and Extension Center, 1997-1998. Oregon Agricultural Experiment Station Special Report No. 1000, Oregon State University, Corvallis.
Table 1. Main effects of cultivar, between-row spacing, and rate of applied N on stand, plant height and width, and flower development in edamame, NWREC, 1999. Seedlingsz/ Canopy heighty Canopy widthy SPAD Units 20 ft inches inches 29 July 12 Aug. Cultivar Butterbeans 44 14 15 33 33 Shironomai 32 14 15 32 32 Significance *** NS NS NS * Spacing 20 inches 38 14 15 33 33 30 inches 38 14 15 32 32 Significance NS NS NS NS * N rate (lb/acre) 0 37 12 13 27 27 36 39 13 15 32 30 72 37 14 16 34 34 108 39 15 16 36 37 Significance NS L** L*** L*** L*** zCounts made 17 June. yMeasured 29 July. ***,**,*,NSSignificant at 0.1, 1.0, and 5.0 percent probability level and nonsignificant, respectively. L = linear. Table 2. Main effects of cultivar, between-row spacing, and rate of applied N fertilizer on edamame yields, NWREC, 1999. Total wt Unfilled 1 bean/ 2+ beans/ Wt 100 No. of Mean bean No. of Pod wt./ Marketable (g/ (tons/ pods pod pods pods beans/ wt plants/ plant pods/plant plot) acre) (g/plot) (g/plot) (g/plot) (g) 25 pods (g) 3 m (g) (g) Cultivar Butterbeans 1,198 2.8 55 403 740 194 55 0.44 24.6 49 31 Shironomai 1,648 3.8 57 513 1,078 263 54 0.51 24.0 70 46 Significance *** *** NS ** *** *** NS ** NS *** *** Spacing 20 inches 1,337 3.8 47 414 876 227 54 0.48 24.9 54 36 30 inches 1,509 2.9 66 501 942 230 55 0.47 23.7 65 41 Significance ** ** NS * NS NS NS NS NS ** * N Rate, lb/acre 0 880 2.0 49 318 512 204 55 0.41 24.4 37 22 36 1,350 3.1 56 402 892 223 54 0.45 24.4 56 37 72 1,683 3.9 67 526 1,089 240 54 0.53 23.7 73 47 108 1,780 4.1 52 585 1,143 247 55 0.52 24.8 73 46 Significance L*** L*** NS L*** L*** L*** NS L***Q* NS *** *** ***,**,*,NSSignificant at 0.1, 1.0, and 5.0 percent probability level and nonsignificant, respectively. L = linear, Q = quadratic. Table 3. Interaction of rate of applied N and cultivar, averaged over two spacings, on weight of 100 marketable pods, total pod weight/plot, and mean bean weight, NWREC, 1999. Applied N Cultivar Wt of 100 pods Total wt/3 m Mean bean wt (lb/acre) (g) (g) (g) 0 Butterbeans 161 772 0.32 Shironomai 248 988 0.50 36 Butterbeans 187 1,177 0.44 Shironomai 259 1,523 0.46 72 Butterbeans 222 1,456 0.53 Shironomai 259 1,910 0.52 108 Butterbeans 207 1,386 0.49 Shironomai 286 2,173 0.55 LSD (0.05), Significance 29* 219** 0.09* **,*Significant at 5.0 and 1.0 percent probability levels, respectively. Table 4. Interaction of between-row spacing and cultivar, averaged over four rates of applied N, on the weight of 100 marketable edamame pods and marketable yield per plant, NWREC, 1999. Row spacing, inches Cultivar Wt of 100 pods Marketable pod wt/plant ------------------g--------------------- 20 Butterbeans 185 30 Shironomai 203 41 30 Butterbeans 270 31 Shironomai 256 51 LSD (0.05), Significance 18* 7* *Significant at 5 percent level. Table 5. Interaction of between-row spacing and rate of applied N, averaged over two cultivars, on yield of marketable pods and total pods of edamame, NWREC, 1999. Spacing (inches) N rate (lb/acre) Marketable Total ------g/3 m------ 20 0 426 722 36 827 1,249 72 990 1,558 108 1,262 1,819 30 0 599 1,038 36 957 1,450 72 1,188 1,808 108 1,025 1,741 LSD (0.05) Significance 200** 219* **,*Significant at 5 and 1 percent levels, respectively. Table 6. Main effects of inoculum and rate of applied N on SPAD readings in edamame, NWREC, 2000. 27 July 3 Aug. 10 Aug. 17 Aug. 28 Aug. 11 Sept. Inoculum Without 31 30 31 29 29 21 With 31 32 32 31 33 30 Significance NS NS NS NS *** *** N rate (lb/acre) 0 27 27 28 27 27 18 36 32 30 31 30 30 24 72 33 32 32 31 32 28 108 34 35 36 33 35 33 Significance L*** L*** L*** L** L*** L*** ***,**,NSSignificant at 0.1 and 1.0 percent probability level and nonsignificant, respectively; L = linear. Table 7. Interaction of inoculum and rate of applied N on SPAD readings in edamame on two dates, NWREC, 2000. Date Inoculum Rate of applied N (lb/acre) 0 36 72 108 28 Aug. Without 22 28 31 35 With 33 32 32 36 11 Sept. Without 10 19 25 32 With 25 30 32 34 LSD (0.05) for any two means = 4 for 28 August and 8 for 11 September. Table 8. Main effects of inoculum and rate of applied N fertilizer on edamame yields, NWREC, 2000. Total Unfilled 1 bean/ 2+ beans/ Wt 150 No. of Mean bean No. of Pod wt/ Marketable wt pods pod pod pods beans/ wt plants/ plant pods/plant (g/plot) (g/plot) (g/plot) (g/plot) (g) 25 pods (g) 3 m (g) (g) Inoculum Without 892 122 268 502 343 54 0.52 23.6 38 33 With 917 140 280 497 366 56 0.50 20.2 47 40 Significance NS NS NS NS * NS NS * * * N Rate (lb/acre) 0 957 116 267 573 338 57 0.52 21.0 46 41 36 1,041 121 314 606 359 56 0.52 21.6 50 43 72 829 151 253 425 364 56 0.49 22.8 38 32 108 790 136 262 392 358 51 0.50 22.3 36 30 LSD (0.05) 152 NS NS 107 NS NS NS NS NS * *,NSSignificant at 5 percent probability level and nonsignificant, respectively.