Introduction
Pickling cucumber growers and processors are interested in mechanical harvest of fruit to avoid problems with labor cost and availability. Mechanical harvest requires changes in varieties, number of plants/acre, and cultural practices. We have been conducting trials since 1989 to determine the feasibility of machine harvest in the Willamette Valley. This report focuses on the two most recent growing seasons. Our 1993 and 1994 trials included a number of varieties screened in earlier trials as well as some varieties not previously included, and on the effect of harvest date on maturity.
Varieties thought to have a good length/diameter ratio and processing quality were grown in commercial-scale trials. Each variety was harvested more than once to determine days or heat units to the optimum size distribution and the rate at which size grade increases with time. Pickling cucumbers are graded into four sizes by diameter (See Table 1), with the smaller sizes of greater value to the processor. Growers are paid considerably more per pound for the smaller sizes. "Size grade" is determined from the proportion by weight of fruit falling into each of the four size categories. As a field matures, the tonnage and size grade increase, while the value per ton decreases. Harvest and pricing policy must be adjusted both to fit processor pack needs and to provide adequate return for both processor and grower.
Weeds reduce yield and harvester efficiency. High weed populations may cause expensive downtime for cleaning the harvester pickup head. Chloramben (Amiben) and naptalam (Alanap) have been withdrawn from the market. Other herbicides have proven only partially effective or damaging to the crop. Curbit has found a niche with some growers but inconsistency in weed control and crop injury limits its use. Testing of alternatives has been conducted to identify combinations of herbicides that are effective and non-phytotoxic to the crop. The 1991 and 1992 trials indicated that clomazone (Command) shows promise. Weed control was excellent but there was some damage to the cucumber seedlings in 1991. It was apparent that more information on crop response to clomazone, alone or in combination with other materials, was still needed. A 1993 trial indicated insignificant phytotoxicity problems, but the field was so weed-free that no useful information on efficacy was obtained. This trial was repeated in 1994.
Methods
The variety trials were established in a commercial planting just south of Woodburn, Oregon. The previous crops were oats (1993) and sugar beets (1994). Planting date was 8 June, 1993, and 16 May, 1994, following typical land preparation by the grower. Each variety was planted to an area of slightly less than 1.2 acres. The area for each variety was divided into two approximately equal and separate replicates. The planter seeded six rows at a time on 30-inch centers. Planters were not adjusted for the slightly different seed size of each variety; thus, seeding rate varied somewhat with variety. The goal was to produce a stand of about 80,000 plants/acre.
The entire area received a broadcast, pre-plant application of 40 pounds N and 42 pounds K/acre and a banded application of 60 pounds N and 78 pounds P/acre at planting, for a total N application of 100 pounds/acre. This rate appears from our previous work to be optimal for machine-harvested cucumbers. Irrigation, cultivation, and herbicide application were provided by the grower.
We hand-harvested all varieties three times between 6 and 10 August, 1993, and 18 and 22 July, 1994, using a once-over destructive harvest for each 90-square foot plot. The fruit was weighed and graded on a Kerian 'SpeedSizer' mechanical grader provided by Nalley Fine Foods. The commercial plots of all varieties were harvested by Nalley's personnel using an FMC harvester on 10 August, 1993, and 22 July, 1994. A Byron harvester was also used for one variety in 1993. The machine-harvested samples were weighed and graded at the Nalley receiving station at Cornelius.
The 1994 weed control plots were located at the North Willamette Research and Extension Center to take advantage of the better control it offers for managing operations and a greater weed population. The first of two plantings ('Flurry-M') was direct-seeded with Massey-Ferguson International planters on 24 May, 1994. The herbicide treatments were selected to verify results from the previous year and to establish the lower limit at which Command would be effective when applied alone and in combination with Prefar. The preplant treatments were applied one day before planting with a CO2 backpack sprayer (40 psi) to a dry soil surface and incorporated to 3-inch depth with a PTO-driven power tiller. The pre-emergence treatments were applied immediately after planting and the entire area was then irrigated with approximately 1 inch of water.
The trial design was a randomized complete block with four replications. Four rows were planted to a 12 x 30 foot plot. Weed control ratings represent the mean rating of the three project leaders, each rating independently of the others. Plots were harvested on 3 August.
Commercial (Machine-Harvested) Variety Trial, 1993
It is unfortunate that the harvester could not be scheduled into the variety trial on more than one day. Since processors want to harvest at a mean size grade near 2.5, it is apparent that 'Duke', and to a lesser extent 'Flurry-M', 'Napoleon', and 'Quest', were harvested between one and two days after peak maturity (Table 1). The later-maturing 'Lafayette' and the Sun line were harvested at optimal maturity.
Fortunately, the Byron and FMC harvesters harvested several acres of 'Flurry-M' in the same field on 6 August. This gave a basis for comparison with the hand-picked samples on the same day, and a comparison of the two harvesters. The average grade harvested was 2.35, compared to a hand-picked grade of 2.10 for the same date. However, the FMC harvester picked a mean grade of 2.27 (dollar value of $155 per ton) compared to 2.43 for the Byron harvester (dollar value of $134 per ton). Since the two machines harvested different areas of the field, we cannot be sure whether the differences are due to the machine efficiency or to differences in the maturity of the crop in the different areas. However, the fact that the FMC picked 9.2 percent grade 1 fruit, while the Byron picked only 3.7 percent grade 1, leads us to speculate that the difference lies in harvester efficiency at recovery of small fruit. We were unable to calculate dollar value per acre since the harvested areas for each machine were not reported.
It is obvious from comparing the machine-harvest and hand-picked harvests (Table 2) of the same varieties that machine-harvest led to higher mean grade for all varieties except the SunSeeds line, perhaps due to poorer recovery of small fruit. The average return per acre varied from $729 for 'Duke' to $953 for the SunSeeds line (Table 1). Yields were generally comparable to those in 1992 and less than those of 1991.
Hand-Picked Variety Trial, 1993
We simulated a machine harvest three times over a five-day period (Table 2). 'Duke' was first to reach a mean grade of 2.5 in both the hand-picked and mechanically-harvested plots. 'Napoleon' and 'Flurry-M' were the next varieties to mature, while the SunSeeds line and 'Quest' were the last to mature. Given the better than 0.2 change in grade per day, it may be useful to lump 'Duke', Flurry-M', and 'Napoleon' together as the early maturing varieties, with 'Lafayette', 'Quest', and Sun 3539 about one day later. With unusually hot weather at maturity, there was certainly no more than a 48-hour spread between the earliest and latest varieties. 'Napoleon', and to a lesser extent 'Sun 3539' and 'Flurry-M', were slower to increase in size grade over the harvest period than were the other varieties. This is inconsistent with 1992 when 'Lafayette' was by far the slowest to change size grade. The greatest return per acre, at the harvest nearest optimal maturity, was $1347 for the Sun line and $1346 for 'Quest', but 'Flurry-M' and 'Lafayette' also grossed in excess of $1000 per acre. 'Quest' and 'Lafayette' also performed well in the 1992 hand-picked trial.
The apparent decrease in grade for several varieties between the 9 and 10 August harvests is an artifact. The latter harvest represented only one replicate in the generally weaker and less mature west end of the field. The high percentage of nubs and crooks in this harvest also indicates weakness in this area, perhaps due to low nitrogen status.
Commercial (Machine-Harvested) Variety Trial, 1994
Stands of all varieties except 'Calypso' and 'Bradley' neared or exceeded target population and were greater than the last two years (Table 3). We experienced planter problems with the 'Calypso' pass. 'Bradley' emerged two days later than other varieties and had low seedling vigor. At the time stand counts were taken, not all 'Bradley' seed may have germinated. The lower-than-average yield for 'Bradley' may reflect the fact that it was harvested at low mean size grade, rather than any deficiency in stand.
It is unfortunate that the harvester could not be scheduled into the variety trial on more than one day. It is apparent that 'Bradley,' 'Flurry,' and the Harris-Moran line were harvested a day early. One goal in 1994 was to obtain a high percentage of size grade #3 fruit, suitable for making spears. The entire harvest was probably about a day early for this goal. But with an average size grade for the 10 varieties of 2.44, the cucumbers harvested were of excellent quality.
In comparing the machine- and hand-picked harvests (Tables 3 and 4) of the same varieties, machine-harvest produced a lower mean size grade for four varieties, essentially the same size grade for four varieties, and a greater mean grade for only 'Calypso' and 'Neptune,' indicating excellent recovery of small fruit. In contrast, in 1993 the FMC machine harvested a higher mean size grade than did hand-picking.
The average return per acre varied from $854 for 'Calypso' to $1859 for 'Neptune.' Yields were generally a little less than those in 1991, at about the same maturity, and averaged 2 to 3 tons/acre greater than in 1992 or 1993 (mean across varieties of 10.6 tons/acre in 1994, 12.4 in 1991, 8.0 in 1992, and 7.7 in 1993). The only variety in all four trials was 'Flurry-M'. In 1991 it yielded 11.8 tons/acre, in 1992 only 3.1 tons/acre at similar mean grade. The 1993 yield was 7.1 tons/acre, but the fruit was larger than in 1992. In 1994, at mean grade of only 2.36, 'Flurry' yielded 9.9 tons/acre.
The yields from the hand-picked plots were consistent with this trend. Average hand-picked yield for all varieties declined from 15.1 tons/acre in 1991 to 9.2 tons/acre in 1992 and 8.4 tons/acre when picked at nearly optimal size grade in 1993. Mean hand-picked yield in 1994 rebounded to 11.0 tons/acre, at maturity of 2.5 or less. Hand-picked 'Flurry-M' declined from 8.6 tons/acre in 1991 to 5.7 tons/acre in 1992, but increased to 9.4 tons/acre in 1993 and 9.1 tons/acre in 1994. In addition to weather, other factors may have contributed to the yield differences among years. Stands were greater in 1991 and 1994 than in the other years. The plant population in 1992 and 1993 was lower than the target of 80,000/acre. Specifically for 'Flurry-M', the population was estimated at 65,300/acre in 1991, 55,900 in 1992, 62,000 in 1993, and 94,000 in 1994.
The amount of reject material in loads delivered from the test plots was very low in both years, indicating good weed control and excellent harvester efficiency at expelling dirt and vines. Exceptions were 'Flurry,' which was harvested after dark and in a rush to finish up, and 'Bradley,' a small-vined variety.
Hand-Picked Variety Trial, 1994
The ten varieties were hand-harvested three times over a five-day period (Table 4). 'Atlantis,' 'Discover,' 'Excell,' and 'Lafayette' reached a mean grade of 2.5 by the final hand-harvest, taken only three hours before machine-harvest commenced. With unusually hot weather at maturity, there was certainly no more than a 48-hour spread between the earliest and latest varieties. There were again considerable differences among varieties in the rate of increase in size grade over the harvest period (Figure 1). 'Atlantis' and 'Bradley' increased more than 0.26 mean grade/day, contrasted with only 0.157 for 'HMX 1463.' The varieties 'Flurry' and 'Lafayette' have been grown in each of the last three years. 'Lafayette' increased more rapidly in mean grade than did Flurry in 1993 and 1994. The rate of change was greater in 1994 for both varieties, attributable to the extremely high temperatures at harvest. However, relative rate of increase in grade may not be consistent over all seasons. In 1992, 'Lafayette' was considerably slower to size than was 'Flurry.' Several years of experience and correlation with environmental variables would be necessary to determine whether varietal differences in rate of fruit size increase are predictable.
The greatest return per acre, at the harvest nearest optimal maturity, was $1739 for 'Excell', $1657 for 'Lafayette', and $1656 for Neptune. All varieties exceeded $1100 per acre in the hand-picked trial.
It is apparent when comparing the percentage of nubs and crooks recorded for the last hand-picking with the percentages recorded at the receiving station for the machine-picked fruit, that either the harvester expels many nubs and crooks, or we grade to a higher standard than do the receiving station personnel. Our percentage of nubs and crooks was, on the average, about three times what the receiving station recorded. Nevertheless, most results are consistent between our rating of hand-picked fruit and the grading at the receiving station. Both found 'Flurry' and 'FMX 4841' to have a greater than average nubs plus crooks percentage and 'Bradley' and 'HMX 1463' to have a very low percentage.
Since all hand-harvests were near or below optimal maturity, we were not able to distinguish among varieties in the rate at which they lose dollar value per acre when they pass optimal maturity. All varieties except 'Discover' continued to increase in dollar return per acre between the second and third harvests.
Modeling Cucumber Development in Response to Temperatures
We previously concluded that the number of accumulated heat units (North Carolina State-Washington State model; sum of Tmax-15.5 oC with a 32 oC cutoff) required to reach maturity for a given variety increased with later plantings, indicating that maturity is not a linear function of heat unit accumulation in the Willamette Valley. Results in 1991 through 1994 were consistent with this conclusion. In 1991, it took only 583 AHU to surpass optimal maturity for the variety 'Calypso' with a 5 June planting date. However, in 1990, 690 AHU were required for 'Calypso' to reach a mean grade of 2.5 from a 3 July planting date. In 1994, with a 16 May planting date, only 552 AHU were required to bring 'Calypso' to a mean grade of 2.57 (machine-harvest).
From 1991 through 1994 data on 'Flurry-M', it can also be concluded that the model is not consistent between seasons with different weather patterns. In 1991, with a relatively cool, wet June, it took about 540 AHU to mature 'Flurry-M'. In 1992, however, with unusually high temperatures in June, it took 600. In 1993, with an unusually cool and wet June and July, but with very hot weather near harvest, 'Flurry-M' needed only 501 heat units to reach optimal maturity. In 1994, temperatures were near long-term averages from planting until just a few days before harvest. 'Flurry-M' had not reached a mean grade of 2.5 with 552 AHU. Apparently, vine development, flowering, and fruit development in 1993 occurred normally, despite the very cool weather, and the warmer weather in 1994 did not advance maturity.
It is now apparent that, even with the high temperature cutoff of 32 oC, the North Carolina State-Washington State model gives too much weight to above-normal temperatures at the pre-bloom stage of plant development. Furthermore, the model cannot anticipate any delay in pollination and fruit set that may result from inadequate bee activity at first bloom. It is also our belief that the base temperature of 15.5 oC used in the model is too high. Use of a 10 oC base, and the mean rather than the maximum temperature, along with a penalty for cold weather, may provide a superior model of plant response to temperature.
Weed Control Trial
The high rate of Command, 0.25 lb active/acre, applied pre-emergence produced yields significantly greater than most other treatments and was the only treatment to outyield the hand-weeded control (Table 5). Even at the lower rate, pre-emergence Command was an excellent treatment. Pre-plant incorporation of Command resulted in significantly lower yields and reduced weed control compared to the same rates of Command applied pre-emergence. The high rate of Command, pre-emergence, reduced stands slightly compared to most other treatments in both trials, but yield was not adversely affected. The combination of Command with either Prefar or Alanap did not improve weed control or yield compared to Command alone. The combination of Curbit and Alanap was superior to Curbit alone for both weed control and yield. Registration of Command would provide a valuable tool for growers, particularly in view of its low cost.
Table 1. Nalley's-OSU pickle variety trial, 1993 machine-harvest results Variety %1s %2s %3s %4s %N&C %Rej Tot lb. Grade $/T $/A T/A Duke 2.7z 13.4 55.6 19.7 7.8 0.8 20400 3.01 85 729 8.6 Flurry-M 3.0 28.1 55.5 8.3 4.7 0.5 17160 2.73 110 775 7.1 Lafayette 4.8 33.2 53.8 3.2 4.2 0.8 18130 2.58 123 949 7.7 Napoleon 6.4 20.8 47.6 22.2 2.5 0.6 19140 2.88 101 813 8.1 Quest 2.4 27.4 62.8 2.7 4.7 0.2 17800 2.70 113 849 7.5 Sunex 3539 3.7 44.8 46.7 1.8 2.8 0.1 17040 2.48 133 953 7.2 zSize categories are as follows: 1) 0.5-1.0 inch diameter, 2) 1.0-1.5 inch, 3) 1.5-2.0 inch, 4) over 2.0 inch. N&C=nubs plus crooks. Rej=reject material, mainly dirt clods, stems, and roots. Grade is the weighted average of the percentage in each size category, excluding N&C. Table 2. Nalley's-OSU pickling cucumber variety trial, 1993 hand-picked harvest results Variety Harvest date %1s %2s %3s %4s %N&C Grade T/acre $/T $/acre Duke Aug. 6z 18.7 54.7 19.6 1.0 6.0 2.03 3.9 185 715 Aug. 9 6.4 21.8 48.0 16.7 7.0 2.81 8.3 105 871 Aug. 10 4.3 18.6 48.2 8.2 20.7 2.72 8.4 98 826 Flurry-M Aug. 6 15.7 55.3 23.0 1.2 4.8 2.10 4.6 179 818 Aug. 9 3.7 25.6 54.4 8.4 7.9 2.73 9.4 108 1020 Aug. 10 3.2 29.1 45.1 9.0 13.5 2.69 9.7 107 1039 Lafayette Aug. 6 26.2 65.9 5.2 0.0 2.7 1.79 3.8 213 798 Aug. 9 4.7 39.9 44.4 3.6 7.4 2.51 9.2 128 1175 Aug. 10 9.8 30.2 47.4 3.5 9.1 2.49 6.7 132 879 Napoleon Aug. 6 12.3 55.1 27.2 3.6 1.7 2.23 3.7 176 648 Aug. 9 4.6 21.8 61.7 5.8 6.1 2.73 7.5 111 833 Aug. 10 8.7 20.6 56.4 7.4 7.0 2.67 5.7 118 677 Quest Aug. 6 26.1 64.1 6.4 0.0 3.4 1.80 3.3 214 706 Aug. 9 4.7 37.3 51.2 1.6 5.2 2.52 9.3 128 1188 Aug. 10 3.8 39.4 47.0 0.8 9.1 2.49 10.6 126 1346 Sunex 3539 Aug. 6 27.8 51.1 18.0 0.0 3.1 1.90 3.7 212 785 Aug. 9 5.2 43.5 44.3 4.2 2.8 2.49 8.5 134 1134 Aug. 10 3.5 42.9 45.7 2.4 5.5 2.50 10.4 129 1347 zHeat units accumulated for each date: Aug. 6, 457; Aug 9, 501; Aug. 10, 510. Accumulated heat units = Sum (Tmax-15.5 oC) with a 32 oC cutoff. Table 3. Nalley's-OSU pickling cucumber variety trial, 1994 machine-harvest results. Variety Plants/A %1s %2s %3s %4s %N&C %Rej lb. Acres Grade $/T $/A T/A Atlantis 101,000 5.1 40.3 49.2 0.8 3.4 1.1 8280 0.30 2.48 133 1859 14.0 Bradley 52,000 6.9 55.1 32.3 1.7 1.4 2.6 8560 0.51 2.30 146 1222 8.4 Calypso 65,000 4.5 36.3 50.3 4.0 4.0 0.9 8220 0.60 2.57 126 854 6.8 Discover 125,000 5.3 40.5 47.7 1.4 4.0 1.1 21425 1.16 2.48 133 1227 9.3 Excell 99,000 4.5 38.8 49.5 2.1 3.9 1.2 28560 1.16 2.52 129 1588 12.3 Flurry-M 94,000 6.5 43.8 37.3 0.6 7.6 4.2 2895 0.15 2.36 133 1319 9.9 FMX 4841 80,000 5.6 44.1 43.1 1.5 4.2 1.5 6815 0.29 2.43 135 1576 11.7 HMX 1463 72,000 6.1 59.7 30.9 0.2 1.7 1.4 13165 0.61 2.26 149 1609 10.8 Lafayette 83,000 5.8 44.8 44.4 0.8 3.3 0.8 26240 1.21 2.42 138 1489 10.8 Neptune 103,000 7.9 42.3 43.4 1.9 2.5 2.1 8840 0.37 2.41 140 1678 12.0 Table 4. Nalley's-OSU pickling cucumber variety trial, 1994 hand-picked harvest Variety Harvest date %1s %2s %3s %4s %N&C Grade T/acre $/T $/acre Atlantis July 18z 51.5 41.0 0.3 0.0 5.0 1.48 3.2 249 794 21 10.7 52.7 22.3 0.0 14.2 2.14 8.4 150 1255 22 5.4 33.0 46.0 3.3 12.3 2.54 11.8 122 1438 Bradley July 18 80.0 20.0 0.0 0.0 0.0 1.20 1.7 309 526 21 16.2 63.9 9.0 0.0 10.8 1.92 7.1 173 1230 22 6.8 47.0 35.3 0.0 10.9 2.32 10.5 137 1445 Calypso July 18 59.7 30.7 3.9 0.0 5.7 1.41 1.9 262 495 21 9.1 44.4 32.3 0.0 14.2 2.27 6.9 140 975 22 6.5 40.7 44.7 0.0 8.1 2.42 8.6 136 1170 Discover July 18 40.6 50.3 0.0 0.0 9.1 1.55 3.0 225 667 21 10.2 50.3 28.2 1.5 9.8 2.23 9.1 148 1353 22 5.0 28.7 45.6 2.2 18.5 2.55 11.8 115 1353 Excell July 18 30.5 53.0 5.3 0.0 11.2 1.72 4.8 200 964 21 7.1 43.3 37.6 1.4 10.6 2.37 10.2 136 1388 22 4.4 27.6 56.5 3.2 8.3 2.64 14.7 118 1739 Flurry-M July 18 49.4 37.8 3.6 0.0 9.3 1.50 2.5 239 602 21 10.7 42.6 26.8 2.3 17.6 2.25 6.2 138 856 22 8.0 37.3 35.4 1.5 17.8 2.37 9.1 129 1175 FMX 4841 July 18 34.6 43.0 4.2 0.0 18.2 1.63 3.4 199 688 21 4.4 50.3 21.1 0.0 24.1 2.22 7.5 126 941 22 3.3 43.5 37.7 1.8 13.7 2.44 9.0 124 1116 HMX 1463 July 18 46.2 42.3 8.2 0.0 3.3 1.61 3.5 238 837 21 7.9 70.6 16.4 4.0 1.1 2.17 9.0 159 1427 22 8.3 57.0 31.2 0.0 3.5 2.24 9.7 153 1483 Lafayette July 18 31.9 55.2 2.3 0.0 10.5 1.67 2.9 206 608 21 7.6 48.2 35.3 0.0 8.9 2.30 8.1 142 1151 22 3.7 30.2 52.7 3.2 10.2 2.62 14.1 117 1657 Neptune July 18 54.0 39.7 0.0 0.0 6.3 1.42 2.8 253 709 21 15.4 48.7 21.7 0.0 14.2 2.07 7.1 159 1130 July 22 11.1 43.5 38.1 0.0 7.3 2.29 11.1 149 1656 zHeat units accumulated for each date were: July 18, 489; July 21, 536; July 22, 552. Accumulated heat units = Sum (Tmax-15.5 oC) with a 32 oC cutoff. Table 5. 1994 pickling cucumber herbicide trial Treatment Rate, lb Yield Stand count/ Weed density Weed controlx Acre cost of ai/acre kg/plot 3 row feet per sq. ft. rating chemicals Command PREz 0.125 16.5 34 6.5 8.5 $2.84 Command PRE 0.25 19.8 26 4.5 9.4 $5.68 Command PPIy 0.125 9.0 33 19.3 4.3 $2.84 Command PPI 0.25 14.8 33 13.0 5.7 $5.68 Command PPI 0.125 13.8 34 39.8 3.6 $64.14 Prefar 6.0 Command PPI 0.125 13.5 32 8.5 7.0 $56.94 Alanap 4.0 Curbit PRE 1.5 11.8 29 9.5 5.7 $30.58 Curbit PRE 1.5 15.3 33 4.0 9.0 $84.68 Alanap 4.0 Alanap PPI 4.0 17.4 33 9.3 7.2 $115.40 Prefar 6.0 Handweeded --- 13.7 32 21.3 1.2 ---- Untreated --- 3.6 34 23.0 1.2 ---- LSD 0.05 4.6 4 14.7 zPRE: pre-emergence, incorporated with water. yPPI: pre-plant incorporated with rotary tillage. x0=no control, 10=complete control.