Introduction
Broccoli growers in the Willamette Valley use high rates of nitrogen fertilizers, often exceeding 250 to 300 pounds actual N/acre per season. While the common experience has been that these rates are necessary to achieve maximum yields and quality, a considerable portion of the applied fertilizer may not be taken up by the crop. This has raised concerns that the remaining N may be contributing to nitrate pollution of groundwater. Improved efficiency of nitrogen management in broccoli may be possible if the fertilizer could be applied at the time of maximum crop need and placed for maximum contact with the root system. Trials conducted in 1989 and 1990 indicated that 250 pounds N/acre are needed for maximum yields, and that N source (potassium nitrate, sodium nitrate, calcium nitrate, urea, ammonium nitrate, ammonium sulfate) has very little effect on crop yield and nitrogen uptake. The purpose of the 1991 and 1992 trials was to study the response of broccoli yield to a wide range of rates of nitrogen as well as different methods of fertilizer placement and timing of the sidedressed applications. In addition, plots fertilized at 250 pounds per acre were used to determine the course of nitrogen uptake over time.
Methods
In 1991, 'Gem' broccoli was direct-seeded in a Willamette silt loam, pH 5.7, at the NWREC on 12 June. Plot preparation included a broadcast and incorporated application of triple superphosphate at 200 pounds/acre, potassium sulfate at 240 pounds/acre, boron at 2.0 pounds/acre, trifluralin at 0.75 pounds/acre, and chlorpyrifos at 1.3 pounds/acre. Fifty pounds N/acre as urea were shanked in at 2 inches beneath and 2 inches to the side of the seed row on the appropriate plots. All other plots except a zero-nitrogen check received a broadcast application of 50 pounds N/acre on 13 June, followed immediately by an irrigation of 0.5 inches. Plot size was 15 feet with four rows on an 80-inch bed. Spacing between rows alternated between 12 and 28 inches. The seedlings were thinned to approximately 10 inches between plants in the row on 26 June. The plots were sprinkler-irrigated as necessary and harvested on 21 and 28 August, and 4 September. Only main shoots were cut.
Urea was the sole nitrogen source. Of the total of 13 treatments that received N fertilizer, the remainder of the nitrogen was applied on 16 July for 12 treatments. For the remaining treatment, half the remaining N was applied on 9 July, the other half on 29 July. Treatments consisted of various combinations of fertilizer rates, and methods and timing of fertilizer placement (Table 1). In addition, two sets of plots were reserved for application of 15N-enriched urea to monitor N uptake and movement of applied N through the soil profile. Treatments were in randomized complete block design with four replications.
Following completion of the last harvest and soil sampling for residual N, the four blocks were split and half of each block seeded to 'Wheeler' cereal rye on 17 Sept. Broccoli residue was left standing.
In 1992, methods were similar except as follows. The seeding date was 9 June. Plot size was 20 feet with four rows on an 80-inch bed. Spacing between rows was 16 inches. The initial application of nitrogen was broadcast immediately after seeding and irrigated in. Depending on treatment, all N was applied at planting; 40 pounds per acre were applied at planting and the remainder on 14 July; or 40 pounds were applied at planting, half the remainder on 7 July, and the final application on 27 July (Table 8). The plots were harvested on 18 and 25 August.
The cereal rye cover crop was seeded on 3 Sept. after mowing the broccoli and lightly disking the soil surface.
Results and Discussion
1991
When all N fertilizer was broadcast, yield increased with increasing rate of N to a maximum at 250 pounds N/acre (Table 2). Not all N rates were included when the second N application was banded between the paired 12-inch rows or when the first N application was shanked-in rather than broadcast. However, the trend in yield response to increasing rate of N was similar to the yield response when all N was broadcast (Tables 3 and 4). The greatest yield was obtained when the sidedressed N was split between two applications at four and seven weeks after planting rather than a single application at five weeks (Trt. 11, Table 2). However, the yield with the split application was not significantly greater than the yield with a single sidedressing of N at the same rate. These results are consistent with those obtained in 1990, when maximum yield occurred at 250 pounds N, regardless of source of the applied N. Rates of N in excess of 250 pounds do not appear to increase potential return to the grower.
Method of N placement did not significantly affect yield in this trial. Banding the sidedressed N between the closely-spaced paired rows was no more efficent than spinning the N over the entire area at this plant population (31,360/acre) and row spacing (Table 3). Apparently, fertilizer landing in the 28-inch space between paired rows was effectively utilized. Alternatively, concentrating the N between the paired rows may have resulted in excessive N in the band or exposure of too small a fraction of the root system to the available N. Likewise, shanking the initial fertilizer application in a band 2 inches to the side and 2 inches beneath the seed line did not increase efficiency of utilization as reflected in mean head weight (Table 4). Also, there was a single comparison of shanked at-planting fertilizer and banded sidedressed fertilizer at a total N rate of 100 pounds/acre (Treatment 4). This treatment produced an average head weight for the season of 133 g, less than that obtained for all-broadcast N at 100 pounds (141 g) or the shanked-broadcast combination at 100 pounds N/acre (136 g).
Thus, it appears that at high plant populations and small between-row spacings, broadcast applications are as effective as any other means of application. Irrigation in this trial was controlled to prevent movement of nitrogen out of the root zone. A total of 10 inches of water was applied and there was no significant rainfall. Wells (vacuum lysimeters) were installed in the 15N plots. There was no movement of water below the two-foot depth. A comparison of pre-plant and post-harvest soil samples taken down to the 40-inch depth indicated no movement of nitrate or ammonium below a depth of 20 inches during the growing season. Only at the highest rate of applied N was there any evidence of residual nitrate or N movement to a depth greater than 10 inches (Table 5). With greater irrigation amounts or significant precipitation, band placement might still keep more of the applied N in the root zone.
Results from the 15N uptake studies indicate that about two-thirds of N taken up by broccoli fertilized at the optimal 250 pounds/A N rate comes from applied fertilizer. Total N uptake was 310 pounds/A and essentially all the 15N was taken up. This indicates that an optimally-fertilized and irrigated broccoli crop leaves almost no unutilized N in the soil. However, non-uniform plant stand and growth in these plots caution against extrapolating the results to commercial fields.
The cereal rye cover crop planted in 1991 was very effective in removing residual nitrate and ammonium from the soil profile on plots fertilized with 250 pounds N/acre (Table 6). The plots without cover show an increase in nitrate, but not ammonium, concentration with depth. The plots with a cover crop show a much lower level of nitrate and ammonium at all depths and less tendency for higher levels of nitrate in the 20 to 40-inch depth. The cover crop did not accumulate a great amount of dry matter. However, up to 33 pounds N/acre were trapped in the cover crop biomass at the highest rate of N application to the broccoli crop (Table 7). When there was no broccoli left standing, more than 40 pounds N/acre were taken up by the rye following broccoli fertilized with 250 pounds N/acre. The amount of nitrogen in the standing broccoli crop, which was not killed by the mild winter, was not determined.
Table 1. List of N application treatments, 1991 broccoli N utilization trial, NWREC No. Total N applied N applied at planting N applied at five weeks ----------------------lb/A-------------------------------------- 1 0 0 0 2 100 50 broadcast 50 broadcast 3 100 50 shanked 50 broadcast 4 100 50 shanked 50 banded 5 100 50 broadcast 50 banded 6 175 50 broadcast 125 broadcast 7 175 50 broadcast 125 banded 8 250 50 broadcast 200 broadcast 9 250 50 broadcast 200 banded 10 250 50 shanked 200 broadcast 11 250 50 broadcast 0z 12 250 50 broadcast 200 broadcast 15N 13 250 50 broadcast 15N 200 broadcast 14 325 50 broadcast 275 broadcast z100 broadcast at 4 weeks; 100 broadcast at 7 weeks. Table 2. Effect of N rate on yield and mean head weight of broccoli when all fertilizer is broadcast, NWREC, 1991 Mean head weight (g) Trt. No. N rate (lb/A) First harvest All harvests Total yield (T/A) 1 0 95 91 2.9 2 100 137 141 5.5 6 175 145 153 5.4 8 250 154 158 5.9 11 250 159 162 6.0 14 325 123 140 4.1 LSD (0.05) 48 40 1.8 Table 3. Comparison of N rate on mean head weight of broccoli when all fertilizer is broadcast versus band placement of the second N application, NWREC, 1991 Broadcast-broadcast Broadcast-band Trt. No. N rate (lb/A) Mean wt. (g) Trt. no. N rate (lb/A) Mean wt. (g) 2 100 141 5 100 129 6 175 153 7 175 145 8 250 158 9 250 161 Meanz 151 145 zNo significant difference between means (P=0.05). Table 4. Comparison of N rate on mean head weight of broccoli when all fertilizer is broadcast versus shanked-in placement of the at-planting N application, NWREC, 1991 Broadcast-broadcast Shank-broadcast Trt. No. N rate (lb/A) Mean wt. (g) Trt. no. N rate (lb/A) Mean wt. (g) 2 100 141 3 100 136 8 250 158 10 250 151 Meanz 150 144 zNo significant difference between means (P=0.05). Table 5. Soil nitrate and ammonium concentrations (ppm) before planting and after broccoli harvest, NWREC, 1991 N rate, lb/A 0 100 175 250 325 Depth of sample (inches) Nitrate Pre-plant ---------Post-harvest----------- 0-10 6.3 0.1 0.4 3.4 9.0 23.3 10-20 3.3 0.8 0.1 0.7 2.4 4.2 20-30 2.0 0.6 0.1 0.3 1.2 2.0 30-40 2.4 0.8 0.3 0.4 1.0 1.7 Ammonium 0-10 3.2 3.7 4.0 7.6 12.9 30.8 10-20 3.7 2.4 2.7 2.3 4.2 12.1 20-30 6.7 3.7 4.9 2.8 4.8 5.2 30-40 8.2 3.0 3.1 2.3 2.7 4.0 Table 6. Effect of a cereal rye cover crop on soil nitrate and ammonium concentrations (ppm) in broccoli plots fertilized with 250 pounds/acre nitrogen, NWREC, 15 April, 1992 Depth of sample (inches) With cover Without cover Significancez Nitrate 0-10 0.07 0.41 * 10-20 0.09 0.57 ** 20-30 0.12 3.17 ** 30-40 0.18 5.38 ** Ammonium 0-10 1.05 2.36 * 10-20 1.23 2.04 * 20-30 1.27 2.20 * 30-40 1.25 1.96 NS z*,**,NS: Differences significant at the 5% and 1% levels and no significant difference, respectively. Table 7. Nitrogen content, yield, and nitrogen uptake by a cereal rye cover crop in April, 1992, following various rates of N applied to a broccoli crop in 1991, NWREC N rate, lb/acre N content (%) Dry matter yield, lb/acre N uptake, lb/acre 0 1.55 430 6.2 100 1.48 580 7.6 175 1.29 1840 23.0 250 1.63 1510 24.4 250z 1.27 3530 42.5 325 1.45 2330 33.5 Significance NS ** ** zBroccoli plants removed from this plot but were left standing after harvest in all other plots.
1992
Averaged over timing of N application, yield increased with increasing rate of N to 180 pounds N/acre (Table 9). Rates of N application beyond this level had little effect on yield. In the previous experiments, yield peaked at 250 pounds N/acre, with a definite decrease at 325 pounds/acre.
Splitting the N application to provide most of the N nearer the time that the plants experience greatest uptake did not increase yield (Table 10). Even at less than optimal rates of N, yield was greatest when all N was applied at planting. This was a late-spring planting in a very dry year and with carefully controlled irrigation. Thus, N loss to leaching was probably not a factor. Results may have differed for a planting early in the season when leaching of the early N application would be more likely.
It is critical to have an adequate supply of N at planting. Even after a dry winter and with significant available N in the soil at planting (5.6 ppm nitrate in the surface six inches of soil), yield was increased by applying N at planting. Delaying the first N application to four weeks after planting reduced mean head weight from 163 g to 148 g for plants receiving a total of 180 pounds N/acre. In the comparison of the single application versus a 2-way or 3-way split, perhaps the split applications would have been more favorable if a greater proportion of the nitrogen had been applied at planting.
The broccoli crop effectively depleted soil mineral nitrogen at rates of applied nitrogen up to 180 pounds/acre (Table 11). At higher rates the soil nitrate, but not ammonium, concentration was elevated slightly. The results obtained at 240 pounds N/acre are almost identical to those obtained with the optimal rate of 250 pounds in 1991.
Four years of results indicate that the optimum nitrogen rate for broccoli grown on a Willamette silt loam at a plant density of 30,000 to 40,000 per acre is near 250 pounds/acre. Rates in excess of 300 pounds/acre leave significant amounts of residual nitrate available for leaching, whereas rates of 250 pounds/acre or less are efficiently utilized by the crop. Winter cover crops can catch a large proportion of the residual nitrate. Neither method of placement of N nor the timing of N application appear to have much effect on the efficiency of N utilization by broccoli.
Table 8. List of N application treatments, 1992 broccoli N utilization trial, NWREC No. Total N applied At planting At five weeks At seven weeks ---------------------------lb/A------------------------------- 1 0 0 0 0 2 60 40 20 0 3 120 40 80 0 4 180 40 140 0 5 240 40 200 0 6 300 40 260 0 7 60 60 0 0 8 120 120 0 0 9 180 180 0 0 10 240 240 0 0 11 300 300 0 0 12 120 40 40z 40 13 180 40 70z 70 14 240 40 100z 100 15 300 40 130z 130 16 180 0 90z 90 _____________________________________________________________________________ zAt 4 weeks rather than 5 weeks. Table 9. Main effect of N rate on yield and head size of broccoli, NWREC, 1992 N rate Mean head wt., Mean head wt., total Total yield (lb/acre) first harvest (g) of two harvests (g) (tons/acre) 0 76 52 1.6 60 112 95 3.9 120 167 142 5.5 180 195 167 6.8 240 196 169 6.7 300 205 170 7.4 LSD (0.05) 41 31 1.0 Table 10. Effect of splitting nitrogen application on broccoli head size at several rates of applied nitrogen, NWREC, 1992 N rate (lb/acre) Single applic. Two applics. Three applics. --------------------g/head---------------------- 60 101 88 -- 120 159 137 129 180 177 163 162 240 170 169 168 300 171 164 176 Mean (120-300 lb) 169 158 159 Table 11. Effect of rate of broadcast nitrogen on soil nitrate and ammonium concentrations (ppm) following final broccoli harvest, NWREC, September, 1992 N rate, lb/A 0 60 120 180 240 300 LSD(.05) Depth of sample (inches) Nitrate Pre-plant -------------Post-harvest-------------- 0-10 5.6 0.9 0.4 3.1 3.4 9.8 12.1 4.2 10-20 3.8 0.2 0.2 0.9 1.1 4.7 3.6 3.2 20-30 3.1 0.6 0.4 0.3 0.6 1.4 0.9 NSD 30-40 3.8 1.2 0.7 1.2 1.0 1.9 1.4 NSD Ammonium 0-10 4.8 2.0 1.1 1.8 2.0 2.0 2.3 NSD 10-20 2.8 1.8 1.6 2.2 2.7 1.3 1.5 NSD 20-30 5.1 1.5 2.0 1.1 0.9 0.8 1.3 NSD 30-40 4.4 1.5 0.8 1.2 1.1 0.9 1.4 NSD