Plant material and growing conditions.

Plug seedlings of ‘Vates’ collards, ‘Nagoya Mix’ flowering kale, ‘Buttercrunch’ lettuce, ‘Sweet Banana’ pepper, and ‘Sweet 100’ tomato were obtained from a commercial producer. The lettuce and flowering kale were from 288-cell plug flats, whereas the remaining plants came from 512-cell flats. On 6 May 2008, the seedlings were transplanted into 4-inch-diameter plastic containers (495 cm³) that were filled with a commercial soilless substrate (Metro-Mix 360; Sun Gro Horticulture, Vancouver, BC, Canada).

Plants were placed in a greenhouse with temperature set points of 20 °C during the day (0800–2000 hr) and 16 °C at night (2000–0800 hr). A 50% shade curtain was set to automatically close whenever an outdoor weather station recorded instantaneous solar radiation of 600 W·m² and reopen at 400 W·m². Seedlings were irrigated at transplanting with tap water and grown for 2 d before treatments commenced. Two days after transplanting (DAT) and before treatments began, pour-through root-zone pH and EC measurements were taken (Cavins et al., 2000).

Treatments.

Plants were separated into two adjoining greenhouse compartments where irrigation water was provided via subirrigation in one house and overhead irrigation in the second (hand-watering). Temperatures across the experimental period were 19.9 ± 0.12 °C (mean ± se) and 19.7 ± 0.12 °C in the subirrigation and overhead irrigation greenhouses, respectively. In each greenhouse, one bench was reserved for each fertilizer treatment and plants were irrigated thoroughly daily. Fertilizer concentration treatments were 50, 100, 200, 350, or 500 mg·L⁻¹ N from a commercial fertilizer mix that contained 20N–2.2P–16.6K (Jack's Professional LX™ 21-5-20 All Purpose Water Soluble Fertilizer; J.R. Peter's, Allentown, PA). Because the tap water was low in magnesium, each fertilizer treatment was also supplemented with 30 mg·L⁻¹ magnesium added from magnesium sulfate (MgSO₄·7H₂O). The municipal tap water had an EC of 0.4 dS·m⁻¹ and alkalinity of 111 mg·L⁻¹ calcium carbonate (CaCO₃). Fertilizer treatments were mixed in 10-gal stock tanks at 100 times the applied concentration. An injector calibrated to a 1:100 ratio and stock solutions were used to fill the 140-gal reservoirs corresponding to each subirrigation treatment. An injector and the treatment stock tanks were also used for the overhead irrigation treatments. Beginning 6 DAT, a weekly pH and EC measurement was taken for each fertilizer treatment in the subirrigation (reservoir) and overhead watered (hose end) treatments (Table 1) to verify consistent fertilizer supply over time.

Table 1.Electrical conductivity (EC) and pH of the overhead irrigation and subirrigation solutions used to fertilize vegetable transplants. Data are single measurements taken 6, 13, 20, and 27 d after vegetable seedlings were transplanted. Irrigation and fertilizer treatments began 2 d after transplanting.

View Fullscreen

Measurements and statistical analysis.

Beginning at 13 DAT, weekly pour-through root-zone pH and EC measurements of 10 randomly selected plants (across all crops) from each bench were taken using the method cited above. At 27 DAT plant height was measured as the distance from the medium surface to the tallest part of the plant. At 30 DAT plants were destructively harvested at the medium surface to determine shoot FW and stem diameter. Stem diameter was measured on the main stem with a digital caliper at 1 cm above the medium surface. Plants were dried in an oven for 3 d at 70 °C and shoot DW was recorded.

The experiment was arranged as a completely randomized design with treatments arranged as a split plot with irrigation method as the main plot factor and fertilizer concentration as the subplot factor. There were five replicates per cultivar per fertilizer treatment and two replicates with five repeated measures per species per irrigation treatment. Analysis of variance tests (SAS version 9.1; SAS Institute, Cary, NC) were conducted to identify differences in the measured parameters in response to irrigation or fertilizer treatment. Model testing using contrasts was used to determine the best model, linear or quadratic, to fit the data. Regression analysis was carried out using the regression procedure of SAS (version 9.1). Pairwise comparisons between irrigation treatments were conducted using Tukey's honestly significant difference test at α = 0.05.

Leachate ph and ec.

The baseline leachate pH (measured 2 DAT, before fertilizer/irrigation treatments) was 6.2 ± 0.02 (mean ± se) and EC was 2.6 ± 0.17 dS·m⁻¹. For the overhead watering treatment, leachate pH decreased between 13 and 27 d (Fig. 1). Final leachate pH of overhead irrigated plants ranged from 5.3 to 6.3 and was lowest for the 350 and 500 mg·L⁻¹ N fertilizer treatments. For the subirrigated plants, leachate pH decreased slightly between 13 and 27 d for the 200 to 500 mg·L⁻¹ N treatments, but the pH of the 50 and 100 mg·L⁻¹ N treatments increased between 13 and 20 d and then stabilized for the remaining duration. Final subirrigation leachate pH was higher than the overhead irrigation leachate pH ranging from 5.8 to 6.6, again with the high concentration fertilizer treatments having the lowest substrate pH. Regardless of irrigation method, substrate ECs between 13 and 27 d were steady or slightly decreased for the 50, 100, and 200 mg·L⁻¹ N treatments. However, substrate EC increased between 13 and 27 d for the 350 and 500 mg·L⁻¹ N fertilizer treatments, increasing by ≈1.5 dS·m⁻¹ in the 500 mg·L⁻¹ N treatment. By the end of the experiment, leachate EC was greater in the overhead treatment (5.3 dS·m⁻¹) than the subirrigation treatment (4.7 dS·m⁻¹). The conventional thought is that subirrigation promotes the accumulation of excess salts in the substrate; this was not indicated by our leachate measurements. However, others report that subirrigated containers accumulate excess salts only in the top-third of the substrate (Kang et al., 2004; Richards and Reed, 2004) due to water evaporation from the substrate surface. Our final pour-through measurements were taken 27 d after experiment initiation. Both Kang et al. (2004) and Pinto et al. (2008) conducted long-term experiments where pour-through EC measurements of subirrigated plants were compared with overhead watered plants. As in our experiment, these researchers also found that for the first 30–40 d following experiment initiation, measured EC values in the highest fertilizer treatment were greater in overhead watered plants as compared with subirrigated plants. After a longer time period (i.e., across 70–90 d), measured pour-through EC was greater in subirrigated plants. In general, because salts are accumulated in the upper profile, pour-through leachate measurements may not accurately reflect pH and EC conditions coinciding with the region of the substrate where roots are actively growing. Therefore, we propose that 1:2 dilution method for substrate pH and EC analysis may be more appropriate for subirrigated plants.

View Figure

• Download as Powerpoint
• Download Figure

Collards growth.

Dry weight, FW, height, and stem diameter of collards were significantly affected by fertilizer concentration and irrigation method (Table 2). Regression analysis shows that 400 mg·L⁻¹ N may be the optimal fertilizer concentration for collards, whereas concentrations greater than 400 mg·L⁻¹ N cause a slight decline in FW and DW. Shoot DW with 350 mg·L⁻¹ N was 3.9- and 3.2-fold greater than 50 mg·L⁻¹ N plants receiving overhead irrigation or subirrigation, respectively (Fig. 2). Stem diameter increased with increasing fertilizer concentration across the entire fertilizer range (Fig. 2). Stem diameter may be used as an indicator of plant strength (Kang et al., 2004), which is important for shipping and plant establishment. Plant height increased as fertilizer concentration increased from 50 to 200 mg·L⁻¹ N and was not significantly different as fertilizer concentration increased to 500 mg·L⁻¹ N. Subirrigated collards exhibited increased growth parameters across all fertilizer concentrations compared with overhead watered plants. There are two possible explanations for this observation. First, the relatively high fertility needs of collards were perhaps more easily satisfied in subirrigation. Second, subirrigation supplies water more uniformly compared with overhead irrigation (Cantliffe and Soundy, 2000), meaning that each plant receives a thorough watering to container capacity during each irrigation event (Pinto et al., 2008); whereas consistent, even application with overhead irrigation can be difficult to achieve. In fact, we found that when averaged across all fertilizer treatments FW/DW of all crops in our experiment was greatest when plants were subirrigated as opposed to overhead watered.

Table 2.Two-way analysis of variance for the main effects and interactions of fertilizer concentration [Fert (50, 100, 200, 350, and 500 mg·L⁻¹ nitrogen)]^z and irrigation [Irr (overhead irrigation and subirrigation)] on plant dry weight (DW), fresh weight (FW), height, and stem diameter of five vegetable transplant crops. The mean square error (MSE) is presented for the linear and quadratic regression of measured values in response to applied fertilizer rate for overhead (_Ovr) and subirrigation (_Sub).

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Kale growth.

Fertilizer concentration and irrigation method significantly affected FW, DW, and height of kale (Table 2). Kale reached its highest FW and DW at 350 mg·L⁻¹ N. There was significant interaction between irrigation methods on FW, DW, and stem diameter (Table 2); i.e., regression analysis indicated subirrigated plants reached optimal (maximum) FW with the 300 mg·L⁻¹ N treatment, whereas overhead irrigated plants reached optimal FW at 400 mg·L⁻¹ N (Fig. 2). The FW and DW of subirrigated kale was sensitive to high fertilizer concentration as both declined at N concentrations greater than 350 mg·L⁻¹ N. Overhead irrigated kale DW and FW did not significantly decline at concentrations greater than 350 mg·L⁻¹ N. The difference in growth patterns in the two treatments suggests that accumulating salts reduced growth in the subirrigation treatment. Although greater accumulated salts in subirrigated plants are not indicated from Fig. 1, the pour-through leachate method may not have accurately accounted for salts levels across the entire substrate volume. When averaged across all fertilizer treatments, the FW, DW, and height of subirrigated plants were generally greater than overhead irrigated plants. There were no significant differences in stem diameter between irrigation methods, yet stem diameter did follow a general quadratic relationship in both irrigation methods.

Lettuce growth.

There was a significant interaction between irrigation method and fertilizer concentration on the FW and DW of lettuce (Table 2). Subirrigated plants reached their highest DW and FW at 200 mg·L⁻¹ N, whereas overhead irrigated plants reached their highest DW and FW at 350 mg·L⁻¹ N (Fig. 2). Further, FW and DW of subirrigated plants was reduced by high concentration fertilizer treatments (350 and 500 mg·L⁻¹ N) (Fig. 2). Stem diameter was not significantly affected by irrigation treatment; regarding fertilizer treatment, the only significant reductions are observed at 500 mg·L⁻¹ N (Fig. 2). Height of overhead irrigated plants was shortest at 50 mg·L⁻¹ N averaging 19.8 ± 0.9 cm, greatest at 350 mg·L⁻¹ N averaging 34.7 ± 1.5 cm, and declined as fertilizer concentration increased to 500 mg·L⁻¹ N averaging 24.9 ± 0.6 cm. Our results are similar to those of Masson et al. (1991) who reported that shoot DW of overhead irrigated ‘Ithaca’ lettuce grown in a 288-cell plug tray was greatest at 300 mg·L⁻¹ N for unlighted plants and 400 mg·L⁻¹ N for plants receiving supplemental light. Similarly, Tremblay and Senecal (1988) reported that shoot growth of lettuce seedlings was greater at 350 mg·L⁻¹ N as compared with 150 mg·L⁻¹ N. However, higher N fertility reduced the root:shoot ratio and was considered not acceptable for high quality vegetable transplants for field production (Tremblay and Senecal, 1988).

Pepper growth.

Fresh weight and DW of pepper were significantly affected by fertilizer and irrigation method (Table 2). Subirrigated pepper appeared sensitive to soluble salt accumulation. The greatest pepper FW and DW occurred at 200 mg·L⁻¹ N but declined dramatically at 350 mg·L⁻¹ N (Fig. 2). The FW of overhead plants was greatest at 100 mg·L⁻¹ N and did not vary significantly at higher fertilizer concentrations. Similarly, overhead watered plant DW increased significantly between 50 and 100 mg·L⁻¹ N but did not vary significantly at higher fertilizer concentrations.

Height of pepper plants was significantly affected by fertilizer concentration, but these differences were quite small (Table 2). However, pepper height did differ significantly between irrigation treatments at each fertilizer concentration except 50 and 100 mg·L⁻¹ N. Stem diameter of pepper was largely unaffected by irrigation method. At the end of the study, pepper and tomato plants did not have flowers under any of the treatments; plant fertilizer requirements may change once the plant is in a reproductive stage.

Other researchers have found that the optimal shoot DW and fruit DW of ornamental ‘Treasures Red’ pepper was reached at 200 mg·L⁻¹ N for both subirrigated and overhead irrigated plants (Kang et al., 2004). Bar-Tal et al. (1990) reported the optimal N and P concentrations for subirrigated pepper seedlings grown in 15-cm³ cells to be 5 mm N (70 mg·L⁻¹ N) and 0.5 mm P (15 mg·L⁻¹ P). Increasing the fertilizer concentration to 10 mm N and 1 mm P did not increase shoot DW.

Tomato growth.

Fresh weight, DW, and height of tomato were significantly affected by fertilizer concentration and irrigation method (Table 2). FW of overhead irrigated tomato was 17.6 ± 0.9 g/plant at 50 mg·L⁻¹ N and increased to 80.7 ± 4.8 g/plant at 350 mg·L⁻¹ N (Fig. 2). For subirrigated plants, the greatest FW was reached at 200 mg·L⁻¹ N. The FW under both irrigation regimes decreased at 500 mg·L⁻¹ N. The DW of both overhead irrigated and subirrigated tomato increased significantly between 50 and 200 mg·L⁻¹ N but did not vary significantly at higher fertilizer concentrations. Across all fertilizer concentrations, tomato DW was generally greater under subirrigation mean 9.3 ± 0.6 g/plant, compared with overhead irrigated plants mean 5.9 ± 0.4 g/plant.

Reduced fertilizer concentration was an effective method of controlling plant height of tomato. The height of overhead plants increased significantly as fertilizer concentration increased from 50 to 350 mg·L⁻¹ N. However, withholding fertilizer to control height is done so at a direct cost to overall plant growth. The DW of overhead tomato was 7.6 ± 0.3 g/plant at 350 mg·L⁻¹ N but was 2.7 ± 0.2 g/plant at 50 mg·L⁻¹ N. Tomato stem diameter was unaffected by irrigation method but tended to follow a quadratic pattern in response to increasing fertilizer concentration. Similarly, Rideout and Overstreet (2003) reported that reducing P fertilizer concentration from 3.6 to 1.5 mg·L⁻¹ reduced height, DW, and stem diameter of tomato seedlings grown in 128-cell flats. Unlike our experiment, shoot DW of tomato seedlings grown in 128-cell plug trays increased as N fertilizer concentration increased from 100 to 400 mg·L⁻¹ (Masson et al., 1991).

Fertilizer response as influenced by irrigation method.

There was a significant fertilizer concentration by irrigation method interaction for FW and DW of kale, lettuce, and pepper (Table 2). Collards FW exhibited a significant interaction as well, but not DW (Table 2). This relationship indicates a possible gain in fertilizer use efficiency under subirrigation. To further elucidate this relationship, quadratic regression equations were used to determine the fertilizer concentration range that provided ≥90% of optimal DW (Fig. 3). Shoot DW was relatively unaffected by irrigation regime for collards and tomato (Fig. 3). However, to reach 90% of optimal DW of kale, lettuce, and pepper less fertilizer is required under subirrigation, compared with overhead irrigation. For these three crops, deleterious effects were predicted at high fertilizer concentration (greater than 350–440 mg·L⁻¹ N) under subirrigation, but not with overhead irrigation.

View Figure

• Download as Powerpoint
• Download Figure

A few reports have directly compared optimal fertilization rates of overhead irrigation and subirrigation, with variable results. The optimal fertilizer concentration for vegetative growth of ‘Treasures Red’ ornamental pepper was 200 mg·L⁻¹ N regardless of irrigation method (Kang et al., 2004). Optimal growth of subirrigated vinca (Catharanthus roseus ‘Pacifica Red’) was reached at 2 mm K (78 mg·L⁻¹), whereas 4 mm K (156 mg·L⁻¹) was required for top-watered plants (Haley and Reed, 2004). Growth of overhead irrigated poinsettia (Euphorbia pulcherrima ‘Gutbier V-14 Glory’) was optimal at 250 mg·L⁻¹ N, whereas a reduced concentration (175 mg·L⁻¹ N) was optimal for subirrigated plants (Dole et al., 1994.) However, DW of ‘Red Sails’ poinsettia was unaffected by fertilizer concentration for both subirrigated and overhead irrigated plants across the range of 100 to 325 mg·L⁻¹ N (Cox, 2001). Optimal fertilizer concentration for DW of peace lily (Spathiphyllum ‘Sensation’) was unaffected by irrigation method when plants were grown in a peat-based medium, but plants growing in a coir-based medium had an optimum DW at 8 mm N (112 mg·L⁻¹ N) for subirrigation and 16 mm N (224 mg·L⁻¹ N) for top irrigation (Mak and Yeh, 2001).

Besides irrigation method, several production factors are known to influence the fertilizer concentration response of a given plant species, including environmental conditions such as temperature (Kang and van Iersel, 2001) and light intensity (Masson et al., 1991), growth medium (James and van Iersel, 2001; Mak and Yeh, 2001; Poole and Conover, 1992), leaching fraction (Yelanich and Biernbaum, 1993), concentration/proportion of specific nutrients (Haley and Reed, 2004), and salt ions that can accumulate in the root zone (Massa et al., 2009). Taken together, these results suggest that species- and environment-specific guidelines should be developed for determining whether to decrease fertilizer concentrations when adapting fertilizer guidelines developed for overhead watering to subirrigation systems.