Plant material and experimental conditions.

The experiment was carried out at the University of Almeria (lat. 36°49′N, long. 2°24′W). Rooted cuttings of fern leaf lavender were obtained from a local nursery with an average DW of 3.4 g and transplanted into 1.5-L polyethylene containers (one plant per container) containing peatmoss and perlite 80:20 (by volume). During the 8 weeks of the trial (7 May to 1 July 2010), the containers were placed in a 150-m² greenhouse. The microclimatic conditions inside the greenhouse for the experimental period were monitored continuously with a data logger (HOBO model H 08-004-02; Onset Computer Corp., Bourne, MA). Average maximum/minimum temperatures were 36/18 °C, and average maximum/minimum of 82%/40% relative humidity, and an average maximum/minimum of photosynthetically active radiation (PAR) of 235/215 µmol·m⁻²·s⁻¹.

Experimental design and treatments.

The experiment consisted of three treatments using different salinities in a standard nutrient solution reported by Jimenez and Caballero (1990) for an adequate growth of ornamental plants in the Mediterranean region. The solution contained (in mg·L⁻¹): 68 phosphates (H₂PO₄⁻), 372 nitrates (NO₃⁻), 192 sulfates (SO₄²⁻), 117 potassium (K⁺), 80 calcium (Ca²⁺), and 34 magnesium (Mg²⁺) amended with different concentrations of sodium chloride [NaCl (sole salinizing agent)] to achieve electrical conductivity (EC) levels of the irrigation water (EC) of either 2.0 (T₁ or control), 4.5 (T₂) or 7.5 (T₃) dS·m⁻¹. The standard nutrient solution was derived from tap water (pH 8.1, EC 0.9 dS·m⁻¹, 50 mg·L⁻¹ Ca²⁺, 34 mg·L⁻¹ Mg²⁺) and phosphoric acid, nitric acid, potassium nitrate, and calcium nitrate. The volume of nutrient solution added manually every day to each container was 70 mL (45% of the container capacity) for each saline treatment during the experimental growing period. The runoff of each container was collected weekly by placing a plastic collection bucket under each container. The buckets were tightly fitted to the containers to prevent evaporation of runoff between collection events and containers were also elevated to prevent them from sitting in the runoff. The levels of EC were established according to the data reported by Contreras (2014) who established that the minimum and maximum EC values were of 2 and 7.5 dS·m⁻¹ in the aquifers used for irrigation in the southeastern coastal areas of Spain. The experimental design consisted of three salinity treatments, four blocks, and four plants (one plant per container) per block giving a total of 12 plants plus border plants. Border plants (36) were placed around the perimeter of the treatment plants to maintain uniform growing conditions for treatment plants.

Monitoring of irrigation and runoff water volume and chemical composition.

Four containers per treatment were randomly selected during the experimental growing period to determine the volume of runoff. Runoff was collected weekly and sample aliquots (25 mL) were filtered through 0.45-µm membrane filters and frozen until nutrient analyses were conducted. From each aliquot, EC and pH values were determined by conductivity meter (models C66; Milwaukee Instruments, Rocky Mount, NC) and pH meter (model pH52, Milwaukee Instruments); and the concentrations of NO₃⁻-N, PO₄²⁻-P, and K⁺ were determined by high-performance liquid chromatography (HPLC) [model 883 Basic IC Plus, anions ion exchange column model Metrosep A SUPP 4, cations ion exchange column model Metrosep C4 100, IC conductivity detector range (0–15,000 µS·cm⁻¹); Metrohm, Herisau, Switzerland)] as described by Csáky and Martínez-Grau (1998). During the experimental growing period, nutrients applied and runoff per plant (in milligrams) were calculated by multiplying the nutrient concentration in the nutrient solution or runoff by their respective volume. These values were then used to calculate the nutrients runoff per plant in percent (percentages that reflect the fraction of applied nutrient which was leached by the plant during 8 weeks).

Biomass and WUE.

At the end of the experiment, the plants of the containers randomly designated for the runoff collection were harvested and the substrate gently washed from the roots. The plants were divided into roots, stems, and leaves and washed with distilled water to be subsequently dried with blotting paper. Roots, stems, and leaves were then oven-dried at 60 °C until they reached a constant weight to measure the respective DWs. The plant DW was calculated as the sum of leaves, stems, and roots. The WUE (grams DW per liter of water applied) of each treatment was calculated as the difference between plant DW after 8 weeks and the initial plant DW divided by the volume of water applied during the experimental period.

Plant nutrient uptake.

The oven-dried samples were ground to a fine powder in a mill (Grindomix GM 200; Retsch, Haan, Germany) taking care to clean the mill thoroughly between each sample. Each sample was divided in two subsamples. The soluble ionic form of nitrates was quantified through HPLC as described by Csáky and Martínez-Grau (1998). Samples were also mineralized with sulfuric acid [H₂SO₄ (96%)] in the presence of P-free hydrogen peroxide [H₂O₂ (30% w/v)] at 300 °C and used for the determination of organic N (Krom, 1980), total P (Hogue et al., 1970), and K⁺ (Lachica et al., 1973) concentration. The total N concentration was calculated as the sum of the organic N and NO₃⁻ concentration. From these determinations and with the DW measured, we calculated the plant nutrient uptake as the difference between final tissue nutrient content after 8 weeks and the initial nutrient content. An initial harvest at the beginning of the experiment gave an average of 75 mg N, 9 mg P, and 3 mg K per plant. These values of nutrient uptake were then used to calculate nutrient uptakes efficiencies (percentages that reflect the fraction of applied nutrient accumulated by plant during the 8 weeks).

Statistical analysis.

The experiment was analyzed as a randomized complete block design, and the values obtained for each plant and each variable were considered as independent replicates. The analysis of variance and the least significant difference tests at P < 0.05 were used to assess the differences between treatments. All statistical analyses were performed using Statgraphics Plus (version 5.1 for Windows; Statpoint Technologies, Warrenton, VA).

Biomass and WUE.

Among the treatments, plant DW was not significantly affected by salinity (Table 1). These results may be due to the lower levels of salinity applied in this experiment to exhibit differences in plant DW in fern leaf lavender. It is typical to find a reduction of growth and plant DW under saline conditions, which induces water stress, caused by the greater difficulty of water absorption, and ionic stress, related to the sodium ion effect (Munns and Tester, 2008).

Table 1.

Influence of salt treatments on plant dry weight, volume of water leached and water use efficiency of fern leaf lavender grown in 1.5-L (0.40 gal) containers containing peatmoss and perlite 80:20 (v/v). Standard nutrient solutions amended with different concentrations of sodium chloride to achieve EC levels of the irrigation water of 2.0 (T₁ or control), 4.5 (T₂), or 7.5 (T₃) dS·m⁻¹.^z

View Table

Although, the volume of saline irrigation water was the same for each treatment during the crop growing period, the volume of water leached per plant increased with increasing NaCl in the irrigation water because the plants could be salt stressed, and thus had higher leaching rates (Table 1). Leachate volumes per plant were 1.6 and 2.2 times higher in T₂ and T₃; respectively, compared with the control plants (T₁). These results agree with Hanson et al. (2006) who reported that an increase of salinity in the root zone decreased the water uptake in plants and consequently increased leaching.

The WUE was not significantly affected by salinity because the volume of nutrient solution was the same for each treatment and also because there were no significant differences in growth in plants due to the salt treatments (Table 1). Different results were found by Alvarez and Sánchez-Blanco (2014) who reported a decrease of WUE related with a growth decrease in red bottlebrush (Callistemon citrinus) production in containers irrigated with saline water.

Irrigation and runoff water quality.

The EC values of the runoff increased compared with the nutrient solution, except for T₁ (Table 2). The runoff and nutrient solution had similar concentrations of NO₃⁻-N and K in all the treatments and may be due to the plants having these elements in excess, whereas PO₄²⁻-P concentrations exhibited differences between runoff and nutrient solutions. The lower concentrations of PO₄²⁻-P in runoff with respect to the nutrient solutions may be related with the nutrient demands of this species during the experimental period and also may be bound in the substrate. From an environmental perspective, the average of NO₃⁻-N concentrations in this experiment (88–93 mg·L⁻¹) exceeded the threshold of 10 mg·L⁻¹ of NO₃⁻-N for drinking water by about 9-times [U.S. Environmental Protection Agency (USEPA), 2010] and the threshold of 50 mg·L⁻¹ established by European Environmental Agency (2014) by 1.8-times. Likewise, the average of PO₄²⁻-P concentrations (8.4–9.0 mg·L⁻¹) exceeded the threshold of 0.025 mg·L⁻¹ established by USEPA (2010) for drinking water by 360-times. These higher concentrations of nutrients in the runoff may result in undesirable changes in natural ecosystems due to eutrophication (Brady and Weil, 1999; Howarth and Marino, 2006).

Table 2.

Average weekly values of pH, electrical conductitity (EC) and concentrations of nutrient elements [nitrate-nitrogen (NO₃^--N), phosphate-phosphorus (PO₄^2--P) and potassium (K⁺)] in nutrient solution (NS) and runoff (R) of fern leaf lavender grown in 1.5-L (0.40 gal) containers containing peatmoss and perlite 80:20 (by volume). Standard nutrient solutions were amended with different concentrations of sodium chloride to achieve EC levels of the irrigation water of 2.0 (T₁ or control), 4.5 (T₂), or 7.5 (T₃) dS·m⁻¹.^z

View Table

Nutrient balance.

Plant nutrient uptake was affected by salinity treatments (Table 3). Plant nitrogen uptake decreased with higher EC, which may be due to Cl⁻ competing for binding sites with nitrogen (Abdelgadir et al., 2005). Plant phosphorus uptake increased with higher EC (4.5 and 7.5 dS·m⁻¹) compared with the control plants. This increase could be related to the rise due to the energy (ATP) required to transport the excess of ions into the vacuoles (Mengel and Kirkby, 2001). Similar results for nitrogen and phosphorus uptake were reported by García-Caparrós et al. (2016), who studied the nutritional behavior of three ornamental potted plants [indian aloe (Aloe vera), flaming katy (Kalanchoe blossfeldiana), and treasure flower (Gazania splendens)] grown under moderate salinity. Plant potassium uptake did not show a clear response to salinity. Conversely with these results, Cassaniti et al. (2009) reported a decrease on potassium uptake in ornamental potted plants such as paper flower (Bougainvillea glabra), juniper-leaf grevillea (Grevillea juniperina), and jelly bush (Leptospermum scoparium) subjected to salinity mainly due to the competition effects between Na⁺ and K⁺ ions which are likely to share the same transport system (Parida and Das, 2005).

Table 3.

Nutrient balance of fern leaf lavender grown in 1.5-L (0.40 gal) containers containing peatmoss and perlite 80:20 (v/v) under different salinity levels. Standard nutrient solutions amended with different concentrations of sodium chloride to achieve EC levels of the irrigation water of 2.0 (T₁ or control), 4.5 (T₂), or 7.5 (T₃) dS·m⁻¹.^z

View Table

The amount of nutrients applied per plant, except for the increase in sodium and chloride of the salinity, were the same in each treatment according to the recommendations given by Jimenez and Caballero (1990) for the growth of ornamental plants under Mediterranean conditions. It should be noted that the N application rates of ornamental container crops can be 15 times higher than fertilizer rates for agronomic field crops (Chen et al., 2001) and that P application rates were 100 times higher than plant requirements (Borch et al., 1998; Lin et al., 1996) resulting in low uptake efficiencies and excessive nutrient loss via runoff.

Under the saline conditions reported here, the uptake efficiency of N and P showed different trends. The nitrogen uptake efficiency (NUE) declined significantly with saline stress (28% lower compared with the control), whereas phosphorus uptake efficiency (PUE) increased with higher EC (4.5 and 7.5 dS·m⁻¹) (23% and 26%, respectively, compared with the control). Similarly, Ristvey et al. (2007) also reported an increase in NUE (from 12% to 38%) and PUE (from 11% to 50%) by reducing the N and P fertilization to match plant growth requirements of azalea (Rhododendron) plants. The values of NUE in our experiment may be overestimated as a result of available nutrients in the soilless potting substrate as reported by Sandrock et al. (2005) who suggest that N may be mineralized in soilless potting media. Similar findings were reported by other researchers in soilless media, who reported a high percentage of N immobilized and denitrified by microorganisms (Cabrera, 2003; Scagel, 2003) or volatilized (Rathier and Frink, 1989). On the other hand, K⁺ uptake efficiency (KUE) did not show a clear response to salinity as happens with the plant K⁺ uptake discussed above. These results disagree with Scagel et al. (2011) who reported an increase of KUE in different azalea cultivars with two different rates of N fertilizers.

Salt stress increased the nutrient runoff per plant and consequently the percentage of runoff compared with the control plants. The increase of N, P, and K runoff can be because the organic potting substrates have little NO₃⁻-N and PO₄²⁻-P retention qualities and very poor cation exchange capacities, especially those that are pine bark or peat based (Marconi and Nelson, 1984). The values of runoff per plant in our experiment ranging for N (65–144 mg), P (6–14 mg), and K (102–225 mg), whereas Ristvey et al. (2007) reported values of runoff for N (4–152 mg) and P (2–16 mg) in an experiment with different fertilizer rates being necessary to point out that these results are not directly comparable to ours due to container size and spacing differences.

The increase in runoff consequently assumed that the nutrient runoff expressed in percentage increase compared with the control plants (28%, 9%, and 27% for N, P, and K, respectively). These runoff percentages were lower compared with the data reported by other researchers. Hershey and Paul (1982) reported 12% to 48% of N runoff for potted chrysanthemums irrigated with N solutions of 100–300 mg·L⁻¹, Cabrera (2003) reported 16% to 58% of runoff in a N balance of two container-grown ornamental plants irrigated with different N concentrations, Warsaw et al. (2009) reported in container-grown ornamental plants that compared with their control, N loading in runoff was reduced, on average, 38% and 59% with their 100% and 75% of daily water use (DWU) application rates. Likewise, P runoff in their 100% and 75% DWU treatments were 46% and 74%, respectively, lower than their control. Similarly, Pershey and Cregg (2015) also reported in four conifers (Pinophyta) grown in containers that the control of the irrigation based on DWU reduced runoff NO₃⁻-N by 36% and 67% and PO₄³⁻-P runoff by 38% and 57% with 100% and 75% DWU irrigation applications, respectively. A final major observation was that to complete the nutrient balance, the total nutrients applied could not be accounted for at the end of the experiment as proposed Ristvey et al. (2007) (41% to 43% in N, 44% to 78% in P, and 23% to 44% in K).