Plant material and acclimation period.

One-year-old rooted cuttings of ‘Emerald’ SHB were washed clean of substrate using tap water. Cuttings were transplanted to a previously described hydroponic system (Poonnachit and Darnell, 2004) in which each plant was exposed to 2 L of constantly aerated nutrient solution. Plants were acclimated in a nutrient solution containing 0.5 mm KNO₃, 0.5 mm K₂PO₄, 1.0 mm MgSO₄, 0.5 mm CaCl₂, 0.045 mm H₃BO₃, 0.01 mm MnSO₄, 0.01 mm ZnSO₄ with 0.3 µM CuSO₄, 0.2 µM Na₂MoO₄, and 45 µM Fe provided as Sequestrene 330 (10% iron(III)-diethylenetriamine pentaacetic acid) (Becker Underwood, Inc., Ames, IA). Nutrient solution was buffered to pH 5.5 using 5.0 mm 2-(4-morpholino)-ethane sulfonic acid (MES). Nutrient solution pH was adjusted back to pH 5.5 using 1 M HCl or 1 M KOH as needed, three times each week. Fresh nutrient solution was provided every 7 d to maintain nutrient concentrations. Debris, algae, and detached roots were removed from the reservoir to minimize impact on solution pH. Acclimation lasted 21 d in both experiments. Plant fresh weight (FW) was measured nondestructively at the end of acclimation.

All experiments were carried out in a temperature-controlled greenhouse set to maintain air temperatures between 15.6 and 26.7 °C. Natural light was supplemented by metal halide lamps with a photosynthetic photon flux density output of 367 nmol·m^–2·s^–1for 14 h every day.

Expt. 1: Nitrogen form effect.

After acclimation, one half of the plants were transferred to a nutrient solution in which 5 mm NO₃^– (delivered as KNO₃) was the only form of N. The other half of the plants received a nutrient solution in which 5 mm NH₄⁺ [delivered as (NH₄)₂SO₄] was the only form of N. Additional salts were not added to compensate for K⁺ and SO₄^– concentrations. Considering that the magnitude of N uptake is much greater than K⁺ and SO₄^– uptake (Havlin et al., 2013), it is reasonable to assume that most changes in rhizosphere pH were the result of N form as opposed to these counterions. In addition, KNO₃ and (NH₄)₂SO₄ have been used as blueberry fertilizers in previous scientific studies with SHB (Nunez et al., 2015). Nutrient solutions were adjusted to pH 5.5 during nutrient solution preparation, but they did not contain any pH buffer. All other nutrient concentrations were the same as during the acclimation period. Nutrient solution in reservoirs with no plants was used as the control for each treatment. Nutrient solution pH, N uptake, and leaf chlorophyll content were measured periodically. The treatment period lasted 35 d.

This experiment followed a completely randomized design with one variable (N form) and two treatments (NH₄⁺ and NO₃^–). Individual plants were the experimental unit. There were four single-plant replications per treatment. All data except for rhizosphere pH were analyzed using one-way analysis of variance (ANOVA) in R version 3.5.1 (R Core Team, 2017). Rhizosphere pH data were analyzed using a linear mixed model (see section Rhizosphere pH measurement).

Expt. 2: Ammonium uptake effect.

After acclimation, plants were transplanted to a split-root hydroponic system. Plants were held upright using garden stakes and Velcro straps (Supplemental Fig. 1). Root systems were separated into two parts of similar volume. Each half of the root system was placed in a different 1.5-L reservoir (termed reservoir A and reservoir B). In reservoir A, all plants were exposed to a complete, buffered nutrient solution that contained 5 mm NH₄⁺. NH₄⁺ concentration treatments were delivered in reservoir B. Seven plants were exposed to a nutrient solution containing 5 mm NH₄⁺ in reservoir B, whereas seven different plants received a nutrient solution without N in reservoir B. NH₄⁺ was delivered as (NH₄)₂SO₄ in all solutions. Nutrient solutions were adjusted to pH 5.5 during nutrient solution preparation. Solutions in reservoir A were buffered with 5.0 mm MES. Solutions in reservoir B were not buffered to track pH change. All other nutrient concentrations were the same as during the acclimation period. In summary, treatments were (reservoir A/reservoir B) NH₄⁺/NH₄⁺ and NH₄⁺/0N. Reservoirs with no plants were used as controls for each treatment. Nutrient solution pH, N uptake, and leaf chlorophyll content were measured periodically. Root vitality was measured 30 d after transplant to the split-root system. The treatment period lasted 35 d. At the end of this period, plants were harvested destructively to measure plant and organ dry weight and tissue N content.

This experiment followed a completely randomized design with one variable (NH₄⁺ concentration in reservoir B) and two treatments (0 mm NH₄⁺ and 5 mm NH₄⁺). Individual plants were the experimental unit. There were seven single-plant replications per treatment. All data except for rhizosphere pH were analyzed using one-way ANOVA in R version 3.5.1 (R Core Team, 2017). When appropriate, reservoir A and reservoir B were compared via paired one-tail t tests (α = 0.05). Rhizosphere pH data were analyzed using a linear mixed model (see section Rhizosphere pH measurement).

Rhizosphere pH measurement.

Nutrient solution pH was measured three times per week during the treatment period in both experiments using a standardized portable pH meter (Accumet AP110; Fisher Scientific, Hampton, NH). Solution cross-contamination and disturbance of fine roots was avoided during pH measurement. Rhizosphere pH for a single treatment on a single day was compared with controls using one-sample single-tail t tests (α = 0.05) performed in R version 3.5.1 (R Core Team, 2017). Treatments were compared through linear mixed-effect analysis in R version 3.5.1 using package lme4 (Bates et al., 2015). Midweek pH data were used. Treatment, treatment duration, and their interaction were designated as fixed effects; repeated measures per subject and week were designated as random sources of error. Statistical significance was determined by likelihood ratio tests comparing the full model against a model without the effect being investigated. Linear regression slopes are provided only as a reference.

Nitrogen uptake.

Nutrient solution volume was measured at the end of each week. About 20 mL solution was collected in scintillation vials from each reservoir. Samples were filtered through a fine-wire mesh to remove debris, acidified with 1 drop 1 N HCl, and stored frozen at –20 °C. Samples were submitted for analysis at an institutional laboratory (University of Florida IFAS Analytical Research Laboratory, Gainesville, FL), where NO₃^– and NH₄⁺ concentrations were measured by automated colorimetry according to Environmental Protection Agency protocols 353.2 and 351.2, respectively (U.S. Environmental Protection Agency, 1993a, 1993b). Both NO₃^– and NH₄⁺ concentrations were measured in all nutrient solution samples. Nutrient solution volumes and concentrations were used to determine N content after 7 d of plant exposure. N removal from the nutrient solution was computed by subtracting N content after 7 d from the N content of fresh nutrient solution. N uptake was normalized according to whole-plant FW. Nitrification rates were estimated from the NO₃^– concentration at day 7 in control reservoirs that received NH₄⁺-containing nutrient solution.

Chlorophyll status and N content.

Plants were harvested destructively after 5 weeks of treatment. Canes, leaves, and roots from each reservoir were dried to a constant weight and ground until they passed through a #20 mesh screen. N content in these samples was measured through combustion–thermal conductivity at a commercial laboratory (Waters Agricultural Laboratories, Camila, GA) according to Association of Analytical Communities protocol 2.4.02 (Association of Analytical Communities, 2012). Chlorophyll status was measured nondestructively using a leaf greenness SPAD 502 meter (Konika Minolta, Osaka, Japan). Unitless measurements were made on three positions on the youngest fully expanded leaf of each plant at the end of the acclimation period and during week 3 of treatment.

Root viability.

Relative electrolyte leakage (REL) was used to assess root viability per McKay and White (1997). Root tissue samples (≈250 mg) were collected from each reservoir. Root samples contained a third-order root and all the attached lower order roots. This sampling method has been used previously to study SHB roots (Nunez et al., 2015). Roots were blotted dry, submerged in 16 mL deionized H₂O, and incubated for 24 h under continuous shaking at ≈60 rpm. Electrical conductivity was measured immediately after incubation and after autoclaving for 30 min at 121 °C using a standardized conductivity meter (Accumet XL30, Fisher Scientific). Deionized water incubated with no roots was used as an assay control.

Nitrogen form effect on rhizosphere pH.

In our experiment, cumulative NH₄⁺ uptake by ‘Emerald’ SHB was significantly greater than NO₃^– uptake (5.66 mg N/g FW vs. 4.11 mg N/g FW, P = 0.03). Nitrification in the nutrient solution was negligible (average, 0.50 µmol NO₃^–/reservoir/d). Although it is possible that the low replication number in this experiment affected the analyses, these trends are consistent with previous studies in which SHB and other acid-loving plants exhibited a preference for NH₄⁺ uptake over NO₃^– uptake (Merhaut and Darnell, 1995; Poonnachit and Darnell, 2004; Townsend, 1966). Low NO₃^– uptake is also consistent with the significantly lower chlorophyll status observed in NO₃^–-grown plants compared with NH₄⁺-grown ones during week 3 of treatment (36.28 SPAD vs. 59.78 SPAD, P = 0.02). FW and dry weight (DW) data were lost as a result of human error, but observations indicated that biomass accumulation mirrored N uptake and chlorophyll status trends.

NH₄⁺-grown ‘Emerald’ SHB plants acidified their rhizosphere consistently and significantly within 3 d of each solution change (Fig. 1, P ≤ 0.0001 in all cases). NO₃^–-grown plants alkalized their nutrient solution starting on day 5 of the experiment (P ≤ 0.01 in all cases). N form-dependent rhizosphere acidification and alkalization is consistent with previous reports in blueberry (Merhaut and Darnell, 1995; Nunez et al., 2015), and other woody (Gogorcena et al., 2001; Hawkins et al., 2008; Marschner et al., 1991; Suzko et al., 2018; Valentinuzzi et al., 2015) and herbaceous (Lei et al., 2015; Santi and Schmidt, 2009) plants. When analyzed through a linear mixed-effect model, the interaction between N source and treatment duration affected rhizosphere pH significantly (χ² = 10.18, df = 1, P ≤ 0.01). Midweek rhizosphere pH of NH₄⁺-grown plants was nearly constant (linear regression slope = – 0.01), whereas midweek rhizosphere pH of NO₃^–-grown plants increased over time (linear regression slope = 0.15). N source also affected rhizosphere pH (χ² = 48.64, df = 1, P ≤ 0.001). NO₃^– uptake raised rhizosphere pH by an average of 3.52 units compared with NH₄⁺ uptake (Supplemental Fig. 2). Treatment duration did not affect rhizosphere pH significantly (χ² = 3.59, df = 1, P = 0.06).

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a complete nutrient solution containing 5 mm NO₃^– (dashed lines) or 5 mm NH₄⁺ (solid lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in reservoirs with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

View Full Size

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a complete nutrient solution containing 5 mm NO₃^– (dashed lines) or 5 mm NH₄⁺ (solid lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in reservoirs with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a complete nutrient solution containing 5 mm NO₃^– (dashed lines) or 5 mm NH₄⁺ (solid lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in reservoirs with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

Although NH₄⁺-containing fertilizers are the basis for blueberry nutrition, NO₃^– is not absent from the blueberry rhizosphere. Fertilizers used for conventional (Kennedy, 2011; Liu et al., 2012; Smith and Harris, 2017) and organic (Larco et al., 2013) blueberry production can contain mixes of NH₄⁺ and NO₃^–. Also, nitric acid—which dissociates into NO₃^–—can be used to acidify blueberry fertigation (Gallegos-Cedillo et al., 2018). In addition, blueberry soils can have high nitrification rates (Hanson et al., 2002). Our results suggest that using NO₃^– for blueberry fertilization might lead to rhizosphere alkalization. In blueberry, high rhizosphere pH can lead to nutrient deficiencies, stunted growth, and low survival (Finn et al., 1993; Tsuda et al., 2014). Thus, NO₃^– use would exacerbate the need for soil amendments and soilless substrates for blueberry cultivation. Further research is necessary to establish the extent to which NH₄⁺:NO₃^– combinations affect rhizosphere pH under field conditions.

The observed rhizosphere acidification in NH₄⁺-grown ‘Emerald’ SHB plants represents a conundrum, because it is not possible to determine whether these plants acidify their rhizosphere independently from the effect of NH₄⁺ uptake. Previously, rhizosphere acidification capacity has been used as a proxy to identify genotypes with broad soil adaptation (Suzko et al., 2018) or superior nutrient uptake capacity (Lei et al., 2015). Thus, determining whether SHB is capable of acidifying its rhizosphere independently of N form might be a first step in breeding blueberries that tolerate neutral or higher pH soils. This trait has been identified as a priority for blueberry breeders in the United States and Canada (Gallardo et al., 2018).

Rhizosphere acidification with and without NH₄⁺ uptake.

‘Emerald’ SHB plants were transplanted to a split-root hydroponic system to distinguish basal rhizosphere acidification from NH₄⁺ uptake-related rhizosphere acidification. Thirty days after transplant, the REL of roots from the same reservoir was not significantly different (P_{reservoir A} = 0.35, P_{reservoir B} = 0.19). Abiotic stress increases root REL (McKay and White, 1997). REL results in this experiment suggest that roots were not stressed by manipulation and transplant to the split-root hydroponic system.

‘Emerald’ SHB plants in treatment NH₄⁺/NH₄⁺ took up significantly more N than plants in treatment NH₄⁺/0N (3.95 mg N/g FW vs. 2.73 mg N/g FW, P = 0.04). N uptake from reservoir A was not significantly different between treatments (average = 2.46 mg N/g FW, P = 0.10). All N uptake in treatment NH₄⁺/0N was from reservoir A, because reservoir B did not contain N. These results are consistent with previous studies in which plants that were exposed to low nutrient concentrations in one half of their root systems took up less nutrients overall (Ma and Rengel, 2008; Shen et al., 2005).

Nutrient solution NH₄⁺ concentration affected root N concentration. Roots in reservoir B in treatment NH₄⁺/0N exhibited significantly lower N concentrations than roots in reservoir B in treatment NH₄⁺/NH₄⁺ (P < 0.001). Roots in reservoir B in treatment NH₄⁺/NH₄⁺ exhibited significantly greater N concentrations than roots in reservoir A in all treatments (P < 0.01), suggesting there was greater NH₄⁺ uptake from nutrient solutions that were not buffered. Contrasting N uptake and root N concentration in reservoir B did not affect whole-plant N status. Treatments did not affect plant DW (average = 29.38 g, P = 0.42), root N concentrations in reservoir A (P = 0.76), or leaf N concentration (P = 0.31, Fig. 2). All leaf N concentrations were within previously established sufficiency ranges for blueberry (Smith and Harris, 2017; Strik and Vance, 2015). In addition, leaf chlorophyll status did not change between acclimation and treatment periods (49.1 SPAD vs. 52.4 SPAD, P = 0.21), and it was not significantly different between NH₄⁺/NH₄⁺ and NH₄⁺/0N plants during the treatment period (55.0 SPAD vs. 49.9 SPAD, P = 0.11). These results suggest that overall plant N status was unaffected by the treatments in the 5-week treatment period used in this experiment. It is likely that prolonged cultivation under these conditions could lead to N-deficient plants in treatment NH₄⁺/0N.

N concentration in roots and leaves of hydroponically grown ‘Emerald’ southern highbush blueberry as determined by combustion–thermal conductivity. Plants were grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺) or no N (treatment NH₄⁺/0N). Different letters indicate significant differences between treatments (P ≤ 0.05). ns = not significant.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

View Full Size

N concentration in roots and leaves of hydroponically grown ‘Emerald’ southern highbush blueberry as determined by combustion–thermal conductivity. Plants were grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺) or no N (treatment NH₄⁺/0N). Different letters indicate significant differences between treatments (P ≤ 0.05). ns = not significant.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

N concentration in roots and leaves of hydroponically grown ‘Emerald’ southern highbush blueberry as determined by combustion–thermal conductivity. Plants were grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺) or no N (treatment NH₄⁺/0N). Different letters indicate significant differences between treatments (P ≤ 0.05). ns = not significant.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

Roots in reservoir B of all treatments acidified their rhizosphere consistently and significantly starting 2 d after each solution change (Fig. 3, P ≤ 0.0009 in all cases). The magnitude of rhizosphere acidification by roots exposed to NH₄⁺ in their nutrient solution was comparable between Expts. 1 and 2, and it is consistent with previous reports (Marschner et al., 1991). When analyzed through a linear mixed-effect model, the interaction between NH₄⁺ concentration in reservoir B and treatment duration significantly affected rhizosphere pH (χ² = 8.36, df = 1, P < 0.01). Midweek nutrient solution pH remained relatively constant in treatment NH₄⁺/NH₄⁺ (linear regression slope = –0.17) (Supplemental Fig. 3), suggesting that rhizosphere acidification in these plants responded to a constant stimulus, such as NH₄⁺ uptake. In contrast, midweek nutrient solution pH decreased over time in treatment NH₄⁺/0N (linear regression slope = –0.31), suggesting that rhizosphere acidification in these plants increased in response to changing conditions. It is possible that diminishing root N concentrations in reservoir B of treatment NH₄⁺/0N increased the rhizosphere acidification capacity in these roots (local response). In addition, it is possible that the entire root system in treatment NH₄⁺/0N responded to NH₄⁺ uptake from reservoir A (systemic response). Additional research is necessary to elucidate whether local or systemic responses are involved in this apparent acclimation.

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺, solid lines) or no N (treatment NH₄⁺/0N, dashed lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in bottles with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

View Full Size

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺, solid lines) or no N (treatment NH₄⁺/0N, dashed lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in bottles with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

Rhizosphere pH of ‘Emerald’ southern highbush blueberry plants grown in a split-root hydroponic system in which reservoir A contained a complete nutrient solution buffered to pH 5.5, and reservoir B contained a nonbuffered nutrient solution containing 5 mm NH₄⁺ (treatment NH₄⁺/NH₄⁺, solid lines) or no N (treatment NH₄⁺/0N, dashed lines). Fresh nutrient solution at pH 5.5 was delivered every week. Asterisks indicate significant differences from nutrient solution pH in bottles with no plants at P ≤ 0.05.

Citation: HortScience horts 54, 5; 10.21273/HORTSCI13764-18

• Download Figure

NH₄⁺ concentration in reservoir B (χ² = 7.45, df = 1, P < 0.006) and treatment duration (χ² = 12.63, df = 1, P < 0.0004) also affected rhizosphere pH significantly. Ammonium uptake decreased rhizosphere pH by an average of 0.75 units beyond basal rhizosphere acidification (Supplemental Fig. 3). These results are consistent with an increase in H⁺-ATPase activity as a consequence of cation uptake. Molecular and genetic evidence for this phenomenon are available for model species [see Hachiya and Sakakibara (2017) for review], but not for blueberry. Root organic acid exudation increases when P and micronutrients are scarcely bioavailable (Marschner et al., 1987). Considering P and micronutrients were equally abundant in both treatments, it is unlikely that organic acid exudation was involved in rhizosphere pH change in our experiment.

Previously, lack of rhizosphere acidification in Vaccinium spp. was ascribed to relaxed selection in the acidic soils where these species evolved (Nunez et al., 2015). Our results contradict this notion, as ‘Emerald’ SHB plants are, in fact, capable of acidifying their rhizosphere. In the future, additional genotypes and related taxa should be tested to determine whether there is variation in this trait among cultivated and wild Vaccinium spp.

Other woody and herbaceous species are capable of acidifying their rhizosphere even when they take up NO₃^– (Gogorcena et al., 2001; Hawkins et al., 2008; Santi and Schmidt, 2009), suggesting their H⁺ extrusion rates (from H+-ATPase activity or organic acid exudation) are greater than H⁺ influx rates through the H⁺/NO₃^– symport. In our experiment, ‘Emerald’ SHB plants were capable of acidifying their rhizosphere when grown with NH₄⁺ or without N, but not when grown with NO₃^–. These results suggest that ‘Emerald’ SHB H⁺ extrusion rates are less than H⁺ influx rates caused by NO₃^– uptake. Considering that blueberry plants are exposed to NH₄⁺ and NO₃^– under typical field conditions (Hanson et al., 2002), further research is necessary to determine whether they can create an acidic microenvironment around their roots in agricultural settings.

In conclusion, N uptake is the main driver of rhizosphere pH in ‘Emerald’ SHB. NH₄⁺ uptake, in particular, appears to be essential for maintaining low rhizosphere pH. These findings advance our understanding of rhizosphere chemistry in acid-loving plants and provide a baseline for testing NH₄⁺:NO₃^– mixes and other Vaccinium spp. in the future.