Content uploaded by Gustavo Fernando Kreutz
Author content
All content in this area was uploaded by Gustavo Fernando Kreutz on Mar 17, 2023
Content may be subject to copyright.
HORTSCIENCE 58(4):467–474. 2023. https://doi.org/10.21273/HORTSCI17040-22
Identifying Lettuce Accessions for
Efficient Use of Phosphorus in
Hydroponics
Gustavo F. Kreutz
Horticultural Sciences Department, Everglades Research and Education Center,
University of Florida, 3200 East Palm Beach Road, Belle Glade, FL 33430, USA
Jehangir H. Bhadha
Department of Soil, Water, and Ecosystem Sciences, Everglades Research and Education
Center, University of Florida, 3200 East Palm Beach Road, Belle Glade, FL 33430, USA
Guodong D. Liu
Horticultural Sciences Department, University of Florida, 2550 Hull Road,
Gainesville, FL 32611, USA
Germ
an V. Sandoya
Horticultural Sciences Department, Everglades Research and Education Center,
University of Florida, 3200 East Palm Beach Road, Belle Glade, FL 33430, USA
Keywords. breeding, Lactuca sativa, nutrient film technique, phosphorus use efficiency
Abstract. Lettuce (Lactuca sativa L.) is the most common leafy vegetable produced hydroponically
in the United States. Although hydroponic systems are advantageous due to lower pest and disease
pressure, and reduced water and nutrient requirements, the increasing prices of fertilizers, including
phosphorus (P), still influences the profitability of hydroponic production of lettuce. Characterizing let-
tuce germplasm capable of producing high yield using less P inputs may help reduce fertilizer use,
production costs, and P loads in wastewater. In this study, 12 lettuce accessions were grown in four
experiments in a nutrient film technique system. In the first two experiments, the treatments consisted
of two P concentrations (3.1 and 31 mg·L
21
). Lettuce cultivated with 3.1 mg·L
21
of P had variable
shoot and root biomass, root–shoot ratio, P uptake efficiency, and P utilization efficiency, indicating
the existence of genetic variation. Five accessions (‘Little Gem’, 60183, ‘Valmaine’, BG19-0539, and
‘Green Lightning’) were considered efficient to P because produced similar shoot biomass with the
low and high P treatments. In the third and fourth experiments, the treatments consisted of two P
sources (monosodium phosphate (NaH
2
PO
4
) and tricalcium phosphate [TCP; Ca
3
(PO
4
)
2
]. Initially, ex-
tra 5 mM of calcium (Ca) was added to the TCP solution to reduce the TCP solubility and, hence, P
bioavailability to plants. All accessions produced similar shoot and root weight with both treatments,
indicating that the TCP treatment did not cause low-P stress to the plants. After, the extra Ca concen-
tration added to TCP was increased to 10 mM, resulting in low-P stress and a significant reduction
in shoot weight of all accessions. Despite the severe P stress, ‘Little Gem’and 60183 were among the
accessions with the least shoot weight reduction in the TCP treatment. Variability was observed in
root biomass root–shoot ratio among accessions under the TCP treatment, suggesting that
lettuce accessions responded differently to P stress conditions. The genetic variation for P
use efficiency (PUE) and PUE-related traits in lettuce grown hydroponically suggests the
feasibility of breeding new lettuce cultivars from elite lettuce germplasm adapted to low
P availability in hydroponics.
Lettuce (Lactuca sativa L.), one of the
most consumed vegetables worldwide, is a
versatile crop that can be produced in a wide
range of production systems, from field to
greenhouse (Ahmed et al. 2021; Sandoya
2019; Sandoya et al. 2021). In recent years,
hydroponic systems such as floating raft, nu-
trient film technique, and vertical towers have
been increasingly adopted for lettuce produc-
tion in the United States (Resh 2022). This
phenomenon is, in part, a result of the identi-
fication of lettuce cultivars suitable for hydro-
ponic production that allow growers to
achieve yields similar to those observed in
field cultivation (Resh 2022). Hydroponic
systems allow cultivation of lettuce with
better management of water, nutrients, light
and temperature, lower pressure from pests
and diseases, greater yield per unit area, and
shorter life cycle (Resh 2022; Sharma et al.
2018).
Despite the many advantages, hydroponic
farming presents some constraints especially
regarding the high capital costs to establish
and operate these systems (Resh 2022). Addi-
tional challenges include the high costs of
fertilizers and the environmental risks associ-
ated with nutrient losses, especially in open
systems such as rockwool, sand, and sawdust
cultures that are incapable of recycling them
(Choi et al. 2011; Resh 2022). Among these
nutrients, phosphorus (P) is an element essen-
tial to plants that derive from nonrenewable
sources and is commonly associated with
eutrophication (Raghothama 1999). More-
over, the high global demand for phosphate
fertilizers and their price fluctuations cause
risks to farming operations due to increased
production costs (Sarvajayakesavalu et al.
2018). The drawbacks of P fertilizer use in
hydroponic systems can be mitigated by
breeding and adopting P-efficient cultivars
that produce similar yield in solutions with
lower nutrient inputs.
Phosphorus use efficiency (PUE), a con-
cept defined as higher capacity of plants to
produce economic yield per unit of applied P,
could reduce P inputs in crop production
while maximizing productivity (Fageria et al.
2017). P-efficient cultivars may present
higher capacity to absorb P from growth
medium due to improved morphological
and physiological mechanisms such as su-
perior root architecture and density (Lan
et al. 2015; Wen et al. 2019). Alternatively,
P-efficient cultivars may use internal P more
efficiently through higher capacity of internal
P transport, distribution, allocation, and remo-
bilization (Parentoni et al. 2012). For instance,
higher root-to-shoot (R–S) biomass ratio in
rye (Secale cereale L.) and wheat (Triticum
spp.) confer these crop species with superior P
uptake efficiency, whereas less efficient crops
such as bean (Phaseolus vulgaris L.), onion
(Allium cepa L.), and tomato (Lycopersicon
esculentum Mill.) tend to show low R–S ratios
(Raghothama 1999). In potato (Solanum tu-
berosum L.), PUE was correlated with total
plant biomass and total P uptake (Sanda~
na,
2016). Similarly, higher P uptake and P utili-
zation were associated with increased PUE in
mustard (Brassica juncea L.) (Aziz et al.
2006).
Hydroponic studies have been conducted
to understand the effect of P limitation on
physiological parameters and growth of
maize (Zea mays L.), sorghum [Sorghum bi-
color (L.) Moench], potato, rice (Oryza sativa
L.), and lettuce (Bera et al. 2018; Delaide
et al. 2016; Islam et al. 2019; Lee et al. 2021;
Neocleous and Savvas 2019; Nirubana et al.
2020; Sapkota et al. 2019). Nevertheless,
further research is needed to identify and
characterize lettuce accessions with higher
PUE in hydroponics. P-efficient lettuce ac-
cessions could benefit hydroponic growers by
reducing P inputs and/or improving the ef-
ficiency of P fertilizers applied to the nutri-
ent solutions, especially in systems like
Received for publication 12 Dec 2022. Accepted
for publication 3 Feb 2023.
Published online 17 Mar 2023.
We acknowledge the hatch project FLA-EREC-
005599. The Plant Breeding Graduate Initiative
from the Plant Breeding Working Group and the
Dean of Research Office of the University of Florida
Institute of Food and Agricultural Sciences. We
thank Heriberto Trevino for his help conducting the
experiments and Dr. Abul Rabbany for technical as-
sistance conducting P analyses in the Soil, Water,
and Nutrient Management Lab.
Current affiliation for G.F.K.: Department of Plant
Sciences, North Dakota State University, 1360 Al-
brecht Boulevard, Fargo, ND 58102, USA
G.V.S. is the corresponding author. E-mail:
gsandoyamiranda@ufl.edu.
This is an open access article distributed under the
CC BY-NC-ND license (https://creativecommons.
org/licenses/by-nc-nd/4.0/).
HORTSCIENCE VOL. 58(4) APRIL 2023 467
aquaponics where P solubility is reduced
due to high pH (Anderson et al. 2017).
There is already identified lettuce that re-
sponds differently to a 50% reduction of
the recommended P rate application in field
conditions (Kreutz et al. 2022); therefore,
genetic variation for PUE in greenhouse
can exist but warrants research. However,
PUE in field does not often correlate with
PUE in a greenhouse in a variety of crops
(Parentoni et al. 2012); consequently, ac-
cessions with higher PUE in field might
not be efficient in low P in greenhouse and
vice versa. In field, PUE is conditioned by
environmental factors (e.g., rainfall and
soil temperature), plant–soil interactions,
and pest, weed and disease incidence (Pa-
rentoni et al. 2012). In contrast to fields,
greenhouses allow for a more controlled
growing environment for lettuce, espe-
cially in terms of temperature, water avail-
ability, and pest and disease control.
Therefore, the objective of this study was
to identify and characterize lettuce acces-
sions capable of producing acceptable
yield in suboptimal P conditions. To iden-
tify elite lettuce genotypes, mimicked soil
solutions with low-P bioavailability were
employed; a set of lettuce accessions was
tested for PUE and PUE-related traits with
two P concentrations and two different P
sources in hydroponics.
Materials and Methods
Plant material. A set of 12 lettuce acces-
sions were tested for PUE in hydroponic
settings in a glasshouse at the University of
Florida Institute of Food and Agricultural
Sciences (UF/IFAS) Everglades Research
and Education Center (EREC), in Belle
Glade, FL. Most accessions, except for one
(BG19-0539), were previously tested for
PUE in field conditions and considered as
P-efficient or P-inefficient (Kreutz et al.
2022). Seeds of all accessions, that in-
cluded six romaine, five crisphead, and one
Latin (Table 1), were previously increased
by the UF/IFAS Lettuce Breeding Program.
Experiment description. Four experiments
(E1, E2, E3, E4) were conducted from
Oct through Dec 2020 (E1), from Dec 2020
through Feb 2021 (E2), from Dec 2021
through Jan 2022 (E3), and from Jan through
Mar 2022 (E4), respectively. Seeds of the 12
accessions were germinated in rockwool
cubes (Gro-Block, Grodan Rockwool B.V.,
The Netherlands), and seedlings were trans-
ferred to two nutrient film technique (NFT)
structures (CropKing Inc., Lodi, OH, USA)
at 10–16 d after sowing and at a plant density
of 26 plants/m
2
. The two NFT systems were
located inside of a glasshouse with natural
sunlight and semicontrolled temperature ad-
justed by an air-conditioner.
Both NFT systems were supplied with a
modified Howard Resh solution (Resh, 2022)
with the following composition: 7.5 mM
KNO
3
, 4 mM Ca(NO
3
)
2
·4H
2
O, 0.5 mM NH
4
NO
3
,0.5mMNaNO
3
, 2 mM MgSO
4
·7H
2
O,
50 mM KCl, 50 mMH
3
BO
3
,12mMMnSO
4
·4H
2
O, 2 mMZnSO
4
·7H
2
O, 1.5 mMCuSO
4
·5H
2
O, 0.1 mM(NH
4
)
6
Mo
7
O
24
·4H
2
O, and
10 mM NaFeEDTA. In E1 and E2, treatments
consisted of two P levels; each NFT system
was supplied with a unique concentration of
phosphoric acid (H
3
PO
4
): 0.1 mM (3.1 mg·L
1
)
or 1 mM (31 mg·L
1
), herein named low P
and high P, respectively. In E3 and E4, each
NFT system was supplied with a unique P
source: 1 mM NaH
2
PO
4
(monosodium phos-
phate; MSP) or 0.5 mM Ca
3
(PO
4
)
2
(tricalcium
phosphate; TCP; K
sp
52.7 × 10
33
). In E3, an
additional 5 mM of Ca was added to the TCP
treatment in the form of CaCl
2
to limit the
availability of P. Because differences in plant
biomass among the two P treatments were not
significant in E3, 10 mM of Ca was added
to the TCP treatment in E4 to reduce P
availability.
In all experiments, nutrient solutions were
prepared using deionized water obtained
from the Soil, Water, and Nutrient Manage-
ment Laboratory at EREC. Each NFT system
was connected to a 100-L reservoir contain-
ing the nutrient solution. The 12 lettuce
accessions in all experiments were arranged
in a completely randomized design with three
replicates on each P treatment, and each
replicate consisted of a single plant. Electrical
conductivity (EC) of the solutions was moni-
tored daily and maintained within a range
of 1.4 to 1.8 dS/m, and pH was adjusted to
6.0 ± 0.1 (optimal pH range for lettuce) by
adding hydrochloric acid (HCl) or sodium
hydroxide (NaOH). Nutrient solutions were
discarded, and tanks were recharged every 2
weeks to avoid any unbalance of ions.
Data collection. Accordingtohorticul-
tural maturity, plants were harvested at 51,
58, 42, and 41 d after sowing in E1, E2,
E3, and E4, respectively. Plants were eval-
uated for root and shoot fresh weight
(grams), root and shoot dry weight (DW,
grams). To measure shoot and root weight,
plants were harvested, separated into roots
and shoots, and weighed. Following har-
vest, roots and shoots were oven-dried at
65 Cfor5dtoobtainshootandrootDW.
Shoot and root tissue samples were then
subjected to total-P (TP) extraction to de-
termine P concentration.
Briefly, the TP extraction protocol con-
sisted of weighing 0.4 g of dried ground plant
tissue into a 20-mL glass scintillation vial.
Samples were then placed in a mufflefurnace
and burnt to ashes at 550 Cfor5h30min.
Once samples reached room temperature,
they were moistened by adding five drops of
deionized water. Each sample then received
2 mL of 6 M hydrochloric acid (HCl) and
was maintained at room temperature for 2 h.
The volume of each vial was then gaged to
20 mL, filtered with qualitative P5 filter paper
(12.5 cm in diameter), and transferred to
15-mL polypropylene test tubes. The total P
concentration of samples was determined us-
ing an inductively coupled plasma optical
emission spectrometer (ICP-OES Agilent
Technologies 5110, Santa Clara, CA, USA)
at the UF/IFAS Soil, Water, and Nutrient
Management Laboratory.
The following PUE parameters were esti-
mated for all accessions in each of the P treat-
ments: R–S biomass ratio, represented by the
proportion of root biomass relative to shoot
biomass; relative P uptake efficiency (PUpE;
mg P mg·L
1
P), characterized by the total
plant P content per unit of applied P;
Table 1. Lettuce accessions grown in four hydroponic experiments.
Accession Type PI number
i
Breeder PUE characterization
ii
Green Lightning Crisphead PI 599597 Progeny Advanced Genetics, Inc. P-inefficient
H1078 Crisphead N/A UF/IFAS P-efficient
Honcho II Crisphead PI 601591 Seminis Vegetable Seeds, Inc. P-efficient
Lantana Crisphead PI 658143 3 Star Lettuce, LLC P-inefficient
Sun Devil Crisphead PI 603974 Progeny Advanced Genetics, Inc. P-efficient
60183 Romaine N/A UF/IFAS P-efficient
BG19–0539 Romaine N/A UF/IFAS N/A
Floricos 83 Romaine N/A UF/IFAS P-inefficient
Manatee Romaine PI 641790 3 Star Lettuce, LLC P-inefficient
Okeechobee Romaine PI 658142 3 Star Lettuce, LLC P-efficient
Valmaine Romaine PI 543959 UF/IFAS P-inefficient
Little Gem Latin PI 617959 Vilmorin, S.A. P-efficient
i
Plant introduction number obtained from US Department of Agriculture, National Plant Germplasm System (https://npgsweb.ars-grin.gov/gringlobal/
search).
ii
Characterization for phosphorus use efficiency based on Kreutz et al. (2022). Accessions classified as P-efficient had a head weight reduction of 20% re-
duction at half-P rate compared with standard-P rate in field experiments.
N/A 5not available; UF/IFAS 5University of Florida Institute of Food and Agricultural Sciences.
468 HORTSCIENCE VOL. 58(4) APRIL 2023
P utilization efficiency (PutE), characterized
by the total biomass produced per unit of ab-
sorbed P (g DW mg·P
1
), as described by
Hammond et al. (2009) and Neto et al.
(2016) (Table 2).
Statistical analyses. For E1 and E2, analy-
ses of variance (ANOVAs) for shoot weight,
root weight, shoot TP, and root TP were
performed among accessions, P treatments,
experiments, and their respective interactions.
All factors were treated as fixed effects. Be-
cause the concentration of Ca added to the
TCP treatment in E4 was twice the Ca con-
centration used in E3, data from E3 and E4
were analyzed separately. Thus, ANOVA for
shoot weight, root weight, shoot TP, and root
TP was performed among accessions, P treat-
ments, and the accession × P treatment inter-
action. All factors were considered as fixed
effects. For all experiments, an additional
ANOVA was conducted to identify differ-
ences in PUpE and PUtE following the model
previously described.
Accessions were then pairwise compared
(ttests) to identify nonsignificant differences
for each trait among P rates. One accession
was considered as P-efficient when its head
weight reduction under P stress was less than
20% compared with the optimal P treatment.
In all analyses, least square means were
generated using the lsmeans statement and dif-
ferences for accessions within each treatment
were determined by using Fisher’s protected
least significantdifferencetestatalevelofsig-
nificance of P50.05. Pearson correlation
coefficients were calculated between shoot
weight reduction, R–S biomass ratio, PUpE,
and PUtE for each of the two P treatments.
The coefficients were based on genotypic
means across experiments and replicates. All
analyses were carried out using GLIMMIX
and CORR procedures in SAS software ver.
9.4 (SAS Institute Inc., Cary, NC, USA).
Results
Lettuce responds to different phosphorus
levels (E1 and E2). Applying only 10% of the
optimal P level in the P solution aided to
characterize germplasm that produced similar
yield at low P compared with high P treat-
ment in the NFT system. Significant differ-
ences (P<0.05) were detected for shoot and
root weight among accession (G), P treatment
(T), experiment (E), and the interactions G×E
and T×E (Supplemental Table 1). The G×T
interaction was found to be slightly signifi-
cant (P50.0434) for shoot weight, but not
forrootweight(P50.3033) (Supplemental
Table 1).
Although lettuce shoot weight decreased
when plants were grown at low P concentra-
tion, five lettuce accessions (‘Little Gem’,
60183, ‘Valmaine’,BG19-0539,and‘Green
Lightning’) presented a shoot weight reduc-
tion of 20% or less with low P as compared
with high P. ‘Little Gem’and 60183 were the
accessions with the most similar shoot weight
across the two treatments (shoot weight re-
duction of –6% and 11%, respectively)
(Table 3; Fig. 1).
In contrast to shoot weight, the reduction in
P concentration from 31 to 3.1 mg·L
1
caused
a highly significant increase (P<0.0001) in
the overall root biomass in lettuce. Five acces-
sions (H1078, ‘Honcho II’,‘Little Gem’,
‘Okeechobee’,‘Sun Devil’) showed similar
(P>0.75) root weight in low and high P
(Table 3; Fig. 1). At low P, the romaine breed-
ing line 60183 and cv. Valmaine had a signifi-
cantly higher root weight compared with high
Ptreatment(P50.0040 and P50.0086, re-
spectively), indicating root growth promotion
of these accessions when subjected to P stress
(Fig. 1).
The 10% of P in the nutrient solution led to
significant differences in shoot TP among acces-
sions (P<0.0001), P treatments (P<0.0001),
experiments (P<0.0001), and the interactions
G×E (P50.0002), T×E (P<0.0001), and
G×T×E (P<0.0001). Meanwhile, P treatment
(P<0.0001), experiment (P50.0179), and
the interaction T×E (P<0.0001) were the only
statistically significant factors for root TP of
lettuce (Supplemental Table 1). With low P, all
accessions showed a significant (P<0.05) re-
duction in shoot TP compared with high P treat-
ment. ‘Little Gem’showed the greatest shoot
TP concentration with both, low and high P
(Supplemental Fig. 1). These results indicate the
presence of genetic variation for shoot TP con-
centration in lettuce grown under suboptimal
and optimal conditions. Considering root TP, a
significant decrease was observed under low P
vs. high P for most accessions, except for ‘Flori-
cos 83’and ‘Lantana’(P50.0890 and P5
0.0752, respectively) (Supplemental Fig. 1).
While no significant differences (P>0.05) in
root TP were detected among accessions at low
P treatment, accessions ‘Sun Devil’,‘Green
Lightning’, and 60183 had the greatest root TP
concentration at high P (Supplemental Fig. 1).
In lettuce, PUE is dependent on multiple
plant traits, including root morphology,
PUpE, PUtE. P-efficient accessions tend to
Table 2. Definition, abbreviation, formula, and unit of phosphorus use efficiency parameters estimated for 12 lettuce accessions grown in four hydroponic
experiments.
Measurement Abbreviation Formula
i
Unit
Root-to-shoot ratio R–S ratio (Root DW) / (Shoot DW) –
Phosphorus uptake efficiency PUpE (Tissue TP × Plant DW) / (P
i
applied) mg P mg·L
1
P
Phosphorus utilization efficiency PUtE (Plant DW) / (Tissue TP × Plant DW) g DW mg·P
1
i
Tissue total-P estimated on the basis of dry weight (DW).
Table 3. Shoot and root weight reduction (%) and respective standard error and Pvalues, root-to-shoot (R–S) biomass ratio, P uptake efficiency (PUpE,
mg P mg·L
1
P applied), and P utilization efficiency (PUtE, g DW mg
1
P) of 12 lettuce accessions grown under low and high P in E1 and E2.
Accession Shoot wt reduction
i,iii
Pvalue Root wt reduction
i,iii
Pvalue
R–S ratio
ii,iii
PUpE
iii
PUtE
iii
Low P High P Low P High P Low P High P
60183 11 ± 17 0.5382 59 ± 18 0.0040 0.47 bc 0.29 d 10.33 a 2.06 0.31 g 0.13
BG19–0539 17 ± 17 0.3104 24 ± 18 0.0813 0.52 ab 0.42 bcd 7.45 cd 2.46 0.47 abc 0.14
Floricos 83 32 ± 17 0.0453 30 ± 18 0.1093 0.49 bc 0.32 cd 6.25 d 2.15 0.51 ab 0.17
Green Lightning 20 ± 17 0.2870 36 ± 18 0.0527 0.64 a 0.51 b 6.48 cd 2.01 0.46 bcd 0.12
H1078 36 ± 17 0.0056 4 ± 18 0.8093 0.43 bc 0.35 cd 6.87 cd 2.52 0.43 cde 0.14
Honcho II 22 ± 17 0.1333 5 ± 18 0.7722 0.36 c 0.38 bcd 7.01 cd 2.04 0.38 ef 0.14
Lantana 28 ± 17 0.0473 39 ± 18 0.0657 0.48 bc 0.39 bcd 6.57 cd 1.75 0.47 abc 0.16
Little Gem 6 ± 17 0.8514 1 ± 18 0.9487 0.52 ab 0.68 a 6.55 cd 1.41 0.33 fg 0.13
Manatee 24 ± 17 0.0432 13 ± 18 0.2563 0.50 abc 0.39 bcd 9.56 ab 3.13 0.42 cde 0.13
Okeechobee 55 ± 17 <0.0001 1 ± 18 0.9154 0.51 ab 0.36 cd 6.62 cd 3.18 0.53 a 0.16
Sun Devil 27 ± 17 0.0929 1 ± 18 0.9436 0.51 ab 0.41 bcd 6.74 cd 2.08 0.40 de 0.15
Valmaine 14 ± 17 0.3443 34 ± 18 0.0086 0.50 ab 0.44 bc 8.34 bc 2.84 0.47 abc 0.14
Average 23 <0.0001 20 <0.0001 0.49 0.42 7.40 2.30 0.43 0.14
i
Negatives values indicate higher weight at low P compared with high P treatment.
ii
Root-to-shoot biomass ratio estimated on a dry weight basis.
iii
Values in the same column followed by lowercase letters each column indicate significant differences (P<0.05) among accessions. No letters indicate
nonsignificant difference (P>0.05) according to least significant difference protected test.
DW 5dry weight.
HORTSCIENCE VOL. 58(4) APRIL 2023 469
produce greater root biomass when grown in
P-deficient conditions, as noted by the greater
root weight of lettuce plants with low P com-
pared with high P. As a consequence, 10 of
12 accessions showed a greater R–S biomass
ratio at low P. Among these, the crisphead
‘Green Lightning’had the greatest R–Sratio
under low P, indicating that this accession re-
sponded the most to the low-P stress by in-
creasing its root biomass relative to the shoot
biomass (Table 3).
It was also found that accessions differed
significantly (P<0.05) in PUpE and PUtE at
low P treatment. However, PUpE and PUtE
were associated with specific accessions. For
instance, the romaine breeding line 60183
and ‘Manatee’were the most efficient acces-
sions at absorbing P from the solution
(PUpE), whereas romaine ‘Okeechobee’and
‘Floricos 83’used internal P more efficiently
(PUtE) (Table 3).
The shoot weight reduction of plants under
low P relative to high P was negatively corre-
lated with R–S ratio (r5–0.62; P50.0318)
in this study. This suggests that accessions
with greater R–S ratio in optimal P conditions
are less affected by shoot biomass reduction
when grown with low P. In contrast, shoot
weight reduction was found to be positively
correlated with PUpE at high P (r50.58;
P50.0496), and with PUtE at low P (r50.70;
P50.0116) and high P (r50.62; P50.0321).
These results indicate that the reduction
in shoot weight of lettuce grown under
P-deprived conditions is inversely proportional
to the capacity of plants to absorb P (under
high P conditions) and/or use P internally. Ad-
ditionally, a positive and slightly significant
correlation between PUtE of lettuce grown
at low and high P treatments was observed
(r50.60; P50.0412). At low P treatment, a
negative, nonsignificant correlation (r5–0.45;
P50.1428) between PUpE and PUtE was
observed. The lack of correlation between
PUpE and PUtE indicates that these two pa-
rameters may be driven by independent mecha-
nisms in lettuce.
Lettuce responds to different phosphorus
sources (E3 and E4). Although differences
were found to be significant for both shoot
and root weight (P<0.0001 and P<0.0001,
respectively) among accessions, the G×T in-
teraction was not significant for both traits
(P>0.05) in E3. The utilization of TCP as a
source of P plus an additional 5 mM of Ca did
not significantly (P>0.05) affect the shoot
and root weight of lettuce (Supplemental
Table 2; Supplemental Fig. 2). All the 12
tested accessions produced statistically the
same (P>0.05) shoot weight under the TCP
and MSP treatments (Fig. 2).
Only slight differences were found for
root weight in the romaine accessions BG19-
0539 and ‘Valmaine’, which produced less
root biomass when grown in the TCP treat-
ment (P50.0181 and P50.0494, respec-
tively); the opposite was true for the romaine
breeding line 60183 that had significantly
higher root weight when cultivated in the
TCP treatment (P50.0262) (Fig. 2). Like-
wise, no significant differences (P>0.05)
were detected for shoot TP and root TP
among accessions, P treatments, and the G×T
interaction in E3 (Supplemental Table 2).
Most accessions presented statistically the
same (P>0.05) shoot TP and root TP under
both P treatments, except for the romaine
cv. Okeechobee that had significantly less
shoot TP at the TCP treatment (Supplemental
Fig. 3).
Due to the nonsignificant differences
between the two initial P sources, the concen-
tration of Ca added to the TCP treatment was
raised to 10 mM in E4. The addition of
10 mM of Ca caused the shoot weight and root
weight of lettuce to drastically decrease (P<
0.0001 and P<0.0001, respectively) com-
pared with the MSP treatment (Supplemental
Table 2; Supplemental Fig. 2). As a conse-
quence, all 12 accessions experienced signifi-
cant (P<0.05) shoot weight reduction when
grown under TCP (Fig. 3). The cv. Manatee,
cv. Little Gem, and breeding line BG19-0539
showed a highly similar (P>0.50) root weight
under TCP and MSP treatments (Fig. 3). In
contrast, five lettuce accessions (‘Green
Lightning’, H1078, ‘Lantana’,‘Okeechobee’,
and ‘Sun Devil’) yielded significantly less
(P<0.05)rootweightwhencultivatedin
TCP vs. MSP (Fig. 3).
The addition of extra Ca into the TCP
treatment in E4 significantly influenced
lettuce tissue P concentration, as observed by
the differences in shoot TP and root TP
among accessions (P<0.05) and P treatment
(P<0.05) (Supplemental Table 2). The G×T
interaction was found to be nonsignificant
(P>0.05) for both shoot TP and root TP.
All accessions showed a significant reduction
in shoot TP when grown under the TCP
source compared with the MSP treatment
(Supplemental Fig. 4). Meanwhile, only
three accessions (60183, ‘Little Gem’,and
0
50
100
150
200
250
Root - Low P Root - High P Shoot - Low P Shoot - High P
a
a
a
b
a
a
a
b
a
a
a
b
a
a
a
b
b
a
a
a
a
a
A
B
AAAAAAAAA
AAA
AAAAA
AAB
a
a
A
A
Fresh weight (g/plant)
Fig. 1. Least square means of shoot and root weight (g/plant) with 95% confidence intervals of 12 let-
tuce accessions grown under low and high P in E1 and E2. Shoot weight means with different
lowercase letters and root weight means with different uppercase letters within an accession are
significantly different at P#0.05 using the least significant difference test.
0
50
100
150
200
250
Root - TCP Root - MSP Shoot - TCP Shoot - MSP
a
aa
a
a
a
a
aa
AA
aa a
a
a
a
a
aa
a
a
a aa
a
B B
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
Fresh weight (g/plant)
Fig. 2. Least square means of shoot and root weight (g/plant) with 95% confidence intervals of 12 let-
tuce accessions grown under tricalcium phosphate (TCP) and monosodium phosphate (MSP) treat-
ments in E3. Shoot weight means with different lowercase letters and root weight means with
different uppercase letters within an accession are significantly different at P#0.05 using the least
significant difference test.
470 HORTSCIENCE VOL. 58(4) APRIL 2023
‘Manatee’) experienced a significant de-
crease in root TP in the roots under the
TCP treatment (P50.0294, P50.0228,
and P50.0063, respectively) (Supplemen-
tal Fig. 4).
Lettuce plants cultivated in the TCP treat-
ment did not present a higher root growth as
nonsignificant differences in the overall R–S
ratio of lettuce were observed between the
two P sources (TCP and MSP) in both experi-
ments, E3 (P50.1615) and E4 (P5
0.1669). However, significant (P<0.05) ge-
netic variability for R–Sratiowasfound
within the germplasm evaluated in the TCP
treatment (Tables 4 and 5). In E4, crisphead
cv. Honcho II obtained the highest R–Sratio
among all accessions, whereas the romaine
breeding lines 60183 and BG19-0539 had the
lowest R–S ratio (Table 5). These results in-
dicate that under the TCP treatment, some
accessions were more responsive to the lower
P availability in the solution than others, as
observed by the higher root biomass relative
to the shoot biomass.
In E4, lettuce accessions significantly
(P<0.05) differed in P uptake efficiency
(PUpE) only under the MSP treatment. De-
spite the genetic variability in PUpE under
optimum P conditions, TCP with 10 mM of
Ca caused severe P stress to the plants,
likely leading to nonsignificant differences
(P>0.05) in P uptake among accessions.
Under MSP, the crisphead cv. Green Light-
ning and the romaine cv. Manatee had the
highest PUpE scores, indicating that these
accessions were able to uptake more P un-
der optimum conditions than the other ac-
cessions. The crisphead cv. Honcho II had
the lowest PUpE among all the accessions
(Table 5). No significant differences (P>
0.05) in PUtE were detected among acces-
sions in both MSP and TCP treatments
(Table 5).
In both experiments (E3 and E4), PUpE
and PUtE of lettuce accessions grown with
the MSP treatment were found to be nega-
tively correlated (r5–0.68; P50.0304 in
E3 and r5–0.62; P50.0327 in E4). In
addition, the lack of correlation between both
parameters under TCP treatment in E3 and
E4 could be an affirmation that PUpE and
PUtE might be controlled by independent
mechanisms in lettuce when grown under
suboptimal P conditions.
In E3 and E4, shoot weight reduction was
not correlated with R–S ratio, PUpE, or PUtE,
regardless of the P solution in which plants
were grown. Nevertheless, R–S ratio of lettuce
was found to be negatively correlated (r5
–0.67; P50.0168) with PUpE at the TCP
treatment in E4, suggesting that accessions
with greater root biomass relative to the shoot
biomass had a smaller P uptake efficiency,
and vice versa. In E4, PUpE under TCP and
MSP were positively correlated (r50.67;
P50.0169), denoting that those accessions
with superior P uptake in optimal P conditions
also performed well when P availability was
reduced.
Discussion
Lettuce response to different phosphorus
levels. Genetic variation for PUE was identi-
fied in lettuce grown with different P levels
and P sources, as observed by the discrepan-
cies in the percentage of shoot weight reduc-
tion between P treatments and the significant
differences in PUpE and PUtE among the
accessions tested. Applying only 10% of the
optimal P recommended concentration led to
a decrease in the shoot weight of lettuce. In-
traspecific genotypic variation in shoot
weight was detected in five cultivars and
breeding lines producing similar shoot weight
Table 4. Shoot and root weight reduction (%) and respective standard error and Pvalues, root-to-shoot (R–S) biomass ratio, P uptake efficiency (PUpE,
mg P mg·L
1
P applied), and P utilization efficiency (PUtE, g DW mg·P
1
) of 12 lettuce accessions grown under tricalcium phosphate (TCP) and
monosodium phosphate (MSP) treatments in E3.
Accession Shoot wt reduction
i,iii
Pvalue Root wt reduction
i,iii
Pvalue
R–S ratio
ii, iii
PUpE
iii
PUtE
iii
TCP MSP TCP MSP TCP MSP
60183 9 ± 19 0.5949 88 ± 46 0.0262 0.30 b 0.12 1.78 abc 0.65 0.10 0.35
BG19–0539 22 ± 19 0.0991 43 ± 46 0.0181 0.19 cd 0.25 0.94 cd 1.69 0.18 0.28
Floricos 83 5 ± 19 0.7908 24 ± 46 0.3971 0.26 bc 0.24 0.90 bcd –
iv
0.20 –
iv
Green Lightning 6 ± 19 0.7880 46 ± 46 0.1484 0.44 a 0.25 1.08 cd 0.86 0.12 0.27
H1078 6 ± 19 0.6814 30 ± 46 0.1883 0.25 bc 0.24 2.11 a 1.59 0.10 0.15
Honcho II 20 ± 19 0.3893 2 ± 46 0.9619 0.12 d 0.28 0.82 d 0.38 0.16 0.20
Lantana 14 ± 19 0.3735 25 ± 46 0.2834 0.31 b 0.25 1.34 bcd 0.47 0.13 0.34
Little Gem 2 ± 19 0.9164 4 ± 46 0.9088 0.24 bc 0.26 0.75 d –
iv
0.18 –
iv
Manatee 6 ± 19 0.6507 16 ± 46 0.2751 0.27 b 0.27 1.84 ab 1.69 0.15 0.15
Okeechobee 4 ± 19 0.7278 16 ± 46 0.2473 0.28 b 0.22 1.34 bcd 2.63 0.28 0.14
Sun Devil 5 ± 19 0.7558 1 ± 46 0.9848 0.23 bc 0.19 1.42 abcd 1.42 0.13 0.14
Valmaine 8 ± 19 0.6520 28 ± 46 0.0494 0.28 b 0.35 1.58 abc 2.01 0.14 0.13
Average 2 0.5098 15 0.1640 0.26 0.24 1.33 1.34 0.16 0.22
i
Negatives values indicate higher shoot weight at TCP compared with MSP treatment.
ii
Root-shoot biomass ratio estimated on a dry weight basis.
iii
Values in the same column followed by lowercase letter within each column indicate significant differences (P<0.05) among accessions. No letters in-
dicate nonsignificant difference (P>0.05) according to least significant difference protected test.
iv
Data not available due to insufficient biomass for tissue-P analysis.
DW 5dry weight.
0
50
100
150
200
250
Fresh weight (g/plant)
Root - TCP Root - MSP Shoot - TCP Shoot - MSP
a
b
a
a
a
a
a
a
a
aa
aa
bbbb
b
b
bbbbb
A
BA
AA
A
A
BB
A
AA
A
BAA
AA A
BB
AA
A
Fig. 3. Least square means of shoot and root weight (g/plant) with 95% confidence intervals of 12 let-
tuce accessions grown under tricalcium phosphate (TCP) and monosodium phosphate (MSP) treat-
ments in E4. Shoot weight means with different lowercase letters and root weight means with
different uppercase letters within an accession are significantly different at P#0.05 using the least
significant difference test.
HORTSCIENCE VOL. 58(4) APRIL 2023 471
in P-limited conditions. This suggests that let-
tuce germplasm responded differently to the
reduction in P availability in greenhouse, as
described in lettuce grown in field (Kreutz
et al. 2022). Breeding line 60183 and ‘Little
Gem’produced similar head weight under
half-P and standard-P rates in field and green-
house conditions and are considered
P-efficient (Kreutz et al. 2022). ‘Okeechobee’
experienced the greatest reduction in shoot
weight when grown in low P in hydroponics,
contrasting results from field experiments
(Kreutz et al. 2022). These findings indicate
that PUE of lettuce in specific accessions
varies according to the environment and
growing conditions. Likewise, PUE in wheat,
sorghum, and maize is environment-dependent
(Parentoni et al. 2012).
Asignificant increase in root weight of
lettuce when grown at low P observed in this
research is a response of plants to the lower P
availability, leading to a greater R–Sratioin
lettuce (Bertossi et al. 2013; Neocleous and
Savvas 2019). However, the R–S ratio was
not correlated to PUpE at low P treatment,
suggesting that P uptake of lettuce plants was
not improved by the increases in root weight.
This likely occurred due to limited amounts
of P in the solution, which in turn, precluded
the continuous P uptake by roots after all the
P was fully depleted from the solution (Lee
et al. 2021; van de Wiel et al. 2016). As op-
posed to these results, the increase of root
biomass often contributes to a greater P scav-
enging capacity of plants, and consequently,
greater P uptake in fieldconditions(vande
Wiel et al. 2016).
Some lettuce accessions have the capacity
to uptake more P than others. All 12 acces-
sions were in the same NFT system and grown
in the same nutrient solution in the experi-
ments, and therefore the differences observed
in PUpE can be attributed to the genetic
makeup of each accessions. This hypothesis
could explain the significantly greater PUpE
observed for accessions 60183 and ‘Manatee’
at low P treatment. Nevertheless, 60183 and
‘Manatee’were considered P-efficient and
P-inefficient accessions in previous field eval-
uations, suggesting that additional biological
(morphological and/or physiological traits)
and environmental factors may be involved in
PUE in lettuce (Kreutz et al. 2022).
Lettuce response to different phosphorus
sources. In an initial experiment, lettuce ac-
cessions were subjected to TCP [Ca
3
(PO
4
)
2
]
containing an extra 5 mM of Ca to mimic a
growing medium where P is present in non-
bioavailable forms. The bioavailability of
P in the solution depends on the equilibrium
of Ca
12
and PO
4
3
ions [i.e., solubility prod-
uct (K
sp
)] (Lee et al. 2021). When supple-
mental Ca is absent or low, sufficient P is
mobilized from TCP in the nutrient solution;
therefore, the TCP solution had enough P to
support lettuce growth in E3. Consequently,
the discrimination of P-efficient lettuce acces-
sions was not possible, similarly to studies
conducted in wheat, in which the shoot bio-
mass was not reduced when plants were
grown under a treatment containing TCP
combined with 10 mM of extra Ca (Liu et al.
2007). The method has proven to be efficient
at detecting PUE in crops as genotypic vari-
ability for PUE was detected in Indian mus-
tard, potato, and wheat grown in solutions
containing TCP with the addition of Ca (Aziz
et al. 2006; Bera et al. 2020; Lee et al. 2021;
Liu et al. 2007). Thus, each species responds
differently to different levels of P solubility
in the growing medium.
In a follow-up experiment, the concentra-
tion of extra Ca added to the TCP treatment
was increased to 10 mM. The extra Ca added
in E4 likely decreased the solubility of P to
a point in which plants were no longer able
to achieve adequate shoot growth and fa-
vored the precipitation of P. In turn, shoot
weight, root weight, shoot TP, and root TP
of lettuce accessions substantially de-
creased, as observed in wheat (Akhtar et al.
2016). In addition to higher P precipitation,
the extra 10 mM of Ca could have contri-
buted to a Ca-Mg or Ca-K antagonism (in-
hibition of Mg or K uptake due to high Ca
concentration), as previously seen in Vitis
vinifera L. grown hydroponically (Garcia
et al. 1999). Consequently, the uptake of
Mg and/or K could have been reduced,
which would result in lower lettuce yield.
In this research, lettuce had different levels
of internal P utilization (PUtE) at low P.
Interestingly, accessions with the greatest
PUtE values experienced the greatest shoot
weight reduction. The negative correlations
between PUtE and lettuce shoot weight could
be an indication of greater P starvation rather
than superior P utilization by lettuce plants, as
previously reported in rice grown with moderate
to extreme low P levels (Rose and Wissuwa
2012; Wissuwa et al. 1998).
Phosphorus levels versus phosphorus
sources. In this study, lettuce accessions were
tested for PUE under two P levels (low and
high P) and two P sources (MSP and TCP).
The former approach has successfully al-
lowed the identification of two lettuce
accessions (‘Little Gem’and 60183) that pro-
duced highly similar shoot weight under low
and high P. A similar technique has been pre-
viously used to evaluate potato and soybean
[Glycine max (L.) Merr.] accessions in hydro-
ponics (Lee et al. 2021; Ochigbo and Bello,
2014). A possible disadvantage of testing
plants for PUE under low P concentrations is
the complete depletion of P in the nutrient so-
lution upon plant uptake, which may inhibit
root hair formation as reported in Arabidopsis
thaliana L. (Liu et al. 2006). While root hairs
were not closely examined in this study, five
lettuce accessions produced similar root
weight under both P levels, suggesting that
their root growth was not inhibited by the
low P treatment.
In the second approach, lettuce accessions
were tested under different P sources (MSP
and TSP combined with Ca). This method
has been previously applied to test mustard,
Table 5. Shoot and root weight reduction (%) and respective standard error and Pvalues, root-to-shoot (R–S) biomass ratio, PUpE (mg P mg·L
1
P applied), and PUtE (g DW mg P
1
) of 12 lettuce accessions grown under tricalcium phosphate (TCP) and monosodium phosphate (MSP)
treatments in E4.
Accession Shoot wt reduction
i,iii
Pvalue Root wt reduction
i,iii
Pvalue
R–S ratio
ii,iii
PUpE
iii
PUtE
iii
TCP MSP TCP MSP TCP MSP
60183 57 ± 7 <0.0001 37 ± 23 0.1923 0.12 c 0.14 0.48 1.78 bc 0.18 0.10
BG19–0539 48 ± 7 0.0013 16 ± 23 0.5010 0.10 c 0.30 0.68 1.70 bc 0.19 0.12
Floricos 83 58 ± 7 <0.0001 24 ± 23 0.3371 0.14 bc 0.31 0.59 1.41 c 0.21 0.12
Green Lightning 71 ± 7 <0.0001 36 ± 23 0.0103 0.18 bc 0.18 0.95 3.05 a 0.16 0.11
H1078 77 ± 7 <0.0001 54 ± 23 0.0011 0.36 bc 0.16 0.42 2.10 b 0.14 0.12
Honcho II 76 ± 7 0.0117 44 ± 23 0.2157 0.75 a 0.33 0.07 0.62 d 0.14 0.14
Lantana 59 ± 7 <0.0001 38 ± 23 0.0068 0.19 bc 0.31 0.52 2.08 b 0.19 0.11
Little Gem 58 ± 7 0.0006 18 ± 23 0.5095 0.48 ab 0.22 0.35 1.46 c 0.17 0.09
Manatee 73 ± 7 <0.0001 3 ± 23 0.8661 0.44 abc 0.21 0.53 2.91 a 0.20 0.09
Okeechobee 73 ± 7 <0.0001 44 ± 23 0.0069 0.21 bc 0.20 0.40 2.08 b 0.23 0.12
Sun Devil 76 ± 7 <0.0001 52 ± 23 0.0077 0.21 bc 0.26 0.31 1.90 bc 0.19 0.12
Valmaine 69 ± 7 <0.0001 33 ± 23 0.0765 0.32 bc 0.24 0.47 2.16 b 0.17 0.11
Average 66 <0.0001 33 <0.0001 0.29 0.24 0.48 1.94 0.18 0.11
i
Negatives values indicate higher shoot weight at TCP compared with MSP treatment.
ii
Root-to-shoot biomass ratio estimated on a dry weight basis.
iii
Values in the same column followed by lowercase letter within each column indicate significant differences (P<0.05) among accessions. No letters in-
dicate nonsignificant difference (P>0.05) according to least significant difference protected test.
DW 5dry weight.
472 HORTSCIENCE VOL. 58(4) APRIL 2023
wheat, and potato accessions for PUE
(Akhtar et al. 2016; Aziz et al. 2006; Lee
et al. 2021; Liu et al. 2007). The proper
Ca/TSP ratio varies according to the crop species
as intra- and interspecific variation exists for
traits such as P uptake, P mobilization, and Ca
uptake (Akhtar et al. 2016; Lee et al. 2021). For
instance, species with greater Ca uptake led to
more P release in TCP solutions (Lee et al.
2021). In this study, 5 mM of Ca added to the
TSP solution was insufficient to cause yield dif-
ferences in lettuce. In a second trial, 10 mM of
Ca added to the TCP solution led to significant
yield reductions in lettuce. Despite the severe P
stress, ‘Little Gem’and 60183 were among the
accessions with the least shoot weight reduction
in the TCP treatment, confirming the results ob-
served in the low and high P trials. This indicates
that the two approaches resulted in similar find-
ings regardless of the magnitude of the P defi-
cient stress caused by the treatments. However,
further investigations should be conducted to de-
termine the proper Ca/TCP ratio that will allow
the identification of P-efficient lettuce accessions
in hydroponics.
Conclusions
Despite previous efforts to investigate the
effects of P limitation in hydroponic lettuce,
the identification and characterization of let-
tuce accessions grown with different P con-
centrations and sources remained unexplored
before this study. We identified lettuce acces-
sions with superior PUE in hydroponics that
may be used for breeding new elite cultivars
adaptive to suboptimum P conditions. In this
research, genetic variation for PUE was de-
tected in lettuce accessions grown hydroponi-
cally when 10% of P was used in the growing
solution. Accessions 60183 and ‘Little Gem’
are considered efficient when grown at 10%
of the optimal P concentration in hydroponics.
Shoot biomass of lettuce was unaffected when
the TCP had an extra 5 mM of Ca. In contrast,
drastic reductions in yield were observed
when lettuce accessions were grown at TCP
accompanied by an extra 10 mM of Ca, hin-
dering the discrimination of P-efficient and P-
inefficient accessions. Future research should
investigate proper ratios of Ca (between 5 and
10 mM) and P in nutrient solutions that allow
the discrimination of P-efficient lettuce acces-
sions under TCP conditions.
In addition, genotypic variation for R–S
ratio (an indicative of P-stress response),
PUpE, and PUtE was detected across the
different P treatments and offers an oppor-
tunity to use these parameters for selecting
lettuce for PUE breeding. Although, these
PUE-related traits require the measurement
of additional characteristics such as fresh
and dry weight, and P concentration in
shoots and roots that might difficult the use
of these characteristics for selection. In-
stead, the detection of characteristics highly
associated with PUE-related traits can aid
the selection of P-efficient lettuce acces-
sions for hydroponic production. The iden-
tification of indirect traits linked to PUE
should facilitate and expedite the develop-
ment of new P-efficient lettuce cultivars.
References Cited
Ahmed ZFR, Alnuaimi AKH, Askri A, Tzortzakis
N. 2021. Evalu ation of lettu ce (Lactuca
sativa L.) production under hydroponic system:
Nutrient solution derived from fish waste vs. in-
organic nutrient solution. Horticulturae. 7(9):292.
https://doi.org/10.3390/horticulturae7090292.
Akhtar MS, Oki Y, Nakashima Y, Adachi T, Nishi-
gaki M. 2016. Phosphorus stress-induced differ-
ential growth, and phosphorus acquisition and
use efficiency by spring wheat cultivars. Com-
mun Soil Sci Plant Anal. 47(Suppl.):15–27.
https://doi.org/10.1080/00103624.2016.1232089.
Anderson TS, Villiers DD, Timmons MB. 2017.
Growth and tissue elemental composition re-
sponse of butterhead lettuce (Lactuca sativa,
cv. Flandria) to hydroponic and aquaponic con-
ditions. Horticulturae. 3(3):43. https://doi.org/
10.3390/horticulturae3030043.
Aziz T, Maqsood MA, Tahir MA, Ahmad I,
Cheema MA. 2006. Phosphorus utilization by
six brassica cultivars (Brassica juncea L.) from
tri-calcium phosphate: A relatively insoluble P
compound. Pak J Bot. 38(5):1529–1538.
Bera T, McLamore ES, Wasik B, Rathinasaba-
pathi B, Liu G. 2018. Identification of a
maize (Zea mays L.) inbred line adapted to
low-P conditions via analyses of phosphorus
utilization, root acidification, and calcium in-
flux. J Plant Nutr Soil Sci. 181:275–286.
https://doi.org/10.1002/jpln.201700319.
Bera T, Song F, Liu G. 2020. Rapid identifica-
tion of phosphorus-efficient genotypes from
commercially grown tomato (Solanum lycopersi-
cum L.) varieties in a simulated soil solution. J
Hortic Sci Biotechnol. 95(3):395–404. https://
www.tandfonline.com/doi/full/10.1080/14620316.
2019.1684210.
Bertossi APA, Thomazini A, Fonseca AS, Amaral
JFT. 2013. Nutritional efficiency of phosphorus
in lettuce. J Agric Sci. 5(8):125–131. https://
doi.org/10.5539/jas.v5n8p125.
Choi B, Lee SS, Ok YS. 2011. Effects of waste
nutrient solution on growth of Chinese cabbage
(Brassica campestris L.) in Korea. Korean J
Environ Agric. 30(2):125–131. https://doi.
org/10.5338/KJEA.2011.30.2.125.
Delaide B, Goddek S, Gott J, Soyeurt H, Jijakli
MH. 2016. Lettuce (Lactuca sativa L.) growth
performance in complemented aquaponic solu-
tion outperforms hydroponics. Water. 8:467.
https://doi.org/10.3390/w8100467.
Fageria NK, He Z, Baligar VC. 2017. Phosphorus
management in crop production. Boca Raton,
FL: CRC Press. https://doi.org/10.1201/
9781315162096.
Garcia M, Daverede C, Gallego P, Toumi M.
1999. Effect of various potassium-calcium
ratios on cation nutrition of grape grown hy-
droponically. J Plant Nutr. 22(3):417–425.
https://doi.org/10.1080/01904169909365639.
Hammond JP, Broadley MR, White PJ, King GJ,
Bowen HC, Hayden R, Meacham MC, Mead
A, Overs T, Spracklen WP, Greenwood DJ.
2009. Shoot yield drives phosphorus use effi-
ciency in Brassica oleracea and correlates with
root architecture traits. J Expt Bot. 60(7):
1953–1968. https://doi.org/10.1093/jxb/erp083.
Islam MZ, Lee Y, Mele MA, Choi I, Kang H.
2019. The effect of phosphorus and root zone
temperature on anthocyanin of red romaine
lettuce. Agronomy (Basel). 9(2):47. https://doi.
org/10.3390/agronomy9020047.
Kreutz GF, Bhadha JH, Sandoya GV. 2022.
Evaluating phosphorus use efficiency across
different lettuce (Lactuca sativa L.) acces-
sions. Euphytica. 218:28. https://doi.org/10.1007/
s10681-022-02973-6.
Lan P, Li W, Schmidt W. 2015. ‘Omics’ap-
proaches towards understanding plant phospho-
rus acquisition and use, p 65–98. In: Plaxton
WC, Lambers H (eds.). Annual plant reviews
volume 48: Phosphorus metabolism in plants.
https://doi.org/10.1002/9781118958841.
Lee W, Zotarelli L, Rowland DL, Liu G. 2021.
Evaluationofpotatovarietiesgrowninhy-
droponics for phosphorus use efficiency.
Agriculture. 11:668. https://doi.org/10.3390/
agriculture11070668.
Liu G, Dunlop J, Phung T. 2006. Induction of root
hair growth in a phosphorus-buffered culture
solution. Agric Sci China. 5(5):370–376.
Liu G, Dunlop J, Phung T, Li Y. 2007. Physiologi-
cal responses of wheat phosphorus-efficient
and -inefficient genotypes in field and effects
of mixing other nutrients on mobilization of in-
soluble phosphates in hydroponics. Commun
Soil Sci Plant Anal. 38:2239–2256. https://doi.
org/10.1080/00103620701549249.
Neocleous D, Savvas D. 2019. The effects of
phosphorus supply limitation on photosynthe-
sis, biomass production, nutritional quality, and
mineral nutrition in lettuce grown in a recircu-
lating nutrient solution. Scientia Hortic.
252:379–387. https://doi.org/10.1016/j.scienta.
2019.04.007.
Neto AP, Favarin JL, Hammond JP, Tezotto T,
Couto HTZ. 2016. Analysis of phosphorus use
efficiency traits in Coffea genotypes reveals
Coffea arabica and Coffea canephora have
contrasting phosphorus uptake and utilization
efficiencies. Front Plant Sci. 7:408. https://doi.
org/10.3389/fpls.2016.00408.
Nirubana V, Vanniarajan C, Aananthi N, Rama-
lingam J. 2020. Screening tolerance to phos-
phorus starvation and haplotype analysis
using phosphorus uptake (Pup1) QTL linked
markers in rice genotypes. Physiol Mol Biol
Plants. 26:2355–2369. https://doi.org/10.1007/
s12298-020-00903-1.
Ochigbo AE, Bello LL. 2014. Screening of soybean
varieties for phosphorus use efficiency in nutrient
solution. Agric Biol J N Am. 5(2):68–77.
Parentoni SN, Mendes FF, Guimar~
aes LJM.
2012. Breeding for phosphorus use efficiency,
p 67-85. In: Fritsche-Neto R, Bor
em A (eds.).
Plant breeding for abiotic stress tolerance.
Berlin, Heidelberg: Springer. https://doi.org/
10.1007/978-3-642-30553-5_5.
Raghothama KG. 1999. Phosphate acquisition. Annu
Rev Plant Physiol Plant Mol Biol. 50:665–693.
Resh HM. 2022. Hydroponic food production: A
definite guidebook for the advanced home gar-
dener and the commercial hydroponic grower.
Boca Raton, FL: CRC Press.
Rose TJ, Wissuwa M. 2012. Rethinking internal
phosphorus utilization efficiency: A new ap-
proach is needed to improve PUE in grain
crops, p 185–217. In: Sparks DL (ed.). Advan-
ces in Agronomy. Academic Press. https://doi.
org/10.1016/B978-0-12-394277-7.00005-1.
Sanda~
na P. 2016. Phosphorus uptake and utilization ef-
ficiency in response to potato genotype and phos-
phorus availability. Eur J Agron. 76:95–106.
https://doi.org/10.1016/j.eja.2016.02.003.
Sandoya GV. 2019. Advances in lettuce breeding
and genetics, p 459–478. In: Hochmuth G
(ed.). Burleigh Dodds series in agricultural sci-
ence. Burleigh Dodds Science Publishing.
Sandoya GV, Bosques J, Rivera F, Campoverde
EV. 2021. Growing lettuce in small hydroponic
HORTSCIENCE VOL. 58(4) APRIL 2023 473
systems: HS 1422. EDIS. 2021(5). https://doi.
org/10.32473/edis-hs1422-2021.
Sapkota S, Sapkota S, Liu Z. 2019. Effects of nutrient
composition and lettuce cultivar on crop production
in hydroponic culture. Horticulturae. 5:72. https://
doi.org/10.3390/horticulturae5040072.
Sarvajayakesavalu S, Lu Y, Withers PJA, Pavinato
PS, Pan G, Chareonsudjai P. 2018. Phosphorus
recovery: A need for an integrated approach.
Ecosyst Health Sustain. 4(2):48–57. https://doi.
org/10.1080/20964129.2018.1460122.
Sharma N, Acharya S, Kumar K, Singh N, Chaura-
sia OP. 2018. Hydroponics as an advanced
technique for vegetable production: An over-
view. J Soil Water Conserv. 17(4):364–371.
https://doi.org/10.5958/2455-7145.2018.
00056.5.
van de Wiel CCM, Linden CG, Scholten OE.
2016. Improving phosphorus use efficiency in
agriculture: Opportunities for breeding. Eu-
phytica. 207:1–22. https://doi.org/10.1007/
s10681-015-1572-3.
WenZ,LiH,ShenQ,TangX,XiongC,LiH,
Pang J, Ryan MH, Lambers H, Shen J. 2019.
Tradeoffs among root morphology, exuda-
tion and mycorrhizal symbioses for phospho-
rus-acquisition strategies of 16 crop species.
New Phytol. 223(2):882–895. https://doi.
org/10.1111/nph.15833.
Wissuwa M, Yano M, Ae N. 1998. Mapping of
QTLs for phosphorus-deficiency tolerance in rice
(Oryza sativa L.). Theor Appl Genet. 97:777–789.
https://doi.org/10.1007/s001220050955.
474 HORTSCIENCE VOL. 58(4) APRIL 2023
0
5
10
15
20
25
Tissue-P (g kg-1)
Root - Low P Root - High P Shoot - Low P Shoot - High P
Supplemental Fig. 1. Least square means of shoot and root TP (g·kg
1
) with 95% confidence intervals of 12 lettuce accessions grown under low and high P
in E1 and E2.
0
25
50
75
100
125
150
E3 E4 E3 E4
tooRtoohS
Fresh weight (g plant-1)
TCP MSP
Supplemental Fig. 2. Average least square means of shoot and root weight (g·plant
1
) with 95% confidence intervals of 12 lettuce accessions grown under
tricalcium phosphate (TCP) and monosodium phosphate (MSP) treatments in E3 and E4.
0
5
10
15
20
25
Tissue-P (g kg-1)
Root - TCP Root - MSP Shoot - TCP Shoot - MSP
Supplemental Fig. 3. Least square means of shoot and root TP (g·kg
–1
) with 95% confidence intervals of 12 lettuce accessions grown under tri-calcium phos-
phate (TCP) and monosodium phosphate (MSP) treatments in E3.
HORTSCIENCE VOL. 58(4) APRIL 2023 1
0
5
10
15
20
25
Tissue-P (g kg-1)
Root - TCP Root - MSP Shoot - TCP Shoot - MSP
Supplemental Fig. 4. Least square means of shoot and root total-P (g·kg
1
) with 95% confidence intervals of 12 lettuce accessions grown under tricalcium
phosphate (TCP) and monosodium phosphate (MSP) treatments in E4.
Supplemental Table 1. Analysis of variance of shoot and root weight and shoot and root tissue total-P (TP), for the 12 lettuce accessions in E1 and E2.
Shoot wt Root wt Shoot TP
i
Root TP
i
Source of variation Num DF
ii
Pvalue
Accession (G) 11 <0.0001 <0.0001 <0.0001 0.3768
Treatment (T) 1 <0.0001 <0.0001 <0.0001 <0.0001
G×T 11 0.0434 0.3033 0.1335 0.8704
Experiment (E) 1 <0.0001 0.0008 <0.0001 0.0179
G×E 11 0.0058 0.0015 0.0002 0.1938
T×E 1 0.0001 0.0005 <0.0001 <0.0001
G×T×E 11 0.2333 0.1770 <0.0001 0.7268
i
Tissue total-P estimated on the basis of dry weight.
ii
Numerator degrees of freedom.
Supplemental Table 2. Analysis of variance of shoot and root fresh weight, and shoot and root tissue total-P (TP) for the 12 lettuce accessions grown un-
der tricalcium phosphate (TCP) and monosodium phosphate (MSP) treatments in E3 and E4.
Source of variation Num DF
ii
Shoot wt Root wt Shoot TP
i
Root TP
i
E3 E4 E3 E4 E3 E4 E3 E4
Pvalue
Accession (G) 11 <0.0001 <0.0001 <0.0001 <0.0001 0.2804 0.0081 0.3242 0.5283
Treatment (T) 1 0.5098 <0.0001 0.1640 <0.0001 0.1165 <0.0001 0.1057 0.0001
G×T 11 0.9112 0.0002 0.0518 0.3211 0.1481 0.0661 0.7309 0.2456
i
Tissue TP estimated on the basis of dry weight.
ii
Numerator degrees of freedom.
2HORTSCIENCE VOL. 58(4) APRIL 2023
