Received: 8 November 2017
Revised: 23 December 2017
Accepted: 26 December 2017
DOI: 10.1111/pce.13139
SPECIAL ISSUE
Leaf transpiration plays a role in phosphorus acquisition among a large set of chickpea genotypes Jiayin Pang1,2 Megan H.
|
Hongxia Zhao3
Ryan1,2 |
|
Ruchi Bansal4
Kadambot H.M. Siddique
|
Emilien Bohuon5,6
|
Hans Lambers1,6
|
1,2
1
The UWA Institute of Agriculture, The University of Western Australia, Perth, WA 6001, Australia
2
School of Agriculture and Environment, The University of Western Australia, Perth, WA 6001, Australia
3
Institute of Mountain Hazards and Environment, Chinese Academy of Sciences, Chengdu 610041, China
4 Division of Germplasm Evaluation, ICAR‐ National Bureau of Plant Genetic Resources, New Delhi 110012, India 5
Institut Polytechnique UniLaSalle, Beauvais Cedex 60000, France
6
School of Biological Sciences, The University of Western Australia, Perth, WA 6001, Australia Correspondence J. Pang, School of Agriculture and Environment, The University of Western Australia, Perth, WA 6001 Australia. Email:
[email protected] Funding information ARC Future Fellowship, Grant/Award Number: FT140100103
1
|
Abstract Low availability of inorganic phosphorus (P) is considered a major constraint for crop productivity worldwide. A unique set of 266 chickpea (Cicer arietinum L.) genotypes, originating from 29 countries and with diverse genetic background, were used to study P‐use efficiency. Plants were grown in pots containing sterilized river sand supplied with P at a rate of 10 μg P g−1 soil as FePO4, a poorly soluble form of P. The results showed large genotypic variation in plant growth, shoot P content, physiological P‐use efficiency, and P‐utilization efficiency in response to low P supply. Further investigation of a subset of 100 chickpea genotypes with contrasting growth performance showed significant differences in photosynthetic rate and photosynthetic P‐use efficiency. A positive correlation was found between leaf P concentration and transpiration rate of the young fully expanded leaves. For the first time, our study has suggested a role of leaf transpiration in P acquisition, consistent with transpiration‐driven mass flow in chickpea grown in low‐P sandy soils. The identification of 6 genotypes with high plant growth, P‐acquisition, and P‐utilization efficiency suggests that the chickpea reference set can be used in breeding programmes to improve both P‐acquisition and P‐utilization efficiency under low‐P conditions. KEY W ORDS
Cicer arietinum, leaf transpiration, phosphorus, phosphorus‐acquisition efficiency, phosphorus‐use efficiency, phosphorus‐utilization efficiency, photosynthetic phosphorus‐use efficiency, water‐use efficiency
I N T RO D U CT I O N
is expected to be depleted in the not too distant future, with prices rising as the resource becomes more scarce (Fixen & Johnston, 2012).
Low availability of phosphorus (P) is considered a major constraint to
Therefore, breeding crops for improved P‐use efficiency (PUE), that
crop production worldwide (Raghothama, 1999). It has been estimated
is, an improved production of yield per unit of added P fertilizer, is
that 5.7 billion hectares of land worldwide are deficient in P (Batjes,
arguably the best long‐term environmentally sustainable strategy
1997). The inorganic P (Pi) concentration in the soil solution of arable
(Rose, Liu, & Wissuwa, 2013).
land is generally 0.5 million ha (FAOSTAT, 2014). It is a primary crop in developing
sterilized washed coarse river sand. The bottom of each pot had four
countries, contributing to a significant part of human food and animal
10‐mm‐diameter holes to provide free drainage, which were covered
feed (Foyer et al., 2016; Pang, Turner, Du, Colmer, & Siddique, 2017).
by a nylon mesh to prevent any leakage of the sand. In this study, a
In common with many other widely grown crops, chickpea has a nar-
sandy‐based substrate was used as the growth medium as deep sands
row genetic base due to domestication (Varshney et al., 2013). In
and duplex soils with sandy topsoil cover more than two thirds of the
recent years, a unique chickpea reference set consisting of 300 geno-
cropping area in Western Australia (Pang et al., 2011). This growth sub-
types from 29 countries including 267 landraces, 13 advanced breed-
strate had a P‐retention index (a measure of a soil's capacity to sorb P)
ing lines and cultivars, 7 wild relatives, and 13 accessions whose
of 1.5 units, measured as described by Allen and Jeffery (1990). A soil
classification is unknown was developed by the International Crops
with a P‐retention index of .05) on any of the measured parame-
Just prior to the final harvest, 100 genotypes (79 desi, 15 kabuli, and 6 pea‐shaped types) showing visual differences in plant size or leaf symptoms of P deficiency were selected for the measurements of photosynthetic characteristics. Gas exchange was measured on young fully expanded leaves on primary branches between 10:30 h and 12:00 h WST using a LICOR‐6400 with a red/blue LED light source (LI‐COR, Lincoln, NE, USA). Photosynthetic photon flux density at the leaf surface was set at 1,500 μmol m−2 s−1, block temperature −1
at 25 °C, flow rate at 500 μmol s , and ambient CO2 concentration −1
of the incoming gas stream at 380 μmol mol
as in Pang, Turner,
Khan, et al. (2017). Leaves used for photosynthesis measurement were sampled, and projected leaf area measured using an Epson
ters in this study, suggesting that seed P content did not have a significant role in plant growth or P acquisition 7 weeks after sowing. When genotypic effects were significant, the least significant difference at P = .05 is presented in figures. Linear regression analysis was conducted in Sigmaplot version 12.5 (Systat Software, Inc., 2011). Principal components analysis (PCA) based on the correlation matrix was undertaken using Genstat to examine the relationships among plant traits for 266 genotypes or the subset of 100 genotypes. All seven growth traits measured for 266 genotypes and 14 growth traits for the selected 100 genotypes were included for the PCA analysis. PCA biplots are presented with plants traits (PC factor loadings) and genotypes (PC scores).
1680 scanner at 200 dpi and analysed using the routine procedure in the WinRhizo version 4.1c scanner program, followed by drying
4
RESULTS
|
at 70 °C for 48 hr for leaf DW. Leaf mass per area (LMA) of leaves for photosynthesis measurement was calculated as the ratio of leaf DW to leaf area. Mass‐based photosynthetic rate (Amass) was calcu-
4.1
|
Observations
lated from area‐based photosynthetic rate (Aarea) divided by LMA.
Plants of some genotypes showed visual symptoms of P deficiency
Instantaneous intrinsic water‐use efficiency (WUE) was calculated as
starting from 3 weeks after sowing, with leaflet margins of the bottom
the ratio of photosynthetic rate to stomatal conductance (Osmond,
leaves on main stems showing necrosis (indicated by the arrow in
Björkman, & Anderson, 1980).
Figure 1a) or a purple discolouration (Figure 1b). As leaf necrosis
Plants were harvested 7 weeks after sowing. Roots were washed
became more severe and developed upwards over time, some leaflets
carefully, and plants were separated into shoot and roots. Shoots and
from the bottom of stem turned yellow and were shed (Figure 1c). In
roots were dried at 70 °C for 72 hr, and DW was measured. Root mass
some genotypes, no visual symptoms of P deficiency were observed
ratio was calculated as the ratio of root DW to the total plant DW.
in shoots during the entire experiment, for example, in eight desi
4
PANG
ET AL.
FIGURE 1
Some chickpea genotypes showing necrosis (a) or purple colouration (b) of leaflets 3 weeks after sowing, and leaflets were shed starting from the old leaves on the main stems when symptoms developed further in some genotypes (c); whereas no visual phosphorus (P) deficiency symptoms were observed in some other genotypes (d). Some genotypes formed nodules under low‐P conditions (e) [Colour figure can be viewed at wileyonlinelibrary.com]
FIGURE 2
Box plots showing (a) shoot dry weight (DW), (b) root DW, (c) root mass ratio, (d) shoot phosphorus (P) concentration, (e) shoot P content, (f) physiological P‐use efficiency, and (g) P‐utilization efficiency of 266 chickpea genotypes including 203 desi type, 53 kabuli type, and 10 pea‐shaped types grown for 7 weeks in washed river sand supplied with 10 μg P g−1 dry soil as FePO4. The central vertical bar in each box shows the median, the box represents the interquartile range, the whiskers show the location of the most extreme data points that are still within a factor of 1.5 of the upper or lower quartiles, and the black points are values that fall outside the whiskers
PANG
5
ET AL.
genotypes (CICA0912, ICC 1194, ICC 1230, ICC 12928, ICC 16524,
Root mass ratio varied three‐fold among the 266 genotypes
ICC 1882, ICC 4463, and ICC 6279) and six kabuli genotypes (ICC
(P < .001), ranging from 0.21 to 0.61 with a mean value of 0.49 (Fig-
13187, ICC 13523, ICC 12328, ICC 13816, ICC 7668, and ICC
ures 2c and S3).
15406), although these showed contrasting shoot size (Figure 1d). Nodules were observed in 19 of the genotypes (as shown in Figure 1e): ICC 1194, ICC 4418, ICC 16796, ICC 15618, ICC 4814, ICC 7323, ICC 12037, ICC 8350, ICC 15996, ICC 9402, ICC 7305,
4.3
|
Shoot P concentration and shoot P content
ICC 10945, ICC 13524, ICC 3776, ICC 7441, ICC 9636, ICC 1083,
Both shoot P concentration and shoot P content varied greatly
ICC 16374A, and ICC 9862. Almost all of those genotypes only had a
among the 266 genotypes (P < .001 for both; Figures 2d, 2e, S4,
few nodules per plant, except ICC 1194, which had about 20 nodules
and S5). Shoot P concentration varied almost six‐fold, ranging from
per plant.
0.82 mg g−1 DW in ICC 5135 to 4.67 mg g−1 DW in ICC 13523 with a mean value of 1.37 mg g−1 DW. Although shoot P concentration of almost all genotypes ranged from 1 to 2 mg g−1 DW, five
4.2
|
kabuli‐type genotypes (ICC 9137, ICC 8151, ICC 5337, ICC 15802,
Plant growth
and ICC 13523) had a significantly higher shoot P concentration
An eight‐fold difference in shoot DW was found among the 266 geno−1
types (P < .001), ranging from 0.07 g plant −1
0.58 g plant −1
0.32 g plant
in ICC 13523 to
in Genesis Kalkee with a mean shoot DW of
(Figures 2a and S1). Similar to shoot DW, genotypes also
ranging from 2.51 to 4.67 mg g−1 DW (Figure S4). A 7.6‐fold difference in shoot P content was found among the 266 genotypes, ranging from 0.13 to 0.95 mg g
̶ 1
with a mean value of 0.44 mg g
̶
1
(Figures 2e and S5). Four kabuli genotypes (ICC 10885, ICC
varied greatly in root DW (P < .001), ranging from 0.02 g plant−1 in
16796, ICC 15435, and Genesis Kalkee) and one desi genotype
ICC13523 to 0.61 g plant−1 in ICC 15610 with a mean value of
(ICC 7554) had the highest shoot P content among all genotypes
−1
0.32 g plant
(Figure S5).
(Figures 2b and S2).
TABLE 1
Chickpea genotypes belonging to the top 10% in shoot dry weight (DW) among 266 genotypes, with some genotypes also being in the top 10% for root DW, shoot phosphorus (P) content, P‐utilization efficiency, and physiological P‐use efficiency as shown by black circles Genotype
Market type
Shoot DW
ICC 7255
Kabuli
●
Root DW
Shoot P content
P‐utilization efficiency
Physiological P‐use efficiency
●
●
●
●
●
●
●
ICC 8350
Pea‐shaped
●
●
ICC 13357
Kabuli
●
●
ICC 15612
Desi
●
ICC 3391
Desi
●
ICC 4918
Desi
●
ICC 1915
Desi
●
ICC 14669
Desi
●
ICC 9848
Pea‐shaped
●
ICC 8200
Desi
●
ICC 11903
Desi
●
Almaz
Kabuli
ICC 2277
Kabuli
ICC 12379
Desi
●
ICC 7315
Kabuli
ICC 16796
Kabuli
ICC 10885
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Kabuli
●
●
●
●
ICC 3239
Desi
●
ICC 6294
Desi
●
ICC 8318
Desi
●
ICC 15435
Kabuli
●
●
●
●
Genesis090
Kabuli
●
●
●
●
●
ICC 8261
Kabuli
●
●
●
●
●
ICC 13124
Desi
●
ICC 7554
Desi
●
ICC 6306
Desi
●
Genesis Kalkee
Kabuli
●
●
●
●
●
6
4.4
PANG
|
Physiological PUE and P‐utilization efficiency
A 29‐fold difference in physiological PUE was found among the 266 genotypes (P < .001), ranging from 15 g2 DW g−1 P in ICC 13523 to 439 g2 DW g−1 P in ICC 6306 with a mean value of 244 g2 DW g−1 P (Figures 2f and S6). Phosphorus‐utilization efficiency varied five‐fold among the 266 genotypes (P < .001), which ranged from 217 g2 DW g−1 P in ICC 13,523 to 1,100 g2 DW g−1 P in ICC 5135 and had a mean value of 758 g2 DW g−1 P (Figures 2g and S7).
4.5
|
Genotypes with high P efficiency
4.6
|
ET AL.
Leaf photosynthesis, transpiration, and WUE
Area‐based leaf photosynthetic rate (Aarea) showed a two‐fold variation among the selected set of 100 genotypes with contrasting growth performance (P < .001, Figures 3a and S8a). Aarea ranged from 11.7 μmol m−2 s−1 to 27.7 μmol m−2 s−1, with a mean value of 19.0 μmol m−2 s−1. Although there was large variation in LMA among genotypes (range 32.8–58.2 g m−2, data not shown), a strong positive correlation between mass‐based leaf photosynthetic rate (Amass) and Aarea was found (r = .87, P < .001). As a result, Amass showed a similar trend to Aarea among the genotypes (Figures 3b and S8b). Amass varied from
Of the 27 genotypes with the top 10% of shoot DW, six genotypes
0.23 μmol CO2 g−1 s−1 to 0.66 μmol CO2 g−1 s−1 with a mean value
(ICC 8350, ICC 9848, ICC 2277, ICC 7315, Genesis090, and ICC
of 0.43 μmol CO2 g−1 s−1.
8261) also had the top 10% of root DW, shoot P content, P‐utilization
Leaf transpiration rate varied ~three‐fold among the selected 100
efficiency, and physiological PUE (Table 1). Five genotypes (Almaz, ICC
genotypes (P < .001), ranging from 3.4 mmol m−2 s−1 in ICC 9942 to
16796, ICC 10885, ICC 15435, and Genesis Kalkee) belonged in the top 10% of all the above parameters, except P‐utilization efficiency,
9.8 mmol m−2 s−1 in ICC 12037 with a mean value of 6.1 mmol m−2 s −1
(Figures 3c and S8c). Instantaneous intrinsic WUE varied two‐fold
whereas ICC 13357 belonged in the top 10% of all the above parame-
among seed types and the 100 genotypes (both P < .001), which
ters, except for shoot P content. ICC 7255 belonged in the top 10% for
ranged from 43 μmol mol−1 to 91 μmol mol−1 with a mean value of
shoot DW and shoot P content (Table 1).
62 μmol mol−1 (Figures 3d and S8d).
FIGURE 3 Box plots showing (a) area‐ and (b) mass‐based leaf photosynthesis rate, (c) transpiration rate, (d) water‐use efficiency, and (e) photosynthetic P‐use efficiency of 100 chickpea genotypes including 79 desi type, 15 kabuli type, and 6 pea‐shaped types grown for 7 weeks in washed river sand supplied with 10 μg P g−1 dry soil as FePO4. The central vertical bar in each box shows the median, the box represents the interquartile range, the whiskers show the location of the most extreme data points that are still within a factor of 1.5 of the upper or lower quartiles, and the black points are values that fall outside the whiskers
PANG
ET AL.
4.7
|
Photosynthetic PUE
Photosynthetic PUE varied greatly among the selected 100 genotypes (P < .001), ranging from 170 μmol CO2 g−1 P s−1 in ICC 14199 to 437 μmol CO2 g−1 P s−1 in ICC 12037 with a mean value of 295 μmol CO2 g−1 P s−1 (Figures 3e and S9).
7
circle in Figure 5a, had the highest values of shoot and leaf P concentration, shoot P content, Aarea, Amass and leaf transpiration rate, but the lowest P‐utilization efficiency. The biplots in Figure 5 and Pearson correlation analysis in Table S3 show some strong correlations among all 14 plant traits of the selected 100 genotypes. Among 100 genotypes, P concentration in the young fully expanded leaves was positively correlated with leaf transpiration rate (r = .57, P < .001), but negatively
4.8
|
Correlations between traits
correlated with intrinsic WUE at the leaf level (r = −.44, P < .001), whereas there was no correlation with root DW (r = .10, P > .05; Fig-
PCA based on seven plant traits of 266 genotypes explained 83% of
ures 5a, 6a, 6b, Table S3). Shoot P content had a correlation with leaf
variance in the first two components (Figure 4, Table S1). The first
P concentration (r = .66, P < .001) and a weak correlation with transpi-
component (PC1) represented 47% of the variability and accounted
ration rate (r = .23, P < .05; Figure 5a, Table S3). Leaf transpiration rate
primarily for shoot and root DW, and physiological PUE. The second
was not correlated with shoot DW (r = .00, P > .05) or with root DW
component (PC2) represented 35% of the variability primarily compris-
(r = −.05, P > .05; Table S3). WUE had a linear negative correlation with
ing shoot P concentration, shoot P content, and P‐utilization efficiency
leaf stomatal conductance (r = .73, P < .001, data not shown). There
(Figure 4, Table S1). Among 266 genotypes, shoot P content was pos-
was a weak correlation between photosynthetic PUE at the leaf level
itively correlated with both shoot DW (r = .86, P < .001) and root DW
and P‐utilization efficiency at the whole plant level (r = .35, P < .001),
(r = .57, P < .001; Figure 4). Physiological PUE was also positively cor-
whereas no correlation between photosynthetic PUE at the leaf level
related with shoot DW (r = .86, P < .001) and root DW (r = .69,
and physiological PUE at the whole plant level was found (r = .07,
P < .001; Figure 4).
P > .05; Figure 5, Table S3).
For the PCA analysis of 100 genotypes with 14 plant traits included, three principal components were identified explaining 75% of variation (Figure 5, Table S2). PC1 represented 38% of the variation
5
|
DISCUSSION
and was primarily comprised of shoot P concentration, shoot P content, P‐utilization efficiency, Aarea, Amass, leaf P concentration, and leaf
Our first hypothesis that there is a large variation in growth and PUE
transpiration rate, whereas PC2 (shoot and root DW, physiological
among genotypes in the chickpea reference set with diverse genetic
PUE) represented 22% of the variation and PC3 (Amass and photosyn-
background and some genotypes would be superior to the commercial
thetic PUE) 15% of the variation (Table S2). In the biplots of PC1 ver-
cultivars in growth and P‐acquisition and PUE was fully supported.
sus PC2, four kabuli genotypes (ICC 11879, ICC 14199, ICC 15294,
Chickpea genotypes grown under low‐P conditions in this study
and ICC 16796) and one desi genotype (ICC 6877), as shown in the
showed very large variation in all measured plant traits. Five genotypes (ICC 8350, ICC 9848, ICC 2277, ICC 7315, and ICC 8261) and one Australian commercial cultivar (Genesis090) belonged to the top 10% of genotypes for shoot DW, root DW, shoot P content, P‐utilization efficiency, and physiological PUE. The present findings are consistent with results from other studies. For example, Srinivasarao et al. (2006) found a significant difference in growth, P acquisition, and P‐ utilization efficiency among 20 chickpea genotypes cultivated in different regions of India when grown in pots under both P stress and at optimum levels of P supplied as diammonium phosphate. Recently, Keneni et al. (2015) evaluated 155 chickpea genotypes with and without P fertilizer in Ethiopia and found significant genotypic differences in P acquisition and PUE. Those authors identified several landraces that outperformed a commercial cultivar (Natoli) for P acquisition, PUE, and agronomic characters including grain yield; this indicates the possibilities for developing better varieties for PUE using chickpea landraces collected from Ethiopia. Internal P‐utilization efficiency is often lower in plants with high P‐acquisition efficiency as a result of higher tissue P concentrations (Rose, Rose, Pariasca‐Tanaka, Heuer, & Wissuwa, 2011). However, the identification of genotypes with high shoot DW, root DW, shoot P content, P‐utilization efficiency, and
FIGURE 4
Principal component analysis of seven plant traits for 266 chickpea genotypes including 203 desi type, 53 kabuli type, and 10 pea‐shaped type grown for 7 weeks in washed river sand supplied with 10 μg P g−1 dry soil as FePO4. Biplot vectors are trait factor loadings, whereas the position of each genotype of desi (circles), kabuli (triangles), and pea‐shaped (diamonds) types is shown
physiological PUE from the present study indicates the possibility of using the chickpea reference set in breeding programmes to improve both P‐acquisition and P‐utilization efficiency in low‐P environments. In previous studies on P‐acquisition and PUE in chickpea, P fertilizer was always supplied in soluble forms, for example, diammonium
8
PANG
ET AL.
FIGURE 5
Principal component analysis of 14 plant traits for 100 chickpea genotypes including 79 desi type (circles), 15 kabuli type (triangles), and 6 pea‐shaped type (diamonds) grown for 7 weeks in washed river sand supplied with 10 μg P g−1 dry soil as FePO4. Biplot vectors are trait factor loadings, and the position of each genotype is shown for principal component (a) PC1 versus PC2 representing 60% of the variability, (b) PC1 versus PC3 representing 51% of the variability, and (c) PC2 versus PC3 representing 37% of the variability [Colour figure can be viewed at wileyonlinelibrary.com]
phosphate in Srinivasarao et al. (2006) and triple superphosphate in
to either total shoot or root DW (Figure 5, Table S3). Interestingly,
Keneni et al. (2015). Thus, our study extends the knowledge of geno-
we also found a negative correlation between leaf P concentration
typic variation of chickpea in P‐acquisition and P‐utilization efficiency
and WUE at the leaf level, which showed a significant negative corre-
when P is supplied in a sparingly soluble form. Most of the fertilizer
lation with leaf stomatal conductance but not with photosynthetic
applied to crops is rapidly transformed into forms with limited availabil-
rate; this further indicates the involvement of stomatal conductance
ity to plants, such as complexes with oxides and hydroxides of Fe and
and the possible contribution of transpiration‐driven mass flow to P
Al in acid soils (Osborne & Rengel, 2002). Therefore, the development
acquisition. Masle, Farquhar, and Wong (1992) found a significant cor-
of chickpea genotypes that use poorly available P forms would result in
relation between transpiration ratio (ratio of water transpired to car-
lower input, more environmentally friendly production systems of agri-
bon fixed, the inverse of water use efficiency) and leaf P
culture. The very large variation in plant growth, and P‐acquisition and
concentration in sunflower (Helianthus annuus), wheat (Triticum
P‐utilization efficiency under low Fe‐P in this study indicates the utility
aestivum), and tobacco (Nicotiana tabacum) when ample P was supplied
of the chickpea reference set to identify chickpea genotypes superior
for plant growth (Masle et al., 1992). However, such a relationship was
to the varieties released so far. Our findings also justify the need to ini-
much looser when plants were fed with low concentration nutrient
tiate a breeding programme to improve PUE in chickpea and reduce
solutions (Masle et al., 1992). It has been suggested that significant
dependency on commercial fertilizers.
correlations between P acquisition and transpiration rate are mostly
Our second hypothesis that leaf transpiration rate plays an impor-
found when the concentration of nutrients in the root medium is high
tant role in P acquisition of chickpea grown under low‐P conditions in
(Garrish, Cernusak, Winter, & Turner, 2010; Pitman, 1988); however,
sandy soil was fully supported. There was a positive correlation
[Pi] in soil solutions is generally