Interactions among cluster-root investment, leaf phosphorus concentration, and relative growth rate in two Lupinus species1

Transcription

1 RESEARCH ARTICLE AMERICAN JOURNAL OF BOTANY Interactions among cluster-root investment, leaf phosphorus concentration, and relative growth rate in two Lupinus species1 Xing Wang2, Erik J. Veneklaas, Stuart J. Pearse, and Hans Lambers PREMISE OF THE STUDY: Cluster-root (CR) formation is a desirable trait to improve phosphorus (P) acquisition as global P resources are dwindling. CRs in some lupine species are suppressed at higher P status. Whether increased growth rate enhances CR formation due to a dilution of leaf P concentration is unknown. We investigated interactive effects of leaf P status and relative growth rate (RGR) on CR formation in two Lupinus species, which differ in their CR biomass investment. Variation in RGR was imposed by varying day length. METHODS: Lupinus albus and L. pilosus were grown hydroponically with KH 2 at a day length of 6, 10, or 14 h. We used a slightly higher P supply at longer day lengths to avoid a decline in leaf P concentration, which would induce CRs. Cluster-root percentage, leaf P concentrations, and RGR were determined at 22, 38, and 52 d after sowing. KEY RESULTS: Lupinus species grown at similar root P availability, but with a faster growth rate, as dependent on day length, showed a greater CR percentage. Because our aim to achieve exactly the same leaf P concentrations at different day lengths was only partially achieved, we carried out a multiple regression analysis. This analysis showed the CR percentage was strongly and negatively correlated with plant P status and only marginally and positively correlated with RGR. CONCLUSIONS: The two Lupinus species invariably formed fewer cluster roots at higher leaf P status, irrespective of RGR. Differences in RGR or leaf P concentration cannot explain the species-specific variation in cluster-root investment. KEY WORDS biomass allocation; cluster roots; day length; Fabaceae; Lupinus ; multiple regression analysis Phosphorus (P) is an essential macronutrient for plant growth; however, commercially extractable P reserves are declining ( Cordell et al., 2009 ; Gilbert, 2009 ), although probably not as rapidly as previously forecasted ( Fixen and Johnston, 2012 ; Scholz and Wellmer, 2013 ). Phosphorus is already a major limitation for crop production in many parts of the world, in particular South America, Africa, and southern Asia ( Khasawneh et al., 1980 ; Lynch, 2007 ; Raghothama, 1999 ; Runge-Metzger, 1995 ; Stutter et al., 2012 ). Many soils contain substantial levels of P, but in forms not readily available for most crop plants (e.g., wheat and canola [ Holford, 1997 ; Simpson et al., 2011 ]). For sustaining high crop yields, it is critical to improve the efficiency of P acquisition and P use in cropping systems to satisfy 1 Manuscript received 6 June 2015; revision accepted 20 July School of Plant Biology and The Institute of Agriculture, The University of Western Australia, 35 Stirling Highway, Crawley (Perth), WA 6009, Australia 2 Author for correspondence ( xing.wang@graduate.uwa.edu.au) doi: /ajb world demand for food, feed, and fiber, as P resources dwindle and become increasingly expensive. Some lupine species, e.g., Lupinus albus L. (Fabaceae), can acquire P under low-p conditions because they form cluster roots, which exude carboxylates in an exudative burst ( Watt and Evans, 1999 ). Cluster roots are not only present in L. albus and certain other Lupinus species ( Lambers et al., 2013 ), but also in most Proteaceae and several other taxa ( Dinkelaker et al., 1995 ; Lambers et al., 2006 ). In field studies, cluster-root-forming species show little response to P fertilizer addition because the carboxylate release yields a very high acquisition efficiency for P ( Bolland et al., 1999 ). Therefore, carboxylate-releasing cluster roots are considered an adaptive strategy to access poorly available P from soil sources ( Lambers et al., 2008 ). Cluster-root formation and carboxylate exudation are plastic traits such that plants in high-p soil, with a favorable plant P status, minimize investment in these carbon-costly P-acquisition mechanisms ( Lambers et al., 2013 ; Richardson et al., 2011 ). AMERICAN JOURNAL OF BOTANY 102 (9 ): , 2015 ; Botanical Society of America 1529

2 1530 AMERICAN JOURNAL OF BOTANY The regulatory mechanism controlling the distinct pattern of cluster-root formation is still largely unknown. Some researchers reported that many genes and gene expression patterns related to a P-starvation response are affected by 100 mm sucrose ( Karthikeyan et al., 2007 ; Müller et al., 2007 ; Nacry et al., 2005 ). In addition, splitroot experiments and foliar-feeding experiments ( Li et al., 2008 ; Shane et al., 2003 ) have shown that systemic signals associated with P status in the shoot, rather than P availability in the rhizosphere, control cluster-root formation in L. albus. Some P-starvation responses, including inhibition of primary root growth, initiation of lateral root formation, and the formation of root hairs are partly a result of local low-p signaling and partly affected by exogenous auxins ( Gilbert et al., 2000 ; Jungk, 2001 ; Péret et al., 2011 ). As summarized by Pearse et al. (2006), cluster-root formation is regulated by a negative feedback loop, in which cluster roots have a positive effect on P uptake, which enhances P status and in turn suppresses cluster-root formation and exudation in Lupinus species. However, an increased P-uptake rate may not lead to a more favorable plant P status if growth rate is stimulated to the same extent. We presented a revised model in our recent study ( Wang et al., 2013 ) based on the model proposed by Pearse et al. (2006), that cluster-root formation of responsive Lupinus plants increases P-uptake rate, which successively stimulates plant growth and enhances plant P status; however, we found that variation in cluster-root formation among three Lupinus species cannot be explained for species-specific variation in relative growth rate (RGR) or leaf P concentration. To date, all experiments on the effect of P status on cluster-root formation have achieved a different P status by varying P availability, causing a parallel effect on growth rate. It is therefore important to investigate whether plants with high P status invariably downregulate cluster-root formation, irrespective of growth rate. We hypothesized that Lupinus species grown with a similar P availability, but with a higher growth rate, as dependent on light availability, invest more in cluster roots. We also hypothesized that the differences in cluster-root formation between two Lupinus species are associated with differences in either relative growth rate or internal P concentration. Two cluster-root-forming Lupinus species, L. albus and L. pilosus L., were grown at a similar P availability in the root environment, but at different day lengths, to impose different growth rates. Our objective in this study was to investigate whether the negative correlation between plant P status and investment in cluster roots is maintained if differences in growth rate are imposed by a treatment other than P supply. To test our hypotheses, we determined phenotypic and species-specific relationships among clusterroot formation, leaf P status, and plant relative growth rate. MATERIALS AND METHODS Plant growth In November 2012, seeds of Lupinus albus and L. pilosus were germinated in pots filled with washed and sterilized river sand in a glasshouse. At 7 d after sowing, uniformly sized seedlings from each of the two Lupinus species were carefully removed from the pots, and the roots washed free of sand. The stem of each seedling was placed in the center of a gray foam disk, which formed the center lid of a 4-L black plastic container with continuously aerated nutrient solution of the following composition: 400 μm NO 3, 200 μm Ca 2+, 210 μm K +, 154 μm SO 4 2, 54 μm Mg 2+, 0.24 μm Mn 2+, 0.1 μm Zn 2+, μm Cu 2+, 2.4 μm H 3 BO 3, 0.03 μm Mo 4+, 10 μm Fe-EDTA (ph 5.8). The plants were grown in individual containers half-immersed in a temperature-controlled rootcooling tank maintained at C. The experiment was carried out in a UWA glasshouse in a factorial, completely randomized design. Factor A was day length treatments (6, 10, or 14 h d 1 ) and factor B was species. Seven to eight seedlings of each species were used as a group with the same day length treatment. In total, 86 plants (including a few spare plants) formed 12 groups containing three different day length treatments. During these experiments, the average day/night temperature was 26 /19 C, the average day length was approximately 14 h, the average relative humidity was 52%, and the average midday ambient light level was approximately 1200 μmol photons m 2 s 1. The nutrient solution of each container was replaced daily. Cotyledons were removed 14 d after sowing to reduce the transfer of seed P reserves to seedlings. At 14 d after sowing, day length treatments (6, 10, or 14 h d 1 ) were started. Inverted cardboard boxes, which were partly open at the ends, were used to cover the plants for the 6 h and 10 h day length treatments. These cardboard boxes were designed with two flaps on the two opposite ends, which facilitated airflow and exchange and minimized any possible impacts on atmospheric conditions, temperature, and humidity in the experiment, but prevented light entering the boxes. Since the aim was to modify the growth rate of the plants, any effects in addition to those of day length did not compromise the approach. Natural day length in November and mid-december in Perth is h. To restrict the day length to set periods, the boxes were placed at the end of the natural daylight period at 19:00, and removed the next day at 09:00 (for the 10 h treatment) and 13:00 (for the 6 h treatment). For the 14-h treatment, plants were not covered by cardboard boxes. In our preliminary experiment, we used 20-L black plastic tubs with 12 plants in each tub supplied with 10 μm of P (in the form of KH PO4 2 ) for all day length treatments (6 h, 9 h, and 12 h). Plants had different relative growth rates (0.12, 0.15, and 0.17 mg g 1 d 1 ), and leaf P concentrations (8.4, 3.9, and 3.4 mg g 1 dry mass at 29 d after germination) at 6, 9, and 12 h, respectively. The higher plant P concentration of slower-growing plants caused by the shorter day length, reduced cluster-root production from 37% to 5% of total root dry mass. Therefore, we estimated the need and availability of P for plants under different day length treatments, based on the previous experiment, so P in the growth medium was not a major factor that strongly induced cluster-root formation. Thus, P concentration in the growth medium was adjusted to 3, 5, or 6 μm of P for the 6, 10, or 14 h day 1 treatment, respectively, to avoid a decline in leaf P concentration, which would induce cluster-root formation. Four plants of each species from each treatment were harvested at 22, 38, and 52 d after sowing. At the final harvest, day 52, we used only two replicates for L. pilosus, as the other two plants were either too small or too large. Plant measurements Four plants of all of the 86 harvested plants were gently removed from each pot and then rinsed with deionized water. For each plant, root clusters were separated from the rest of the root system and defined as portions of lateral roots containing bottlebrush-like root clusters of more than 10 rootlets/cm ( Johnson et al., 1996 ). Fresh mass of stems including petioles, leaves, and roots (cluster and noncluster) were determined at each harvest, and the dry mass was determined after drying at 70 C for 1 wk. Each of these plant parts was weighed before and after drying (see below), and the root mass ratio (root dry mass/ total plant dry mass) was

3 SEPTEMBER 2015, VOLUME 102 WANG ET AL. CLUSTER ROOTS, INTERNAL P, AND RGR IN LUPINUS 1531 calculated. Relative growth rates (RGR) based on total plant dry mass were calculated using the equation ( Lambers and Poorter, 1992 ) as follows: ln M ln M RGR =, T T1 where M 2 is the plant dry mass at day 52, or day 38, M 1 is the plant dry mass at day 38, or day 22, and T is the number of days of plant growth. Dried material was then ground into a fine powder using a stainless steel ball mill. A subsample ( mg) of recently fully expanded leaves (RFEL) of each Lupinus species from each harvest was taken and digested using a hot concentrated HNO 3 HClO4 (3:1) acid mixture. The solutions became clear when the digestion was completed. Samples in the digests were then analyzed for total P concentration by the malachite green colorimetric method ( Motomizu et al., 1983 ) using a UV VIS spectrophotometer at a wavelength of 630 nm (Shimadzu, Kyoto, Japan). Statistics Data were compared using a one-way ANOVA followed by a least significant difference (LSD) test ( α = 0.05) using Statistix 8.1 (Analytical Software, Tallahassee, Florida, USA). Means are presented with standard errors to indicate the variation of each measurement. Means were tested for significant differences at P Multiple regressions were carried out and results were plotted in three-dimensional graphs using general linear models (GLMs) using the program R version ( R Development Core Team, 2009 ) and the Scatterplot3d Package version ( Ligges and Mächler, 2003 ) in exploring patterns between combinations of cluster-root formation and both P concentration in RFEL and RGR in Lupinus species. For statistical comparisons, differences were considered statistically significant at P RESULTS Plant growth A longer day length resulted in more biomass for both Lupinus species, showing statistically significant differences at all three harvests ( Fig. 1 ). At the third harvest (day 52), L. albus and L. pilosus produced significantly more biomass at the longest day length (14 h), approximately 49% and 21% more than at a day length of 10 h, and 5-fold and 2.5-fold more than that at a 6-h day length, respectively ( F 2,7 = 197, P < 0.001, F 2,8 = 109, P < 0.001, Fig. 1 ). Day length treatments induced differences in RGR. Over the 16 d between 22 and 38 d after sowing, relative growth rates (RGR) were 83, 140, and 141 mg g 1 d 1 for L. albus and 103, 137, and 131 mg g 1 d 1 for L. pilosus at a day length of 6, 10, and 14 h, respectively. Over the 14 d between 38 and 52 d after sowing, the RGRs were 22, 42, and 55 mg g 1 d 1 for L. albus and 16, 43, and 46 mg 1 d 1 for L. pilosus at a day length of 6, 10, and 14 h, respectively. The longest day length (14 h) resulted in a higher RGR, which was 70% and 27% higher for L. albus and 150% and 188% higher for L. pilosus at a 6-h day length from day 22 to 38 and from day 38 to 52, respectively. The root mass ratio (RMR) increased significantly with increasing day length for L. albus in the last two harvests ( F 2,10 = 44, P < 0.05, F 2,9 = 11.2, P < 0.05), and for L. pilosus in the final harvest ( F 2,8 = 14, P < 0.001); the root mass ratio slightly decreased over time for FIGURE 1 Total plant dry biomass of Lupinus albus and L. pilosus at (A) 22 d, (B) 38 d, and (C) 52 d after sowing, when grown at a day length of 6 h (black), 10 h (gray), or 14 h (white); plants were grown in a hydroponic culture supplied with either 3, 5, or 6 μm of P (in the form of KH 2 ) for the respective day length treatments. Error bars represent standard error ( n = 4, with the exception of L. pilosus at day 52, where n = 2). Treatment means marked with the same letter are not significantly different within each group using a one-way ANOVA for each harvest followed by LSD at P both species. At 52 d after sowing, the RMR at a 14-h day length was 37% and 26% greater than at a 10-h day length, and 100% and 16% greater than at a 6-h day length for L. albus and L. pilosus, respectively ( Fig. 2 ).

4 1532 AMERICAN JOURNAL OF BOTANY after sowing ( Fig. 3 ), constituting an average of 16% and 11% of total root mass for L. albus and L. pilosus, respectively. On day 52, L. albus showed the highest percentage of cluster roots at a day length of 14 h, which was 2.3- and 10.5-fold greater than at a 10-h and at a 6-h day, respectively ( F 2,9 = 15.7, P < 0.05,); the percentage of cluster roots in L. pilosus at a 14-h day ( F 2,9 = 8.27, P < 0.05,) was 2- and 14-fold greater than that at a 10 and 6 h day length, respectively ( Fig. 3 ). Cluster-root investment was suppressed at shorter day lengths. Leaf phosphorus concentration We aimed to achieve very similar leaf P concentrations in plants for all the day length treatments, by providing slightly more P to plants grown at longer days. However, this supplement was only partly effective. Both Lupinus species had significantly lower P concentrations at the longer day lengths at the later harvests. At 22 d after sowing, P concentrations in recently fully expanded leaves (RFEL) at different day lengths were the same ( Fig. 4 ). On day 52, RFEL P concentrations at a day length of 14 h was 20% and 56% lower than at a day length of 10 and 6 h in L. albus, and 10% and 44% lower, respectively, in L. pilosus ( F 2,10 = 15.1, P < 0.05, F 2,10 = 15.1, P < 0.05, respectively, Fig. 4 ). Trends of RFEL P concentrations at day 38 were similar to those at day 52 ( Fig. 4 ). Phosphorus concentrations in RFEL at each day length decreased with time for these Lupinus species ( Fig. 4 ). Relationship between the percentage of cluster roots and leaf P concentration There was a significant negative correlation between the percentage of cluster roots and P concentration in RFEL for L. albus (F 4,14 = 26.5, P 0.001) and L. pilosus (F 5,13 = 41.4, P = 0.015) (Fig. 5 ). The strongest inhibition of cluster roots by P concentration in the RFEL was in L. albus at day 52. At this time, cluster-root percentage decreased to less than 3% when P concentration in the RFEL increased to about 6 mg P g 1 dry mass (DM) ( Fig. 5 ). The percentage of cluster roots in this species was as low as 5% at 3 mg P g 1 DM compared with 15% and 12% in L. albus and L. pilosus, respectively, at the same P concentration in the RFEL ( Fig. 5 ). Cluster-root allocation was suppressed at a higher internal P concentration. FIGURE 2 Root mass ratio (ratio of root dry mass to total plant dry mass) of Lupinus albus and L. pilosus at (A) 22 d, (B) 38 d, and (C) 52 d after sowing, when grown at a day length of 6 h (black), 10 h (gray), or 14 h (white); plants were grown in a hydroponic culture supplied with either 3, 5, or 6 μm P (in the form of KH 2 ) for each day length treatment, respectively. Error bars represent standard error ( n = 4, except for L. pilosus at day 52 [ n = 2]). Treatment means marked with the same letter are not significantly different within each group using a one-way ANOVA for each harvest followed by LSD at P Cluster-root percentage In contrast to the small effect on total root investment ( Fig. 2 ), the two Lupinus species produced a much greater percentage of cluster roots at the longest day (14 h) at 38 d Relationship between the percentage of cluster roots and relative growth rate According to a linear regression analysis, there was no correlation between the percentage of cluster roots and relative growth rate (RGR) between day 22 and 38 and between day 38 and 52 in L. albus and L. pilosus (Fig. 6 ). The percentage of cluster roots did not increase significantly with increasing RGR for the two Lupinus species. Relationship between the percentage of cluster roots and both leaf P concentration and relative growth rate Based on multiple regression analyses, there were strong and significant negative effects of P concentration in RFEL ( F 2,6 = 49.3, P and F 2,7 = 54, P for both L. albus and L. pilosus ), and small positive effects of RGR ( F 2,6 = 49.3, P = and F 2,7 = 54, P = for L. albus and L. pilosus, respectively) on the percentage of cluster roots for L. albus and L. pilosus ( Fig. 7 ). We plotted these relations using data on the percentage of cluster roots at day 38 and 52 and data on both P concentrations in RFEL and RGR for the respective growth intervals (days and 38 52) for both species.

5 SEPTEMBER 2015, VOLUME 102 WANG ET AL. CLUSTER ROOTS, INTERNAL P, AND RGR IN LUPINUS 1533 FIGURE 3 Percentage of cluster roots based on the total root mass of Lupinus albus and L. pilosus at (A) 22 d, (B) 38 d, and (C) 52 d after sowing, when grown at a day length of 6 h (black), 10 h (gray), or 14 h (white); plants were grown in a hydroponic culture supplied with either 3, 5, or 6 μm P (in the form of KH 2 ) for each day length treatment, respectively. The percentages were calculated as dry mass of cluster roots divided by total root dry mass times 100. Error bars represent standard error ( n = 4, except for L. pilosus at day 52 [ n = 2]). Treatment means marked with the same letter are not significantly different within each group using a oneway ANOVA for each harvest followed by LSD at P FIGURE 4 Phosphorus (P) concentration of recently fully expanded leaves of Lupinus albus and L. pilosus at (A) 22 d, (B) 38 d, and (C) 52 d after sowing, when grown at a day length of 6 h (black), 10 h (gray), or 14 h (white); plants were grown in hydroponic culture supplied with either 3, 5, or 6 μm P (in the form of KH 2 ) for each day length treatment, respectively. Error bars represent standard error ( n = 4). Treatment means marked with the same letter are not significantly different within each organ group using a one-way ANOVA for each tissue for all there harvests followed by LSD at P 0.05.

6 1534 AMERICAN JOURNAL OF BOTANY FIGURE 5 Relationship between percentage of cluster roots of total root dry mass and phosphorus (P) concentration in recently fully expanded leaves (RFEL) for Lupinus albus and L. pilosus. The solid lines indicate significant regressions ( P 0.05). DISCUSSION The objective of the current study was to investigate whether the negative correlation between investment in cluster roots and plant internal P status is maintained if differences in growth rate and plant P status are induced by a treatment other than P supply, namely, day length. The results show that cluster-root allocation of both Lupinus species was downregulated at a high shoot P status, irrespective of their growth rate. Day length treatments induced species-specific variation in relative growth rate and cluster-root formation for two Lupinus species There were major differences in cluster-root formation for the two Lupinus species in response to day length, which induced differences in RGR. Less biomass accumulated at shorter day lengths, but the decrease was slight relative to the allocation to the total root mass at the final harvest. Similarly, only small responses of root mass ratio to P supply have been reported for L. albus ( Keerthisinghe et al., 1998 ; Nuruzzaman et al., 2006 ). In striking contrast with the minor differences in biomass allocation to the FIGURE 6 Relationship between percentage of cluster roots of total root dry mass at day 38 and relative growth rate (RGR, between day 22 and day 38) and that at day 62 and RGR (between day 38 and day 52) for Lupinus albus and L. pilosus. The broken line indicates no significant regression. root system as a whole, a much greater percentage of cluster roots formed on L. albus and L. pilosus at longer day lengths at day 38 and day 52, compared with plants grown at shorter day lengths. On the basis of a preliminary experiment, we estimated the P requirement for plants under different day length treatments, so P in the growth medium was not a major treatment that strongly induced cluster-root formation. By doing so, we aimed to achieve a similar leaf P concentration at different day lengths and avoid cluster root decline in leaf P concentration at a longer day length and higher RGR. Varying the day length resulted in statistically significant differences in plant RGR. We cannot fully exclude that this approach may have altered an endogenous light-sensing mechanism, as shown in Arabidopsis thaliana in which a circadian rhythm affects the expression of most of the PHT4 family of P transporters ( Guo et al., 2008 ; Wang et al., 2011 ). This approach to vary day length, in turn might affect P acquisition and leaf P concentration differentially depending on the day length. Furthermore, because light intensity affects photosynthetic carbon (C) gain and C supply to cluster roots in L. albus ( Cheng et al.,

7 SEPTEMBER 2015, VOLUME 102 WANG ET AL. CLUSTER ROOTS, INTERNAL P, AND RGR IN LUPINUS 1535 correlated with plant P status, but only marginally and positively with RGR for both L. albus and L. pilosus (Fig. 7 ). This finding agrees with that in our previous report ( Wang et al., 2013 : fig. 3) showing that cluster-root percentage is highly significantly and negatively correlated with plant P status rather than with RGR. Cluster-root formation in the two Lupinus species as dependent on leaf P concentration and relative growth rate Cluster-root formation was significantly correlated with leaf P concentration for both Lupinus species, with a stronger and negative relationship in L. albus and a weaker but also negative correlation in L. pilosus. These findings are in line with previous reports for L. albus (Abdolzadeh et al., 2010 ; Shane et al., 2003 ; Shen et al., 2003 ). A low cluster-root allocation was closely correlated with a higher leaf P concentration in both species. However, because a plant s RGR is correlated with its leaf P concentration, these relationships might imply effects of RGR. However, there were no correlations between cluster-root allocation and RGR for both Lupinus species. When both leaf P concentration and RGR are considered via multiple regression, although RGR and leaf P concentration varied, allocation of biomass to cluster roots was strongly dependent on leaf P concentration and only weakly correlated with RGR. Greater leaf P concentrations strongly suppressed cluster-root formation and higher RGR weakly stimulated cluster-root formation. Taken together, cluster-root formation was downregulated by higher leaf P concentrations, irrespective of RGR. Our results support the finding that cluster-root formation as induced by a low P supply involves signaling in L. albus ( Marschner et al., 1987 ; Shane et al., 2003 ); the same is true for other P-starvation responses in other species ( Doerner, 2008 ; Hammond and White, 2008 ; Jungk, 2001 ; Liu et al., 2010 ). Differences in cluster-root formation between the two Lupinus species cannot be interpreted by speciesspecific differences in leaf P concentration or RGR. Since we did not investigate signaling at a transcriptional level, we can only speculate on the causes of variation in cluster-root allocation in the Lupinus species we investigated. The exact mechanism accounting for the difference of signaling in regulating cluster-root formation in Lupinus species under P starvation needs further investigation. FIGURE 7 Relationship between percentage of cluster roots of total root dry mass and both phosphorus (P) concentration in recently fully expanded leaves (RFEL) between day 22 and day 38. The multiple regressions were made using relative growth rate (RGR between day 22 and day 38), and RFEL [P] at day 38; and also RGR between day 38 and day 52 and RFEL [P] at day 52 for Lupinus albus and L. pilosus. Significant regressions: *** P 0.001, **P 0.01, *P ), day length may also have affected cluster-root formation and leaf P concentration. Our aim to achieve different growth rates but the same leaf P concentrations at different day lengths was only partially achieved because different day lengths did yield different RGRs, but not the same leaf P concentrations. Since the aim to modify the growth rate of the plants was achieved, any effects on leaf P concentration in addition to those on growth rate did not compromise our approach. Thus, we carried out a multiple regression analysis. This analysis showed that the percentage of cluster roots was strongly and negatively Cluster-root formation in different Lupinus species The present results show variation in cluster-root formation between the two Lupinus species but only on day 22. However, this variation cannot be accounted for by a greater leaf P concentration, because L. pilosus, which produced fewer cluster roots, did not invariably show greater leaf P concentrations. In the current study and a previous one ( Wang et al., 2013 ), L. albus invested a considerable fraction of its root biomass in cluster roots, as long as the leaf P status was low. This result is similar to the findings for the same species in the literature ( Shane et al., 2003 ; Shen et al., 2003 ; Pearse et al., 2006 ; Abdolzadeh et al., 2010 ). Similarly, L. pilosus in the present and previous studies ( Wang et al., 2013 ) showed a total suppression of cluster-root biomass investment at greater plant P concentrations, similar to L. luteus ( Pearse et al., 2006 ). In the current study, rates of cluster-root formation were higher at a longer day length. Longer day lengths or higher light intensity will enhance photosynthate production, which probably leads to carbohydrate signaling and sensing as a systemic control of plant responses to P deficiency. This result agrees with findings by Cheng et al. (2014), who showed that in L. albus P-deficiencyinduced cluster-root formation and citrate release are stimulated

8 1536 AMERICAN JOURNAL OF BOTANY by high light intensity, presumably due to increased carbohydrate signaling. Cluster-root formation in different Lupinus species as dependent on a range of other signals Cluster-root formation in these two Lupinus species may involve a complex series of signaling cascades controlling transcription and initiating plant responses to P starvation. An increased sucrose concentration in the leaf upregulates genes encoding transport proteins to transfer organic carbon and sucrose to the phloem, which helps the movement of these compounds to the roots ( Hermans et al., 2006 ; Hammond and White, 2008 ). Shoot-derived sugar signals (sucrose, glucose, and fructose) control plant P-starvation responses, including cluster-root formation ( Liu et al., 2005 ; Müller et al., 2007 ; Zhou et al., 2008 ). Hence, the strength of this signal might vary among lupine species. In addition, auxins play a role in cluster-root formation ( Gilbert et al., 2000 ; Skene and James, 2000 ; Hocking and Jeffery, 2004 ), so there may be differences in the strength of this hormonal signal between the present Lupinus species. Possibly, strigolactones, whose production increases under P deficiency ( Yoneyama et al., 2007 ; López-R á ez et al., 2008 ; Kohlen et al., 2011 ), play a role as well in differences in cluster-root formation among Lupinus species, given their role in lateral root formation and in the regulation of shoot architecture in response to P deficiency ( Kapulnik et al., 2011 ). Also, as discussed, micrornas act as a systemic low-p signal ( Zhu et al., 2010 ), and this signal might also vary among Lupinus species. Future research is warranted to characterize species-specific variation in systemic signal compounds to determine possible consequences for variation in cluster-root formation among species as demonstrated in our study. CONCLUSIONS Cluster-root-forming L. albus and L. pilosus grown at a higher growth rate, as dependent on light availability, enhanced their biomass investment in cluster roots. We conclude that cluster-root formation is invariably downregulated in plants with high P status, irrespective of growth rate. For L. albus, cluster-root investment was strongly and significantly suppressed at higher leaf P concentrations. In L. pilosus, there was also a strong negative correlation between cluster-root allocation and leaf P concentration, but the correlation was weaker than for L. albus. Thus, variation in leaf P concentration or RGR cannot account for species-specific variation in cluster-root investment between the two Lupinus species. ACKNOWLEDGEMENTS X.W. was supported by the University of Western Australia and the China Scholarship Council for UWA China scholarships and by the School of Plant Biology for project funding. We are grateful to Bevan Buirchell, Richard Snowball, and Jon C. Clements for kindly providing the Lupinus seeds, to Ray Scott at the UWA Combined Workshop, to Rob Creasy and Bill Piasini at the School of Plant Biology s Plant Growth Facility for their great help with the glass house experiment, to Susan Barker for providing materials for the experiment, to Mabel Fabiola Delgado Torres and Hiroaki Matsuoka for help with experiments, to Hai Ngo and Elizabeth Halladin for providing glassware, to Jing Zhang and Yan Liu for assisting with harvests, to Greg Cawthray for help with HPLC, to Laura Firth, François Teste, and Allah Ditta for helpful advice on the statistics, and to Joanne Edmondston and many friends for their support and encouragement. LITERATURE CITED Abdolzadeh, A., X. Wang, E. J. Veneklaas, and H. Lambers Effects of phosphorus supply on growth, phosphate concentration and cluster-root formation in three Lupinus species. Annals of Botany 105 : Bolland, M. D. A., K. H. M. Siddique, S. P. Loss, and M. J. Baker Comparing responses of grain legumes, wheat and canola to applications of superphosphate. Nutrient Cycling in Agroecosystems 53 : Cheng, L., X. Tang, C. P. Vance, P. J. White, F. Zhang, and J. Shen Interactions between light intensity and phosphorus nutrition affect the phosphate-mining capacity of white lupin ( Lupinus albus L.). Journal of Experimental Botany. doi: /jxb/eru135 Cordell, D., J.-O. Drangert, and S. White The story of phosphorus: Global food security and food for thought. Global Environmental Change 19 : Dinkelaker, B., C. Hengeler, and H. Marschner Distribution and function of proteoid rests and other root clusters. Botanica Acta 108 : Doerner, P Phosphate starvation signaling: A threesome controls systemic Pi homeostasis. Current Opinion in Plant Biology 11 : Fixen, P. E., and A. M. Johnston World fertilizer nutrient reserves: A view to the future. Journal of the Science of Food and Agriculture 92 : Gilbert, G. A., J. D. Knight, C. P. Vance, and D. L. Allan Proteoid root development of phosphorus deficient lupin is mimicked by auxin and phosphonate. Annals of Botany 85 : Gilbert, N The disappearing nutrient. Nature 461 : Guo, B., S. Irigoyen, T. B. Fowler, and W. K. Versaw Differential expression and phylogenetic analysis suggest specialization of plastid-localized members of the PHT4 phosphate transporter family for photosynthetic and heterotrophic tissues. Plant Signaling & Behavior 3 : Hammond, J. P., and P. J. White Sucrose transport in the phloem: Integrating root responses to phosphorus starvation. Journal of Experimental Botany 59 : Hermans, C., J. P. Hammond, P. J. White, and N. Verbruggen How do plants respond to nutrient shortage by biomass allocation? Trends in Plant Science 11 : Hocking, P. J., and S. Jeffery Cluster-root production and organic anion exudation in a group of old-world lupins and a new-world lupin. Plant and Soil 258 : Holford, I. C. R Soil phosphorus: Its measurement and its uptake by plants. Australian Journal of Soil Research 35 : Johnson, J. F., C. P. Vance, and D. L. Allan Phosphorus deficiency in Lupinus albus (altered lateral root development and enhanced expression of phosphoenolpyruvate carboxylase). Plant Physiology 112 : Jungk, A Root hairs and the acquisition of plant nutrients from soil. Journal of Plant Nutrition and Soil Science 164 : Kapulnik, Y., P. M. Delaux, N. Resnick, E. Mayzlish-Gati, S. Wininger, C. Bhattacharya, N. Séjalon-Delmas, et al Strigolactones affect lateral root formation and root-hair elongation in Arabidopsis. Planta 233 : Karthikeyan, A. S., D. K. Varadarajan, A. Jain, M. A. Held, N. C. Carpita, and K. G. Raghothama Phosphate starvation responses are mediated by sugar signaling in Arabidopsis. Planta 225 : Keerthisinghe, G., P. J. Hocking, P. R. Ryan, and E. Delhaize Effect of phosphorus supply on the formation and function of proteoid roots of white lupin ( Lupinus albus L.). Plant, Cell & Environment 21 : Khasawneh, F., E. Sample, and E. Kamprath Management considerations for acid soils with high phosphorus fixation capacity. In F. E. Khasawneh, E. C. Sample, and E. J. Kamprath [eds.], The role of phosphorus in agriculture, American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Madison, Wisconsin, USA.

9 SEPTEMBER 2015, VOLUME 102 WANG ET AL. CLUSTER ROOTS, INTERNAL P, AND RGR IN LUPINUS 1537 Kohlen, W., T. Charnikhova, Q. Liu, R. Bours, M. A. Domagalska, S. Beguerie, F. Verstappen, et al Strigolactones are transported through the xylem and play a key role in shoot architectural response to phosphate deficiency in nonarbuscular mycorrhizal host Arabidopsis. Plant Physiology 155 : Lambers, H., J. C. Clements, and M. N. Nelson How a phosphorusacquisition strategy based on carboxylate exudation powers the success and agronomic potential of lupines ( Lupinus, Fabaceae). American Journal of Botany 100 : Lambers, H., and H. Poorter Inherent variation in growth rate between higher plants: A search for physiological causes and ecological consequences. Advances in Ecological Research 23 : Lambers, H., J. A. Raven, G. R. Shaver, and S. E. Smith Plant nutrientacquisition strategies change with soil age. Trends in Ecology & Evolution 23 : Lambers, H., M. W. Shane, M. D. Cramer, S. J. Pearse, and E. J. Veneklaas Root structure and functioning for efficient acquisition of phosphorus: Matching morphological and physiological traits. Annals of Botany 98 : Li, H., J. Shen, F. Zhang, C. Tang, and H. Lambers Is there a critical level of shoot phosphorus concentration for cluster-root formation in Lupinus albus? Functional Plant Biology 35 : Ligges, U., and M. Mächler Scatterplot3d An R package for visualizing multivariate data. Journal of Statistical Software 8 : Liu, J. Q., D. L. Allan, and C. P. Vance Systemic signaling and local sensing of phosphate in common bean: Cross-talk between photosynthate and microrna399. Molecular Plant Pathology 3 : Liu, J. Q., D. A. Samac, B. Bucciarelli, D. L. Allan, and C. P. Vance Signaling of phosphorus deficiency-induced gene expression in white lupin requires sugar and phloem transport. Plant Journal 41 : López-Ráez, J. A., T. Charnikhova, V. Gómez-Roldán, R. Matusova, W. Kohlen, R. De Vos, F. Verstappen, et al Tomato strigolactones are derived from carotenoids and their biosynthesis is promoted by phosphate starvation. New Phytologist 178 : Lynch, J. P Roots of the second green revolution. Australian Journal of Botany 55 : Marschner, H., V. Romheld, and I. Cakmak Root-induced changes of nutrient availbility in the rhizosphere. Journal of Plant Nutrition 10 : Motomizu, S., T. Wakimoto, and K. Toei Spectrophotometric determination of phosphate in river waters with molybdate blue and malachite green. Analyst (London) 108 : Müller, R., M. Morant, H. Jarmer, L. Nilsson, and T. H. Nielsen Genomewide analysis of the Arabidopsis leaf transcriptome reveals interaction of phosphate and sugar metabolism. Plant Physiology 143 : Nacry, P., G. Canivenc, B. Muller, A. Azmi, H. Van Onckelen, M. Rossignol, and P. Doumas A role for auxin redistribution in the responses of the root system architecture to phosphate starvation in Arabidopsis. Plant Physiology 138 : Nuruzzaman, M., H. Lambers, M. D. A. Bolland, and E. J. Veneklaas Distribution of carboxylates and acid phosphatase and depletion of different phosphorus fractions in the rhizosphere of a cereal and three grain legumes. Plant and Soil 281 : Pearse, S. J., E. J. Veneklaas, G. R. Cawthray, M. D. A. Bolland, and H. Lambers Carboxylate release of wheat, canola and 11 grain legume species as affected by phosphorus status. Plant and Soil 288 : Péret, B., M. Clement, L. Nussaume, and T. Desnos Root developmental adaptation to phosphate starvation: Better safe than sorry. Trends in Plant Science 16 : R Development Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Raghothama, K. G Phosphate acquisition. Annual Review of Plant Physiology and Plant Molecular Biology 50 : Richardson, A. E., J. P. Lynch, P. R. Ryan, E. Delhaize, F. A. Smith, S. E. Smith, P. R. Harvey, et al Plant and microbial strategies to improve the phosphorus efficiency of agriculture. Plant and Soil 349 : Runge-Metzger, A Closing the cycle: Obstacles to efficient P management for improved global food security. In H. Teissen [ed.], Phosphorus in the global environment: Transfers, cycles and management, Wiley, Chichester, UK. Scholz, R. W., and F.-W. Wellmer Approaching a dynamic view on the availability of mineral resources: What we may learn from the case of phosphorus? Global Environmental Change 23 : Shane, M. W., M. de Vos, S. de Roock, and H. Lambers Shoot P status regulates cluster-root growth and citrate exudation in Lupinus albus grown with a divided root system. Plant, Cell & Environment 26 : Shen, J., Z. Rengel, C. Tang, and F. Zhang Role of phosphorus nutrition in development of cluster roots and release of carboxylates in soil-grown Lupinus albus. Plant and Soil 248 : Simpson, R. J., A. Oberson, R. A. Culvenor, M. H. Ryan, E. J. Veneklaas, H. Lambers, J. P. Lynch, et al Strategies and agronomic interventions to improve the phosphorus-use efficiency of farming systems. Plant and Soil 349 : Skene, K. R., and W. M. James A comparison of the effects of auxin on cluster root initiation and development in Grevillea robusta Cunn. ex R. Br. (Proteaceae) and in the genus Lupinus (Leguminosae). Plant and Soil 219 : Stutter, M. I., C. A. Shand, T. S. George, M. S. A. Blackwell, R. Bol, R. L. Mackay, A. E. Richardson, et al Recovering phosphorus from soil: A root solution? Environmental Science & Technology 46 : Wang, X., S. J. Pearse, and H. Lambers Cluster-root formation and carboxylate release in three Lupinus species as dependent on phosphorus supply, internal phosphorus concentration and relative growth rate. Annals of Botany 112 : Wang, Y., J.-F. Wu, N. Nakamichi, H. Sakakibara, H.-G. Nam, and S.-H. Wu LIGHT-REGULATED WD1 and PSEUDO-RESPONSE REG- ULATOR9 from a positive feedback regulatory loop in the Arabidopsis circadian clock. Plant Cell Online 23 : Watt, M., and J. R. Evans Proteoid roots. Physiology and development. Plant Physiology 121 : Yoneyama, K., Y. Takeuchi, and H. Sekimoto Phosphorus deficiency in red clover promotes exudation of orobanchol, the signal for mycorrhizal symbionts and germination stimulant for root parasites. Planta 225 : Zhou, K., M. Yamagishi, M. Osaki, and K. Masuda Sugar signalling mediates cluster root formation and phosphorus starvation-induced gene expression in white lupin. Journal of Experimental Botany 59 : Zhu, Y. Y., H. Q. Zeng, C. X. Dong, X. M. Yin, Q. R. Shen, and Z. M. Yang MicroRNA expression profiles associated with phosphorus deficiency in white lupin ( Lupinus albus L.). Plant Science 178 :