Plant and Soil 253: 413–427, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands. 413 The interaction of phosphorus and potassium with seed alkaloid concentrations, yield and mineral content in narrow-leafed lupin (Lupinus angustifolius L.) P. Gremigni1,5,6, J. Hamblin1,2, D. Harris3 & W. A. Cowling4 1 Centre for Legumes in Mediterranean Agriculture, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia. 2 Export Grains Centre Ltd, 219 Canning Hwy, South Perth, WA 6151, Australia. 3 Chemistry Centre (WA), 125 Hay Street, East Perth, WA 6004, Australia. 4 School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia. 5 CSIRO, Plant Industry, Centre for Environment and Life Sciences, Private Bag 5 Wembley, WA 6913, Australia. 6 Corresponding author∗ Received 31 July 2002. Accepted in revised form 3 February 2002 Key words: lupanine, narrow-leafed lupin, phosphorus, P–K interaction, potassium, quinolizidine alkaloid Abstract We tested the impact of P deficiency, K deficiency, and their interaction on seed alkaloid concentrations and profile, yield and mineral content in sweet (low-alkaloid) and bitter (high-alkaloid) varieties of narrow-leafed lupin (Lupinus angustifolius L.). P deficiency reduced seed alkaloid concentrations in sweet, but not in bitter, varieties. Under P deficiency, the alkaloid profile in harvested seed of sweet varieties mimicked that of the bitter variety Fest, with 13-hydroxylupanine dominating over lupanine. With adequate or abundant P, lupanine was the predominant alkaloid in sweet varieties. K deficiency was associated with an 8-fold increase of seed alkaloid concentrations in the sweet variety Danja (from 1000 to 8000 mg kg−1 DM), mostly due to the stimulation of lupanine production. There was a significant interaction between P and K that affected seed alkaloid concentrations in two ways: (i) the inhibitory effect of P deficiency was only apparent under K deficiency and (ii) the lowest seed alkaloid concentrations occurred with abundant K (240 mg K kg−1 ) and P (60 mg P kg−1 ). Seed yield of all varieties increased asymptotically with increasing P and reached a maximum at adequate P (30 mg P kg−1 ). There was no impact of K deficiency on seed yield. In sweet and bitter varieties P supply increased seed N, P and Zn concentrations, but not K. In contrast, seed K concentrations increased and P concentrations decreased with increasing K supply. These findings suggest that P fertiliser should be supplemented with K, to avoid high seed alkaloid concentrations stimulated by asymptomatic K deficiency at high P levels. Introduction Many native soils in Australia are severely P-deficient (Bolland, 1998). After conversion to agricultural land over the past 150 years, P deficiency has been ameliorated with phosphate fertilisers (Rengel, 1998). However, P deficiency still occurs, due to removal of P with crops and animal products (Bolland, 1987), leaching from poor sandy soils (Bolland, 1994; Bolland et al., ∗ FAX No: +61-8-9387 8991. E-mail: patrizia.gremigni@csiro.au 1996; He et al., 1998) or P-fixation in heavier soil types (Bolland and Gilkes, 1998; Cornforth, 1997). Conversely, native soils had adequate K, but K deficiency is now an increasing problem especially in high rainfall areas or on sandy high-leaching soils (Bolland et al., 2002). Narrow-leafed lupin (Lupinus angustifolius L.) requires abundant P at early growth stages (Longnecker et al., 1998). When grown on P-deficient soils, narrowleafed lupin has low levels of P in plant tissues, greatly reduced shoot growth and pod set, no lateral branch formation (Snowball and Robson, 1986) and 414 extremely poor seed yield (Bolland, 1987). Similarly, K deficiency reduces early shoot growth and lateral branch formation in narrow-leafed lupin (Snowball and Robson, 1986), and ultimately causes seed yield losses (Cox, 1978), by reducing photosynthetic rates and increasing respiration rates (Marschner, 1995). K fertiliser may alleviate K deficiency symptoms but does not stimulate plant growth when applied too late after seed germination (Yeates, 1988). Narrow-leafed lupin seed is a highly nutritious component of animal feed because of its mineral composition, high protein and fibre concentrations (Petterson et al., 1997). Quinolizidine alkaloids are the main antinutritional factors that limit its use for animal feed and human consumption (Lowen et al., 1995). Seed alkaloid concentration is regulated by the major gene iucundus (iuc) and is generally low (≤200 mg kg−1 DM) in sweet varieties (homozygous for the iuc allele), or high (≥10 000 mg kg−1 DM) in bitter varieties (dominant Iuc allele). The average seed alkaloid concentration of commercial varieties is strictly regulated and in narrow-leafed lupin seed for food or feed it must be <200 mg kg−1 DM (ANZFA, 2001; Culvenor and Petterson, 1986). However, large and unpredictable fluctuations of seed alkaloid concentrations were found in sweet narrow-leafed lupin in different locations in the same year or the same location over different years (Gremigni, 2002; Harris, 1994). A wide range of environmental stresses, including soil nutritional imbalances, is likely to stimulate alkaloid production in lupin (Eilert, 1998; Waller and Nowacki, 1978). Seed alkaloid concentrations in sweet varieties of L. angustifolius increased under K deficiency (Gremigni et al., 2001), and similar effects of K deficiency were previously reported in a sweet variety of L. albus (Scibor-Marchocka, 1970). In contrast, P deficiency was associated with lower alkaloid concentrations in the same L. albus variety (Scibor-Marcocka, 1970). The aim of this research was to examine the effect of P nutrition on bitter and sweet varieties of narrowleafed lupins under controlled glasshouse conditions and to determine if there was an interaction between P and K deficiency that may partly explain the observed fluctuations in lupin seed alkaloid concentrations in southern Australia (Gremigni, 2002; Harris, 1994). Seed yield and mineral content were also measured. The hypothesis was that P deficiency would be associated with reduced seed alkaloid concentrations and that there would be an interaction between P and K deficiency on seed alkaloid concentrations. Materials and methods Two glasshouse experiments were conducted in 1996 and 1998 with the bitter narrow-leafed lupin variety Fest and the sweet varieties Yorrel, Gungurru and Danja following standard glasshouse procedures outlined in Gremigni et al. (2001). Experiment 1. Response of bitter and sweet varieties of narrow-leafed lupin to P The soil used in this pot experiment was Lancelin sand as described in Gremigni et al. (2001). This soil was from natural bushland and typical of the P-deficient coastal sandy soils of the south west of Western Australia. The sand contained a total concentration of 28 mg P kg−1 and a NaHCO3 -extractable P of 2 mg kg−1 (Rayment and Higginson, 1992), which is extremely low and results in severe P deficiency symptoms in a range of crop species such as maize (Moody et al., 1997) and wheat (Elliott et al., 1997). The concentration of other nutrients (mg kg−1 soil) was as described in Gremigni et al. (2001). Soil was air-dried, sieved through a 2-mm stainless steel sieve and thoroughly mixed as described in Gremigni et al. (2001) before placing 6 kg of soil into each plastic pot (19 cm height × 19.5 cm diameter) lined with a polythene bag. Basal nutrients were applied using AR grade salts in deionized water and added in solution to each pot as described in Gremigni et al. (2001). Four levels of P (7.5, 15, 30, 60 mg P kg−1 soil) were added as Ca(H2 PO4 )2 ·H2 O solution. The ‘adequate’ level was considered to be 30 mg P kg−1 soil, equivalent to 45 kg P ha−1 in the field (Bolland et al., 2001). Different amounts of solid CaSO4 were added to maintain a constant Ca2+ supply between the low and high P treatments. After air-drying the soil in the pots, the nutrients were thoroughly mixed throughout the soil by shaking the soil of each pot in a plastic jar. No nitrogen (N) was applied. Soil was watered to field capacity (11% oven-dried soil weight) and allowed to settle for 3 days before sowing. A factorial design of four P levels and four varieties of narrow-leafed lupin was used for the experiment, with pots as experimental units. Pots were arranged in a randomised block design with four replicates on the glasshouse bench. 415 Ten lupin seeds, sieved to a minimum diameter of 5.5 mm and surface-sterilised as in Gremigni et al. (2001), were sown in each pot on 26 July 1996. Each seed was inoculated with 1 mL suspension of Bradyrhizobium sp. (Lupinus) strain WU425. Three weeks after sowing, plants were thinned to three plants per pot. The soil surface of each pot was covered with alkathene beads to limit soil water evaporation, and watered daily to field capacity. Throughout the experiment, all pots within each replicate were rerandomized once a week to minimise possible effects due to the position of the pots on the bench. At flowering on the main stem (approx. 56 days after sowing [DAS] for Yorrel, 59 for Danja and Gungurru and 93 DAS for Fest), main stem length and leaf number on the main stem were recorded for each plant in each pot. At maturity (120 DAS for Danja, Gungurru and Yorrel and 160 DAS for Fest), plants were cut at ground level. Pods and seeds were collected separately from the main stem and uppermost branches (Dracup and Kirby, 1996) from each plant. All plant parts were air-dried in the glasshouse (max. temperature 60 ◦ C) for 6 days. The air-dry weight and number of pods and seeds, as well as air-dry weight of the remainder of the plant parts above ground, were recorded. Leaves lost from each treatment before final harvest were not measured. Total seed yield, above-ground plant weight and harvest index (HI) (the ratio of total seed yield to above-ground plant weight) were recorded for each plant. The three plants harvested from each pot were bulked for chemical analyses. The dried seed material was ground and analysed for total and individual alkaloid and mineral (N, K, P, Ca, Mg, S, Cu, Zn) concentrations as in Gremigni et al. (2001). same rates as in experiment 1. Four levels of K (0, 15, 60 and 240 mg K kg−1 soil) and four levels of P (15, 20, 30, 60 mg P kg−1 soil) were added as K2 SO4 and Ca (H2 PO4 )2 ·H2 O solutions, respectively. The ‘adequate’ level of K was considered to be 60 mg kg−1 (Tang, 1998) and P was 30 mg kg−1 (Bolland et al., 2001) equivalent to 90 kg K ha−1 and 45 kg P ha−1 in the field, respectively. CaCl2 ·2H2 O (solid) was added to balance Ca2+ ions. Nutrients were mixed throughout the soil, soil was watered as in experiment 1 and allowed to settle for 3 days before sowing. Pots were arranged in a randomised block design with four replicates in a root cooling tank at 20±2 ◦ C. Danja seeds were sieved and sterilised as in experiment 1. Ten seeds were sown in each pot on 1 September 1998 and inoculated each with 1 mL suspension of Bradyrhizobium sp. (Lupinus) strain WU425. About 2 weeks after sowing, plants were thinned to three plants per pot and pots were covered with alkathene beads. Throughout the experiment, all pots were rerandomised as in experiment 1. At flowering (49 DAS) main stem length and leaf number were recorded for each plant. At maturity (95 DAS) above ground plant parts, pods and seeds from main stem and uppermost branches from each plant were collected and air-dried as in experiment 1. The dry weight and number of pods and seeds, as well as dry weight of the remainder of the plant parts above ground were recorded. Seed yield and harvest index (HI) were recorded, and the air-dried seed material from three plants per pot was bulked, ground and analysed for alkaloid and mineral concentrations as in experiment 1. Experiment 2. The interaction of P and K supply on seed alkaloids in the sweet narrow-leafed lupin variety Danja The air-dried seed material was milled, extracted (Harris and Wilson, 1988) and analysed for alkaloids by capillary gas chromatography (Priddis, 1983) as described by Gremigni et al. (2001). The variety Danja was chosen because it was the most responsive of the sweet varieties for seed alkaloid concentrations to both K (Gremigni et al., 2001) and P supply (experiment 1). Also, in the bitter narrowleafed lupin variety Fest, neither K (Gremigni et al., 2001) nor P supply (Gremigni et al., 2000; experiment 1) affected seed total alkaloid concentrations. The experimental units were pots filled with 6 kg of Lancelin soil as described in experiment 1. Basal nutrients were added in solution to each pot at the Alkaloid analysis Nitrogen Total nitrogen (N) concentrations in the ground material were measured on a LECO, FP-428, using the Dumas combustion technique (AOAC, 1999). 416 Mineral composition The ground seed material was extracted and analysed for minerals by ICP-AES as described by Gremigni et al. (2001). Statistical analysis The data were analysed using GenStat
5 Release 4.2 statistical software (GenStat Committee, 2000). The relationships between P level and seed total and individual alkaloid concentrations, total plant biomass, seed yield, HI, and level of K applied were modelled by the standard asymptotic curve y = a+b·ekx , where y represents the variable studied, x is the level of P applied, a is the asymptote, b is the range, a + b is the intercept and k is the rate parameter of the fitted curve (Gremigni et al., 2001). Comparisons between means are based either on the standard error of means or the least significant differences at the 0.05 probability level (l.s.d. 0.05 ). Results Experiment 1. Response of bitter and sweet varieties of narrow-leafed lupin to P supply Seed total alkaloids Due to lack of seed production under severe P deficiency (7.5 mg P kg−1 ), it was not possible to apply the asymptotic curve model to describe trends of total alkaloid concentrations in all varieties. The late-flowering bitter variety Fest failed to develop upper branches and upper branch seed, and the sweet varieties Danja, Gungurru and Yorrel produced insufficient upper branch seed for alkaloid analysis at low P levels. Therefore total plant seed alkaloid concentrations are presented as weighted averages of total alkaloid concentrations in seed from the main stem and upper branches (Figure 1). Lupin varieties had different responses in total alkaloid concentrations to P (p < 0.001, ANOVA not presented). The bitter variety Fest produced main stem seed with extremely high total alkaloid concentrations (25 000–30 000 mg kg−1 DM) and was not responsive to P (Figure 1A). Total alkaloid concentrations were highest in Danja and lowest in Yorrel (Figure 1B– D) and there was no significant interaction between variety and P in these varieties (ANOVA not presented). Total alkaloid concentrations of sweet varieties increased from 15 to 30 mg P kg−1 soil and did not increase above 30 mg P kg−1 soil. Seed alkaloid profile In the bitter variety Fest, 13-hydroxylupanine was the dominant alkaloid at all P levels (about 15 000 mg kg−1 DM, 53% of total) (Figure 1A,E), with lower concentrations of lupanine and angustifoline. Sweet varieties had a similar alkaloid profile to Fest under P deficiency (15 mg P kg−1 ), but lupanine increased at 30 and 60 mg P kg−1 and became the predominant alkaloid at abundant P (Figure 1F–H). Angustifoline (Figure 1E–H) and α-isolupanine (data not shown) contributed 18 and 5.6%, respectively, of total alkaloids in the bitter variety Fest, and 10–14 and <1.5% in the sweet varieties. Plant biomass at harvest As confirmed by the analysis of variance using the standard asymptote model, total plant biomass at harvest increased (p < 0.001) as P increased (Figure 2A). Of all varieties, Yorrel was the least responsive to P (Figure 2A). Seed yield and HI Seed yield and HI of all varieties increased with applied P (p < 0.001) (Figure 2B,C). Fest had the lowest seed yield and HI. The HI of Yorrel was the highest due to its lower biomass (Figure 2C). Danja had a significantly different P-response curve for HI than the other varieties (Figure 2C). HI was overestimated, because defoliated leaves were not considered in the evaluation of plant biomass at maturity. Seed yield components Very few pods and seeds were produced under severe P deficiency (7.5 mg P kg−1 ). The average number of pods, seeds, and seeds per pod increased with increasing P applied (p < 0.001, ANOVA not presented) (Table 1). Yorrel had the lowest number of seeds per pod, but produced the largest seeds on main stem (Table 1). Individual seed weight increased with addition of P (Table 1). Individual seed weight was generally lower (p < 0.001) in seed from upper branches (average 0.102 g) than in seed from main stem (average 0.149 g). Individual pod weight increased both on main stem and upper branches (p < 0.001) with increasing levels of P (Table 1). 417 Figure 1. Experiment 1. (A, B, C, D) Concentrations (mg kg−1 DM) of total alkaloids, lupanine, 13-hydroxylupanine and angustifoline, and (E, F, G, H) relative proportions of lupanine, 13-hydroxylupanine and angustifoline (% of total alkaloid concentrations) in seed from the bitter narrow-leafed lupin variety Fest and the sweet varieties Yorrel, Gungurru and Danja, grown at different levels of applied P (15, 30, 60 mg kg−1 ). Values are means of three replicates. Bars are ± standard errors of means. DM = dry matter based on air-dry weight of seed. In Fest seed was available from main stem only, but in the sweet varieties seed was harvested separately from main stem and upper branches, and the values presented here are the weighted averages of main stem and upper branch seed. 418 Seed mineral composition (N, K, P, Ca, Mg, S, Cu, Zn) Seed mineral analysis was not possible in upper branch seed of any variety at the P deficiency treatments of 7.5 and 15 mg P kg−1 due to failure to develop upper branches or to produce sufficient seed for chemical analysis. N In all varieties seed N concentrations increased (p < 0.001) with increasing P. Main stem seed had higher N concentrations than upper branch seed (Table 2). K, P, Ca, Mg and S In the sweet varieties, seed K was not affected by P and was slightly higher (p < 0.001) in upper branch than in main stem seed (9.8 vs. 9.2 g kg−1 ). Fest had the lowest seed K (average 6.8 g kg−1 ) (Table 2). Seed P increased (p < 0.001) in all varieties as applied P increased (Table 2). Seed Ca decreased as P increased (Table 2). Seed Mg and S concentrations were not affected by P supply (Table 2). In the bitter variety Fest, seed S concentrations were 30% higher than in sweet varieties. Cu, Zn Seed Cu decreased (p < 0.001) and Zn increased (p < 0.001) with increasing P (Table 2). Experiment 2. Response of the sweet narrow-leafed lupin variety Danja to P and K supply Figure 2. Experiment 1. Plant biomass at harvest, (A), seed yield (B), and HI (C) of seed from main stem of the bitter variety Fest and from main stem and upper branches of the sweet varieties Yorrel, Gungurru and Danja of narrow-leafed lupin, grown at four levels of applied P (7.5, 15, 30 and 60 mg kg−1 ). Values are means of four replicates. Bars indicate ± standard errors of observations. Biomass measurements are based on air-dry weight. Variance accounted for by model was 97% for plant biomass, 97.8% for seed yield, and 89.2% for HI. Seed total alkaloids Trends for total alkaloid concentrations were similar for seed from main stem and upper branches and results are reported together as weighted averages for the whole plant (Figure 3). A major drop in total seed alkaloid concentrations occurred as P decreased, but only under severe (0 mg K kg−1 ) and mild (15 mg K kg−1 ) K deficiency (Figure 3A). Under K deficiency, the alkaloid response to P reached an asymptote at 60 mg P kg−1 applied (Figure 3A). The significant changes in slope and asymptote with level of K applied indicate a strong interaction between K and P in the total alkaloid concentrations measured in seed (Figure 3A). Seed alkaloid profile Lupanine was the predominant alkaloid but its relative contribution to total alkaloid concentrations decreased with increasing K and decreasing P (Figure 3E, ANOVA not shown). The opposite trend was observed for 13-hydroxylupanine (Figure 3F) and angustifoline 419 Table 1. Experiment 1. Seed yield components of main stem of Fest and main stem and upper branches of Yorrel, Gungurru and Danja grown at four levels of P applied (7.5, 15, 30, 60 mg kg−1 soil). Values are means of four replicates and are expressed as air-dry weight (Wt) on a per plant basis P applied (mg kg−1 ) Pod no. Seed no. Seed/pod Individual pod Wt (g) Individual seed Wt (g) 7.5 15 30 60 NA 1.6 4.3 4.9 NA 4.2 15.6 19.8 NA 2.6 3.5 4.1 NA 0.29 1.34 1.66 NA 0.091 0.124 0.152 Yorrel 7.5 15 30 60 NA 2.2 3.3 4.7 NA 5.2 9.6 15.3 NA 2.5 2.9 3.2 NA 0.49 1.02 1.43 NA 0.142 0.181 0.197 Gungurru 7.5 15 30 60 1.7 3.7 3.9 4.8 1.7 12.0 15.2 19.3 1.0 3.2 3.9 4.0 0.12 1.03 1.53 1.67 0.047 0.110 0.163 0.176 Danja 7.5 15 30 60 1.0 3.1 3.9 4.2 1.2 10.2 15.2 18.0 0.9 3.4 3.8 4.2 0.04 0.89 1.47 1.47 0.025 0.118 0.163 0.169 (Variety × P) applied l.s.d. 0.05 1.2 3.3 0.6 0.26 0.022 15 30 60 1.7 3.5 4.9 2.6 8.6 14.8 1.4 2.4 3.0 0.33 0.73 1.12 0.095 0.126 0.153 Gungurru 15 30 60 0.7 3.4 4.5 0.8 9.1 13.4 0.8 2.6 3.0 0.08 0.75 1.06 0.049 0.111 0.145 Danja 15 30 60 1.0 3.0 5.2 1.0 9.0 18.6 1.1 3.0 3.6 0.08 0.61 1.52 0.063 0.098 0.157 0.7 2.5 0.4 0.19 0.015 L .angustifolius variety Main stema Fest Upper branchesb Yorrel (Variety × P applied) l.s.d. 0.05 a Under severe P deficiency (7.5 mg kg−1 soil) Fest and Yorrel either did not survive or produced no pods and seeds on main stem. b Under severe P deficiency (7.5 mg kg−1 soil) plants of all varieties did not develop upper branches and upper branch seed. Fest failed to develop upper branches and upper branch seed at all P levels. NA – Not available. (Figure 3G). α-Isolupanine represented only 1–2% of seed total alkaloids and its concentrations were highest under severe K deficiency, especially when adequate P was also applied (maximum 39 mg kg−1 DM). At the other K levels α-isolupanine concentration was about 15 mg kg−1 DM, independent of P (data not shown). Plant biomass at harvest Plant biomass at harvest increased (p < 0.001) with applied P and there was a strong interaction with applied K (Figure 4A). The highest biomass production (11.6 g plant−1 ) was observed with adequate K (60 mg kg−1 ) and abundant P (60 mg kg−1 ) levels (Figure 4A). 420 Figure 3. Experiment 2. Total alkaloid (A), lupanine (B), 13-hydroxylupanine (C) and angustifoline (D) concentrations (mg kg−1 DM) and relative proportions of lupanine, 13-hydroxylupanine and angustifoline (% of total alkaloid concentrations) (E, F, G and H) in seed of the sweet narrow-leafed lupin variety Danja grown at four levels of K (0, 15, 60, 240 mg kg−1 ) and four levels of P (15, 20, 30, 60 mg kg−1 ) applied. Values are means of three replicates. Bars indicate ± standard errors of observations in A, B, C and D, or ± standard errors of means (SE) in E, F, G and H. Seed was harvested separately from main stem and upper branches and the values presented here are the weighted averages of main stem and upper branch seed. DM = dry matter based on air-dry weight of seed. Variance accounted for by model was 83.2% for total alkaloids, 78.4% for lupanine, 46.0% for 13-hydroxylupanine, and 66.4% for angustifoline. 421 Table 2. Experiment 1. Mineral (N, K, P, Ca, Mg, S, Cu, Zn) concentrations in seed from main stem of Fest and seed from both main stem and upper branches of Yorrel, Gungurru and Danja grown at four levels of P applied (7.5, 15, 30, 60 mg kg−1 soil). Values are means of three replicates and are based on the air-dry weight of seed P applied (mg kg−1 ) N 15 30 60 47 50 63 Yorrel 15 30 60 Gungurru Danja L. angustifolius variety Main stema Fest Mineral concentration Ca Mg P S (g kg−1 ) Cu Zn (mg kg−1 ) 7.0 6.8 6.6 3.6 2.6 2.6 1.7 1.7 1.8 2.8 2.9 4.1 4.4 4.7 4.3 6.0 4.6 2.9 39 44 45 52 60 62 9.5 9.6 9.5 3.3 2.8 2.6 1.8 1.8 1.7 2.5 3.1 4.1 3.1 3.0 3.1 6.0 4.8 3.5 45 50 54 15 30 60 48 58 61 9.3 8.8 9.1 2.7 2.6 2.4 1.6 1.6 1.7 2.3 2.8 3.9 3.5 3.5 3.4 4.2 4.1 3.1 44 49 54 15 30 60 49 55 58 9.6 9.1 9.3 2.8 2.6 2.6 1.8 1.7 1.7 2.4 2.7 3.5 3.7 3.4 3.2 5.1 4.2 3.4 42 47 54 4 0.7 0.4 0.2 0.4 0.5 0.5 3 (Variety × P applied) l.s.d. 0.05 K Upper branchesa Yorrel 30 60 49 57 10.0 10.0 3.0 2.6 1.7 1.6 2.8 3.8 3.8 3.7 4.7 3.6 48 50 Gungurru 30 60 50 54 9.1 10.0 2.9 2.6 1.5 1.5 2.4 3.6 3.8 4.0 3.7 3.2 48 51 Danja 30 60 59 57 9.8 9.8 2.5 2.3 1.6 1.7 2.6 3.9 3.9 3.6 3.7 3.3 46 52 2 0.4 0.2 0.1 0.3 0.2 0.4 2 (Variety × P applied) l.s.d. 0.05 a Seed mineral analysis was not possible in upper branch seed of any variety under P deficiency (7.5 and 15 mg P kg−1 soil) due to failure to develop upper branches (Fest) or to produce sufficient seed (Yorrel, Gungurru and Danja) for chemical analysis. Seed yield and Harvest Index (HI) Total seed yield increased (p < 0.001) with increasing applied P levels, but was not affected by K. For this reason the asymptotic model predicts the same curve for all K levels (Figure 4B). HI was not greatly affected by P or K and the model accounted for only 17.3% of the variance in this trait (Figure 4C). Seed yield components The number of pods and seeds increased with P. Main stem pods contained an average of three to four seeds, while each upper branch pod had two to three seeds, regardless of K and P levels (Table 3). The size of pods and seeds on upper branches was about one third lower than that on the main stem (Table 3) and increased with P, independent of K (Table 3). Seeds were significantly (p < 0.001) smaller on the upper branches than on the main stem and seed size decreased (p < 0.001) on both the main stem and upper branches with increasing K (Table 3). Seed mineral composition (N, K, P, Ca, Mg, S, Cu, Zn) N Seed N concentrations increased with increasing P levels, except when abundant K was also applied. K, P, Ca, Mg and S Seed K concentrations increased (p < 0.001) with increasing K supply and reached their maximum with abundant P and K. Seed P concentrations increased (p < 0.001) with applied P and were slightly higher in main stem than in upper branch 422 Table 3. Experiment 2. Seed yield components of main stem and upper branches of Danja grown at four levels of K (0, 15, 60, 240 mg kg −1 soil) and four levels of P (15, 20, 30, 60 mg kg−1 soil) applied. Values are means of four replicates. Pod and seed weight are expressed as air-dry weight (Wt) on a per plant basis K applied P (mg kg−1 ) Pod no. Seed no. Seed/pod Individual pod Wt (g) Individual Seed Wt (g) Main stem 0 0 0 0 15 20 30 60 3.7 3.4 4.4 4.6 15.3 17.0 15.3 17.0 4.0 4.0 3.5 3.7 1.34 1.31 1.28 1.30 0.155 0.165 0.154 0.153 15 15 15 15 15 20 30 60 3.6 4.2 4.3 4.9 15.1 16.4 16.7 19.3 4.3 3.9 3.8 4.0 1.20 1.37 1.23 1.15 0.134 0.144 0.140 0.135 60 60 60 60 15 20 30 60 3.7 3.7 5.0 5.8 12.2 15.0 19.1 22.3 3.3 4.1 3.9 3.9 1.11 1.36 1.43 1.37 0.129 0.148 0.138 0.145 240 240 240 240 15 20 30 60 3.9 4.4 6.1 5.1 15.6 16.9 22.8 18.9 4.0 3.8 3.7 3.7 1.40 1.31 1.67 1.14 0.125 0.117 0.127 0.131 Upper branches 0 0 0 0 15 20 30 60 2.4 2.8 2.8 2.8 5.4 5.7 5.4 5.7 1.6 2.3 2.2 2.1 0.40 0.49 0.52 0.49 0.130 0.128 0.153 0.141 15 15 15 15 15 20 30 60 2.6 3.9 4.3 4.2 6.0 8.4 11.8 12.0 2.3 2.1 2.7 2.9 0.42 0.60 0.83 0.74 0.103 0.106 0.133 0.118 60 60 60 60 15 20 30 60 1.9 2.9 4.5 4.4 4.0 7.8 9.9 12.2 2.0 2.8 2.2 2.6 0.30 0.55 0.65 0.75 0.087 0.108 0.104 0.130 240 240 240 240 15 20 30 60 2.0 2.6 3.3 3.3 4.1 6.1 8.1 9.7 2.1 2.2 2.2 2.8 0.33 0.45 0.49 0.57 0.088 0.089 0.090 0.103 1.3 4.8 0.9 0.31 0.025 (Branch × K applied × P applied) l.s.d. 0.05 423 seed (data not shown). Seed Ca concentrations were generally stable regardless of the level of P and K applied, and lower in main stem than in upper branch seed (data not shown). Seed Mg and S concentrations were not affected by P supply. However, Mg increased (p < 0.001) with abundant K, and S increased with K (K was applied as K2 SO4 ) and was higher in upper branch than in main stem seed (data not shown). Cu, Zn Seed Cu decreased with applied P and K. Seed Zn increased (p < 0.001) with P and decreased (p < 0.001) with K (data not shown). Discussion Seed total alkaloids Figure 4. Experiment 2. Plant biomass at harvest (A), seed yield (B) and HI (C) of seed from main stem and upper branches of the sweet narrow-leafed lupin variety Danja, grown at four levels of P applied (15, 20, 30 and 60 mg kg−1 ) and four levels of K applied (0, 15, 60 and 240 mg kg−1 ). Values are means of four replicates. Bars indicate ± standard errors of observations. In B the model indicates no significant differences in asymptote, range and intercept of the four K levels, and a single curve is fitted. In C the model indicated no response of HI to P at 0 and 15 mg K kg−1 and could be fitted only on HI values at 60 and 240 mg K kg−1 applied. Biomass measurements are based on air-dry weight. Variance accounted for by model was 28.8% for plant biomass, 20.3% for seed yield, and 17.3% for HI. P deficiency decreased total alkaloid concentrations in seed of sweet narrow-leafed lupin varieties. There was an asymptotic response to applied P in seed alkaloids, reaching a plateau at abundant P levels. P deficiency had no effect on the high seed alkaloid concentrations of the bitter variety Fest. As expected from previous findings (Gremigni et al., 2001), K deficiency increased total seed alkaloid concentrations in sweet lupin varieties (an 8-fold increase at 0 compared with 240 mg K kg−1 soil). However, there was a significant interaction between K and P – the response of seed alkaloid concentrations to P deficiency was only evident when K was also deficient. This interaction resulted in the lowest seed alkaloid concentrations in sweet varieties when abundant K (240 mg K kg−1 ) and P (60 mg P kg−1 ) were applied together. In Fest, seed total alkaloid concentrations were always high and independent of P and K supply, supporting our previous results (Gremigni et al., 2001). The reduction in seed alkaloid concentrations observed in the sweet varieties Yorrel, Gungurru and Danja under P deficiency may be explained by a depression in enzymatic activities, including the enzymes involved in alkaloid biosynthesis, due to the key role that P plays in controlling the levels of plant enzymes (Marschner, 1995). The same effect was not present in the bitter variety Fest. In legumes, soil P deficiency reduces shoot growth indirectly by inhibiting N fixation, and directly by stimulating the allocation of carbon and nutrient resources to root rather than shoot growth, to allow better root exploration of soil for P (Marschner, 1995). P deficiency also inhibits photosynthesis (Marschner, 1995), and there appears to be 424 a positive correlation between rates of photosynthesis and lupin alkaloid biosynthesis (metabolic processes that both occur in the leaf chloroplasts) in L. polyphyllus (Wink and Hartman, 1982). Hence, P deficiency may depress alkaloid production in sweet narrowleafed lupins through reduced plant photosynthetic rates. In two separate experiments, Scibor-Marchocka (1970) observed that seed alkaloid concentrations in a sweet variety of L. albus decreased with high K or low P supply, which matches our results in L. angustifolius. We may speculate why P deficiency reduces seed alkaloid concentrations in sweet lupin varieties but not in bitter varieties. In our experiments this effect was only observed under the alkaloid-stimulatory conditions of K deficiency. Our hypothesis is that the Iuc allele and downstream protein products are highly active and not inhibited by P or K deficiency. In contrast, the iuc allele products normally have a low activity but are stimulated by K deficiency, and it is only then that the inhibitory effect of P deficiency is expressed. The iuc allele is well known for its pleiotropic effect of increasing sensitivity to Mn deficiency in sweet narrow-leafed lupin varieties (Walton and Francis, 1975). Seed alkaloid profile In both experiments, the proportion of lupanine in seeds of sweet varieties decreased and that of 13hydroxylupanine increased under P deficiency. In experiment 1, P deficiency changed the alkaloid profile of the three sweet varieties to resemble that of the bitter variety Fest (Figure 1) which was dominated by 13hydroxylupanine. In addition, K deficiency stimulated lupanine production 8-fold (Figure 3B), but caused only a doubling of 13-hydroxylupanine concentration under adequate P (Figure 3C). If our hypothesis concerning allele products of iuc is correct, then the iuc allele products primarily affect lupanine production. In contrast, the Iuc allele products favour the 13hydroxylupanine accumulation and are not influenced by P and K deficiency. This research has demonstrated that alkaloid profile may vary with genotype and environment, and therefore alkaloid profile may not be a good criterion for lupin taxonomy, in agreement with previous observations (Wink et al., 2000). Plant biomass and seed yield In both experiments, P supply was the main factor governing plant biomass and seed production, as ex- pected from plant nutrition studies (Marschner, 1995). Reduced growth is a common feature of plants when exposed to deficiencies of nutrients associated with plant metabolism (e.g., N, P and K) (Lambers et al., 1998), but these experiments confirm that asymptomatic K deficiency may cause a major increase in total seed alkaloid concentrations (Figure 3A), without a clear impact on seed yield or biomass (Figure 4A,B). Asymptomatic K deficiency would not be detected in commercial lupin crops and could explain some of the large variation in seed alkaloids in field-grown lupins in Western Australia (Gremigni, 2002; Harris, 1994). Most lupin crops are fertilized with P but not K in Western Australia, and our experiments suggest that regular application of K is necessary to avoid high seed alkaloid concentrations in P-fertilised crops. Seed yield, HI, seed size and seed number all increased with increasing P levels, in both bitter and sweet varieties. This is typical of plant responses to P supply in the field (Bolland and Jarvis, 1996; Bolland et al., 1993) and in the glasshouse (Ma et al., 2002): P increases the number of reproductive sinks, that is, seeds (Marschner, 1995). Seed mineral composition P supply increased N, P and Zn concentrations in seed of bitter and sweet narrow-leafed lupin varieties. N and P are nutrients associated with plant metabolism and their concentrations in plants are closely tied to soil availability of N and P (Lambers et al., 1998). For example, P supply increased seed P concentration in L. angustifolius (Ma et al., 2002) and the combined application of K (120 kg ha−1 ) and P (60 kg ha−1 ) increased N, P and K concentrations in plants of L. luteus (Marschner, 1995). The increase of seed Zn concentrations due to P supply was unexpected, since in most cases a reduction of Zn concentrations as a consequence of P application is observed. The increase of seed Zn concentrations observed in our glasshouse experiments was not due to use of different seed lots (Longnecker and Robson, 1993), or levels of P applied to the pot soil higher than those encountered in soil solutions (Loneragan and Webb, 1993). The sandy soil that we used was virtually deprived of nutrients, thus applied P did not result in P toxicity. However, we cannot rule out completely the Zn contamination of the P salt used in our nutrient solutions, since we used AR grade and not ultra pure salts. Seed K was not affected by P supply, in contrast to previous findings in the field (Jarvis and Harris, 1993), 425 where P fertilisation increased seed K concentrations in the variety Danja. Seed K was strongly associated with K supply, particularly in sweet varieties, which were more efficient than the bitter Fest in mobilising K to the seed. Although seed K is relatively stable in most species (Marschner, 1995), the concentration of K in lupins depends on K supply (Tang, 1998) and lupin seed may accumulate high concentrations of K when ‘luxurious’ amounts of K and adequate P are available (Gremigni et al., 2001; Liu and Longnecker, 2002). Stimulation of alkaloids in the glasshouse environment In the sweet variety Danja, the seed total alkaloid concentrations were in the same range (1440–2120 mg kg−1 DM) of those found in the same variety in experiment 1 (880–2000 mg kg−1 DM) at adequate K, whilst they increased dramatically (up to approx. 8000 mg kg−1 DM) when plants were grown under K deficiency in experiment 2. The combination of K deficiency and adequate P also resulted in high alkaloid concentrations (up to 12 000 mg kg−1 DM) in Danja in a previous glasshouse experiment (Gremigni et al., 2001). Total seed alkaloid concentrations of sweet narrow-leafed lupin varieties in our glasshouse experiments were generally much greater than ‘normal’ field levels for these varieties (Gremigni et al., 2001) and above 200 mg kg−1 DM (ANZFA, 2001; Culvenor and Petterson, 1986). Seed harvested from glasshousegrown bitter and sweet narrow-leafed lupin plants in these experiments was respectively about 2- and 10fold higher than the original seed. It is possible that glasshouse-specific stress factors may interact with nutrient stresses and stimulate lupin alkaloid production, especially in sweet varieties (expressing the iuc allele) (Gremigni et al., 2001). For instance, restricted volume for roots can impair nutrient uptake (BarTal, 1998; Izaguirre-Mayoral and Sicardi de Mallorca, 1999; Marschner, 1995). Considering that lupin roots can grow very deep, especially in sandy soils (Dracup and Kirby, 1996; Hamblin and Hamblin, 1985), the whole plant metabolism may be adversely affected when they are confined in small pots. Other legumes (Phaseolus vulgaris, Vigna unguiculata) grown in root restricted soil volumes had increased chlorophyll concentrations in mature leaves (Izaguirre-Mayoral and Sicardi de Mallorca, 1999). In lupin, free forms of quinolizidine alkaloids are synthesised in the chloro- plasts (Wink and Hartman, 1982) and esterified in the mitochondria (Suzuki et al., 1996). It is possible that alkaloid production in the chloroplasts was stimulated by stress caused either by root restriction volume or by high temperatures and light intensities occurring in the glasshouse. Furthermore, plants growing under nutrient K deficiency conditions are generally more susceptible to a wide range of environmental stresses and respond with increased secondary metabolite production (Marschner, 1995). We may speculate that the increased accumulation of alkaloids observed in sweet narrow-leafed lupin varieties under K deficiency may be the indirect effect of enhanced production of the lupin alkaloid precursor cadaverine. This has been observed in various organs of a restricted number of plant species including pea, and appears to be caused by abiotic injuries (Bouchereau et al., 1999) such as osmotic stress (Aziz et al., 1999; Lefèvre et al., 2001) and heat shock (Shevyakova et al., 2001). Conclusion In the glasshouse, K deficiency increased and P deficiency reduced seed alkaloid concentrations in sweet varieties of narrow-leafed lupins, and there was a significant interaction of K and P. When both K and P were abundant, seed alkaloid concentrations were lowest but, with abundant P, alkaloid concentrations increased dramatically in response to K deficiency. Alkaloid profiles of sweet varieties also changed in response to P and K deficiency, with a stimulation of lupanine under K deficiency and an increase of 13-hydroxylupanine under P deficiency. In the field, soil P fertiliser is applied regularly to lupin-growing soils and is unlikely to be deficient in most commercial lupin crops. In contrast, K fertiliser is not regularly applied and K availability may fluctuate greatly from year to year. We observed large increases in lupin seed alkaloids at K deficiency levels that were not low enough to cause significant loss of plant biomass or seed yield, or K deficiency symptoms. Asymptomatic K deficiency may be partly responsible for the unpredictable fluctuations of alkaloid concentrations in field-grown lupins from different locations (Gremigni, 2002) or even from paddocks on the same property (Harris, 1994). Adequate P and K fertilisation of lupins may help to prevent seed alkaloid concentrations rising above the current maximum acceptable limit of 200 mg kg−1 DM. Further 426 field experiments are required to verify this conclusion outside the glasshouse environment. We have confirmed previous observations (Gremigni et al., 2001) that the glasshouse environment causes exceptionally high seed alkaloid concentrations in bitter and sweet narrow-leafed lupins. However, in the sweet varieties abundant P and K in the glasshouse reduced seed alkaloid concentrations to within the range normally observed in the field (Gremigni, 2002). Acknowledgements The Grains Research and Development Corporation (GRDC) is acknowledged for funding this project (UWA 168) and the Chemistry Centre (WA) for providing facilities for the chemical analyses of plant samples. References ANZFA (Australian New Zealand Food Authority) Food Standards Code. [Homepage of ANZFA], [Online]. 2001, Dec. 1last update. Available: http://www.anzfa.gov.au/foodstandards/ oldfoodstandardscodecontents/partbcerealsandcerealproducts/ b1cerealslegumesf lou681.cfm [2002, Jan. 9]. AOAC 1999 Official Methods of Analysis of AOAC International. 16th Edition, 5th revision, Method 4.2.04 (968.06). Ed. P. Cunniff. p. 13. AOAC International, Gaithersburg, MA, USA. Aziz A, Martin-Tanguy J and Larher F 1999 Salt-stress induced proline accumulation and changes in tyramine and polyamine levels are linked to ionic adjustment in tomato leaf discs. Plant Sci. Shannon 145 (2), 83–91. Bar-Tal A 1998 The significance of root size for plant nutrition in intensive horticulture. In Mineral Nutrition of Crops. Fundamental mechanisms and implications. Ed. Z Rengel. pp. 115–139. Food Product Press, The Haworth Press Inc., Birghmanton, NY. Bolland M D A 1987 Phosphorus deficiency in narrow-leafed lupins. Farmnote no. 62/87, Agdex 161/632. Dept Agric. WA, Perth. Bolland M D A 1994 Residual value of superphosphate for oat and barley grown on a very sandy, phosphorus leaching soil. Fertil. Res. 38 (3), 171–181. Bolland M D A 1998 Plant nutrition: phosphorus. In Soil Guide. A handbook for understanding and managing agricultural soils. Ed. G Moore. pp. 168–175. Bulletin 4343. WA Dept Agric., Perth. Bolland M D A and Gilkes R J 1998 The chemistry and agronomic effectiveness of phosphate fertilisers. In Mineral Nutrition of Crops. Fundamental mechanisms and implications. Ed. Z Rengel. pp. 227–265. Food Product Press, The Haworth Press Inc., Birghmanton, NY. Bolland M D A and Jarvis R J 1996 Effectiveness of different methods of applying superphosphate for lupins grown on sandplain soils. Aust. J. Exp. Agric. 36, 707–715. Bolland M D A, Jarvis R J, Coates P and Harris D J 1993 Effect of phosphate fertilisers on the elemental composition of seed of wheat, lupin, and triticale. Commun. Soil Sci. Plant Anal. 24 (1516), 1991–2014. Bolland M D A, Yeates J S and Clarke M F 1996 Effect of fertiliser type, sampling depth, and years on Colwell soil test phosphorus for phosphorus leaching soils. Fertil. Res. 44 (3), 177–188. Bolland M D A, Sweetingham M W and Jarvis R J 2001 Effect of Pleiochaeta setosa on field responses of Lupinus angustifolius and L. luteus to applications of phosphorus. Aust. J. Exp. Agric. 41, 549–556. Bolland M D A, Cox W J and Codling B J 2002 Soil and tissue tests to predict pasture yield responses to applications of potassium fertiliser in high rainfall areas of south-western Australia. Aust. J. Exp. Agric. 42 (2), 149–164. Bouchereau A, Aziz A, Larher F and Martin-Tanguy J 1999 Polyamines and environmental challenges: recent development. Plant Sci. Limerick 140 (2), 103–125. Cornforth I 1997 Phosphate fertiliser supply and demand. Agric. Sci. May–June 1997, 38–39. Cox W J 1978 Potassium deficiency in lupins – identification, rates, times and method of application. J. Agric. West. Aust. 19, 27–31. Culvenor C C J and Petterson D S 1986 Lupin toxins – alkaloids and phomopsins. pp. 188–198. Proceedings of the 4th International Lupin Conference, August 1986. Perth, WA Dept Agric., Perth. Dracup M and Kirby E J M 1996 Lupin Delopment Guide. The University of Western Australia Press, Perth. 97 pp. Eilert U 1998 Induction of alkaloid biosynthesis and accumulation in plants and in vitro cultures in response to elicitation. In Alkaloids: Biochemistry, ecology and medicinal applications. Eds. M E Roberts and M Wink. pp. 219–238. Plenum, New York. Elliott D E, Reuter D J, Reddy G D and Abbott R J 1997 Phosphorus nutrition of spring wheat (Triticum aestivum L.) 1. Effects of phosphorus supply on plant symptoms, yield, components of yield, and plant phosphorus uptake. Aust. J. Agric. Res. 48, 855–867. GenStat Committee 2000 The guide to GenStat
– 2 Statistics, Ed. R W Payne. NAG Ltd., Oxford. Gremigni P 2002 Control of seed alkaloid levels in narrow-leafed lupins (L. angustifolius L.) through management of plant nutrition and agronomic practices. Ph. D. Thesis, The University of Western Australia, Perth. 234 pp. Gremigni P, Hamblin J and Harris D J 2000 Genotype × environment interactions and lupin alkaloids. Lupin – an ancient crop for the new millennium. Proceedings of the 9th International Lupin Conference, June 1999. Eds. E Van Santen, M Wink, S Weissman and P Römer. pp. 362–365. Klink-Müritz, Germany. ILA, Canterbury, NZ. Gremigni P, Wong M T F, Edwards N K, Harris D J and Hamblin J 2001 Potassium nutrition effects on seed alkaloid concentrations, yield and mineral content of lupins (L. angustifolius). Plant Soil 234, 131–142. Hamblin A P and Hamblin J 1985 Root characteristics of some temperate legume species and varieties on deep, free-draining entisols. Aust. J. Agric. Res. 36, 63–72. Harris D J 1994 Quality production and breeding. Proceedings of the 1st Australian Lupin Technical Symposium, October 1994. Eds. M Dracup and J Palta. pp. 98–105. Perth, WA Dept Agric., Perth. Harris D J and Wilson P E 1988 A rapid manual method of lupin alkaloid analysis. Proceedings of the 5th International Lupin Conference, July 1988. Ed. T Twardowski. pp. 598–601. Poznan, Poland. Izaguirre-Mayoral M L and Sicardi de Mallorca M 1999 Responses of Rhizobium- inoculated and nitrogen-supplied Phaseolus vulgaris and Vigna unguiculata plants to root volume restriction. Aust. J. Plant Physiol. 26, 613–623. 427 Jarvis R J and Harris D J 1993 Effect of phosphate fertilisers on the elemental composition of seed of wheat, lupin, and triticale. Commun. Soil Sci. Plant Anal. 24 (1516), 1991–2014. Lambers H, Stuart Chapin F III and Pons T L 1998 Plant Physiological Ecology. Springer, New York, Berlin, Heidelberg. 540 p. Lefèvre I, Gratia E and Lutts S Discrimination between the ionic and osmotic components of salt stress in relation to free polyamine level in rice (Oryza sativa). Plant Sci. Shannon 161 (5), 943–952. Liu F-H and Longnecker N 2002 Interactive effect of cytokinin and potassium on sink-source relationships in Lupinus angustifolius. Plant Growth Reg. 36, 1–6. Loneragan and Webb 1993 Interactions between zinc and other nutrients affecting the growth of plants. In Zinc in Soils and Plants. Ed. A Robson. pp. 119–131. Kluwer Academic Publishers, Dordrecht. Longnecker N and Robson A 1993 Distribution and transport of zinc in plants. In Zinc in Soils and Plants. Ed. A Robson. pp. 79–91. Kluwer Academic Publishers, Dordrecht. Longnecker N, Brennan R and Robson A 1998 Lupin nutrition. In Lupins as Crop Plants. Biology, production and utilization. Eds. J S Gladstones, C A Atkins and J Hamblin. pp. 121–148. CAB International, Cambridge, UK. Lowen R, Alam F K A and Edgar J A 1995 Lupin bean toxicity. Med. J. Aust. 162, 256–257. Ma Q, Longnecker N and Atkins C 2002 Varying phosphorus supply and development, growth and seed yield in narrow-leafed lupin. Plant Soil 239, 79–85. Marschner H 1995 Mineral Nutrition of Higher Plants. Academic Press, London. 889 pp. Moody P W, Dickson T and Aitken R L 1997 Soil phosphorus tests and grain responsiveness of maize (Zea mays) on Ferrosols. Aust. J. Soil Res. 35, 609–613. Petterson D S, Sipsas S and Mackintosh J B 1997 The chemical composition and nutritional value of Australian pulses. pp. 7–9. Grains Research & Development Corporation, Canberra, ACT, Australia. Priddis C R 1983 Capillary gas chromatography of lupin alkaloids. J Chromatogr. 261, 95–101. Rayment G E and Higginson F R 1992 Australian Laboratory Handbook for Soil and Water Chemical Methods. Australian Soil and Land Survey Handbook. Inkata Press, Melbourne. 330 pp. Rengel Z 1998 Physiological mechanisms underlying different nutritional efficiency of crop genotypes. In Mineral Nutrition of Crops. Fundamental mechanisms and implications. Ed. Z Rengel. pp. 227–265. Food Product Press, The Haworth Press Inc., Birghmanton, NY. Scibor-Marchocka A 1970 Comparative studies on the homologous types of bitter and fodder white lupine. Acta Agrobot. 23(1), 23– 38. Shevyakova N I, Rakitin V Y, Duong D B, Sadomov N G and Kuznetsov V V 2001 Heat shock-induced cadaverine accumulation and translocation throughout the plant. Plant Sci. 161 (6), 1125–1133. Snowball K and Robson A D 1986 Symptoms of nutrient deficiencies. Lupins. pp. 20–21. University of Western Australia Press, Perth, WA. Suzuki H, Koike Y, Murakoshi I and Saito K 1996 Subcellular localization of acyltransferases for quinolizidine alkaloid biosynthesis in Lupinus. Phytochemistry 42(6), 1557–1562. Tang C 1998 Factors affecting soil acidification under legumes I. Effect of potassium supply. Plant Soil 199, 275–282. Waller G R and Nowacki E K 1978 Alkaloid biology and metabolism in plants. Plenum Press, New York. 294 pp. Walton G H and Francis C M 1975 Genetic influences on the split seed disorder in Lupinus angustifolius L. Aust. J. Agric. Res. 26, 641–646. Wink M and Hartmann T 1982 Localisation of the enzymes of quinolizidine alkaloid biosynthesis in leaf chloroplasts of Lupinus polyphyllus. Plant Physiol. 70, 74–77. Wink M, Merino F and Käss E 2000 Molecular evolution of lupins (Leguminosae: Lupinus). In Lupin, an Ancient Crop for the New Millennium. Proceedings of the 9th International Lupin Conference, June 1999. Klink/Müritz, Germany. Eds. E Van Santen, M Wink and P Römer. pp. 287–290. ILA, Canterbury, NZ. Yeates J 1988 Potassium deficiency in lupins. Farmnote no. 19/88. WA Dept Agric., Perth. Section editor: B. Sattelmacher